APBRmetrics

The discussion of the analysis of basketball through objective evidence, especially basketball statistics.
It is currently Sat Nov 01, 2014 7:11 am

All times are UTC




Post new topic Reply to topic  [ 108 posts ]  Go to page Previous  1, 2, 3, 4, 5 ... 8  Next
Author Message
PostPosted: Fri Apr 22, 2011 2:18 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
94by50



Joined: 01 Jan 2006
Posts: 494
Location: Phoenix

PostPosted: Fri Mar 20, 2009 10:31 pm Post subject: Correlation between team age and exceeding expectations Reply with quote
When I was still updating the weekly bell curve standings, Mountain saw something that led me to research this subject. Specifically, he saw that five of the oldest teams in the league this season were all among the league's leaders in winning percentage over their bell curve expectation. To study this further, I took every team from the merger onward and figured their team age, weighted by minutes played. I then gathered the following data:

* Their actual wins
* Their expected wins based on the bell curve method (prorated to 82 games in the lockout year)
* For kicks, the team's home court advantage, as well as the league's HCA for that season

I'm just an amateur, so the ways in which I know to approach this question are limited, but hopefully some more knowledgeable than I can see this from other angles that I may miss.

First thing I see: there's an odd trend in the league home court advantage. I remember Bill Simmons making remarks about how home court "mattered" more in the past, and I think this is what he meant. The league HCA actually jumped from around 3.7 points per game in the early 80s, to a high of 5.84 in 1987-88. For comparison, over the last ten years or so, it's been somewhere in the 3s. I call it an odd trend because the overall pattern is slowly downward, but there's a sharp, temporary spike in the late 80s.

The main question: are older teams likely to outperform their actual winning percentage?

Test #1: I wouldn't expect a trend like this to be linear, but I ran a linear regression to start. Based on that, there's only a 6% connection (r^2 = .06) between team age, and how much the team exceeds its actual winning percentage.

x = team age, y = actual wins - expected wins
.4498x - 12.05 = y: r^2 = .0608

Adjustment: the average age of teams is slightly below 27 (26.957), so instead of raw team age, I'm subtracting 27. Second adjustment: I'm adding a second degree to the equation. I wonder if the effect peaks anywhere.

Test #2: a second-order regression, subtracting age from 27. Not a thing changes - 6% connection.

x = team age - 27, y = actual wins - expected wins
-.0009x^2 + .4552x + .1132 = y: r^2 = .0609

Just a cursory glance at the subject. It appears to me that age and/or experience has little if any measurable impact on a team's ability to outperform its expected winning percentage. What about this year, then? A fluke? I dunno. There's probably something else I should account for that I've not thought of yet.

I'm going to post the workbook on the Yahoo group, to make the dataset available. If anyone has any suggestions, they'd be much appreciated.
Back to top
View user's profile Send private message
Harold Almonte



Joined: 04 Aug 2006
Posts: 576


PostPosted: Fri Mar 20, 2009 11:19 pm Post subject: Reply with quote
What about the correlation between team age and expected injuries?
Back to top
View user's profile Send private message
Mountain



Joined: 13 Mar 2007
Posts: 1492


PostPosted: Fri Mar 20, 2009 11:54 pm Post subject: Reply with quote
Thanks for pursuing the subject 94by50.

If you wanted to go further, given the discrepancy between a very long timeframe and this season, maybe it would be good to look at last 5-10 years by itself? If it matches up more with this season then maybe we found a meaningful trend in the current phase. If it matches the long-term results then I'd be more ready to drop the notion. Maybe the surge of new blood in the 70s and 80s made youth more of a win producer back then? The period since the merger is so long it might contain several trends within it that might offset.

Another angle might to focus on just starters, players likely to be in at the end of the game where the close games (the main source of the actual - expected gap) are decided.
On a quick check of recent good teams I see at least 3 older key guys but don't know the average situation immediately.

Or check the correlation just for playoff teams. Maybe age only correlates strongly with teams that work, with obviously other terms of fit mattering too.
Back to top
View user's profile Send private message
94by50



Joined: 01 Jan 2006
Posts: 494
Location: Phoenix

PostPosted: Sat Mar 21, 2009 12:26 am Post subject: Reply with quote
Mountain wrote:
Thanks for pursuing the subject 94by50.

If you wanted to go further, given the discrepancy between a very long timeframe and this season, maybe it would be good to look at last 5-10 years by itself? If it matches up more with this season then maybe we found a meaningful trend in the current phase. If it matches the long-term results then I'd be more ready to drop the notion. Maybe the surge of new blood in the 70s and 80s made youth more of a win producer back then? The period since the merger is so long it might contain several trends within it that might offset.

Another angle might to focus on just starters, players likely to be in at the end of the game where the close games (the main source of the actual - expected gap) are decided.

I could start by examining the trend on a year-to-year basis. As for focusing on starters only, should I determine starters from games started? Total minutes? Or something else I don't have yet? Or all of the above?
Back to top
View user's profile Send private message
94by50



Joined: 01 Jan 2006
Posts: 494
Location: Phoenix

PostPosted: Sat Mar 21, 2009 12:34 am Post subject: Reply with quote
Harold Almonte wrote:
What about the correlation between team age and expected injuries?

Well, we would believe there to be such a relationship. As to this topic, would injuries (actual or predicted) affect a team's actual performance against its expectation? I've no clue.
Back to top
View user's profile Send private message
Mountain



Joined: 13 Mar 2007
Posts: 1492


PostPosted: Sat Mar 21, 2009 12:53 am Post subject: Reply with quote
Year by year would be comprehensive detective work. I am personally mainly interested in the recent time period but a longer detailed study would give more context if you are so motivated. I see basketball-reference has close to 30 years of games started information.


Would be interesting to check inter-relationship between wins, age and pace by year.

70s and 80s saw a more wide open door and a faster game than the 90s or recently.

Pistons / Bulls slowed things down. making easier for the old guys and old guys style & tricks?

Recent years a bit faster right? Is the best win age moving back from above peak back topwards 27? Lakers Cavs and Magic are right there by your numbers. Is speend being rewarded more? viewtopic.php?t=1962&start=45
3 of the 4 oldest teams slipped in wins and more so in contending status.
I think the Lakers aged up some from last season to get closer to the player and maybe team sweet spot.
Celtics not far off and a reminder that the age mix is worth looking at beyond the minutes weighted age.
How does 3 point shooting frequency and accuracy trend by age? Maybe that has been presented before? I am not going to look for it but maybe somebody else knows or knows where it is. Lots of factors. Does the best age for today vary by position? Maybe some.

The interplay of these things affect cap strategy, draft pick use (keep or trade), coaching, etc.

Last edited by Mountain on Sat Mar 21, 2009 2:32 am; edited 4 times in total
Back to top
View user's profile Send private message
94by50



Joined: 01 Jan 2006
Posts: 494
Location: Phoenix

PostPosted: Sat Mar 21, 2009 2:18 am Post subject: Reply with quote
Mountain wrote:
Year by year would be comprehensive detective work. I am personally mainly interested in the recent time period but a longer detailed study would give more context if you are so motivated. I see basketball-reference has close to 30 years of games started information.

As long as I'm working on it, why not study as much as possible? I see that BR has the games started data, but it isn't loaded into the player season finder, which will make gathering that data exponentially more time-consuming. I can limit the inclusion of players to only those in the top six or seven of minutes played for each team - it might be an effective substitute.
Back to top
View user's profile Send private message
Mountain



Joined: 13 Mar 2007
Posts: 1492


PostPosted: Sat Mar 21, 2009 2:27 am Post subject: Reply with quote
Minutes played more important than starter so using minutes in some fashion (top3, 5 or 7) sounds fine.


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 2:20 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
magicmerl



Joined: 30 Dec 2007
Posts: 27


PostPosted: Mon Mar 16, 2009 2:43 am Post subject: Records vs the top teams Reply with quote
This is an analysis I did last year to see who the best of the best are and attempt to forecast the eventual champion. All the data is pulled from BR's excellent game summary page:
http://www.basketball-reference.com/lea ... games.html

The first three columns are simply the teams win/loss/pct vs the entire league. Teams with a win% of .500 or better are probably a good approximation of a playoff team.

The second trio of columns are those same three columns vs teams with a .500 or better record. These teams are ones expecting to advance in the playoffs.

The third trio are teams record vs the teams who plan on advancing in the playoffs. There are only 4 'Elite' teams who can win consistently against the cream of the league.

Shock surprise: The anemic record of Cleveland vs the other elite teams. Are they really elite? Only Boston, LA, Orlando and (right on the cusp) San Antonio manage to beat other 'playoff' teams more than half the time.

Code:
Team WinsLosseWin % Win500Lose500vs 500 WinElitLoseElitvs Elite %
Los Angeles Lakers 52 13 0.800 25 10 0.714 13 5 0.722
Orlando Magic 48 17 0.738 19 13 0.594 8 4 0.667
Boston Celtics 50 16 0.758 23 11 0.676 8 7 0.533
San Antonio Spurs 44 21 0.677 18 17 0.514 7 7 0.500
Cleveland Cavaliers 52 13 0.800 21 10 0.677 6 7 0.462
Portland Trail Blazers 41 24 0.631 16 19 0.457 9 11 0.450
Houston Rockets 43 25 0.632 19 15 0.559 7 9 0.438
Detroit Pistons 33 31 0.516 10 21 0.323 7 11 0.389
Miami Heat 36 29 0.554 13 20 0.394 7 11 0.389
Atlanta Hawks 38 28 0.576 13 17 0.433 6 10 0.375
Chicago Bulls 30 37 0.448 13 25 0.342 7 12 0.368
Denver Nuggets 42 25 0.627 19 19 0.500 8 14 0.364
Utah Jazz 41 25 0.621 17 16 0.515 5 9 0.357
New Orleans Hornets 41 24 0.631 18 16 0.529 5 11 0.313
New York Knickerbockers 28 37 0.431 11 24 0.314 5 11 0.313
Indiana Pacers 28 39 0.418 13 25 0.342 6 14 0.300
Dallas Mavericks 40 26 0.606 15 18 0.455 5 14 0.263
Milwaukee Bucks 30 38 0.441 9 25 0.265 4 12 0.250
Charlotte Bobcats 28 38 0.424 10 27 0.270 4 13 0.235
Phoenix Suns 35 31 0.530 12 23 0.343 4 16 0.200
Memphis Grizzlies 16 48 0.250 6 29 0.171 4 18 0.182
Philadelphia 76ers 32 31 0.508 8 21 0.276 3 14 0.176
Washington Wizards 15 51 0.227 4 33 0.108 3 14 0.176
Golden State Warriors 23 42 0.354 9 26 0.257 3 16 0.158
Toronto Raptors 23 43 0.348 7 33 0.175 3 17 0.150
Los Angeles Clippers 15 50 0.231 6 33 0.154 2 16 0.111
Sacramento Kings 14 51 0.215 4 34 0.105 2 18 0.100
New Jersey Nets 28 37 0.431 11 23 0.324 1 14 0.067
Oklahoma City Thunder 18 48 0.273 5 33 0.132 1 17 0.056
Minnesota Timberwolves 20 46 0.303 5 32 0.135 1 19 0.050


What do people think? Is this a valid way to look at teams which have title aspirations?
Back to top
View user's profile Send private message
Mountain



Joined: 13 Mar 2007
Posts: 1492


PostPosted: Mon Mar 16, 2009 2:59 am Post subject: Re: Records vs the top teams Reply with quote
magicmerl wrote:


What do people think? Is this a valid way to look at teams which have title aspirations?


Yes. What's better than performance against playoff or elite? Your groupings are different than Sagarin's at USAToday but I give a lot of weight to what these type numbers show.

Now SOS or SRS can vary so the rankings aren't to be taken as too precise but you can see top (Lakers , Magic) and then the next two tiers are pretty large. How far down do you give a meaningful chance to making the conference finals? I'd go pretty deep. To winning the title? I'd go really shallow.
Back to top
View user's profile Send private message
back2newbelf



Joined: 21 Jun 2005
Posts: 94


PostPosted: Mon Mar 16, 2009 7:04 am Post subject: Reply with quote
Why don't you try and see if that would have been a better indicator for playoff success in the past (i'd say no).
Also, teams might look quite different depending on where you make the cut for "elite". The Lakers' games against "elite" were quite a bit easier the way you did this
Back to top
View user's profile Send private message
bastillon



Joined: 04 Nov 2008
Posts: 34


PostPosted: Mon Mar 16, 2009 9:34 am Post subject: Reply with quote
Quote:
The second trio of columns are those same three columns vs teams with a .500 or better record. These teams are ones expecting to advance in the playoffs.

The third trio are teams record vs the teams who plan on advancing in the playoffs. There are only 4 'Elite' teams who can win consistently against the cream of the league.


Are these connected ? And what do you actually mean by "expecting to advance in the playoffs" ? over 0.500 record vs teams with record over .500 vs the entire league ?

Like... lets say Heat are .500+, but they are almost .400 vs teams like themselves(over .500), so they qualify to expected to make the Playoffs, but not to advance after they do make ?
Back to top
View user's profile Send private message Send e-mail
Mountain



Joined: 13 Mar 2007
Posts: 1492


PostPosted: Mon Mar 16, 2009 12:55 pm Post subject: Reply with quote
Look at Sagarin's accounting of performance against top 10 last season and it "predicts" every playoff series all the way to the champ except one. In close calls could go either way. Spurs beat Hornets. Lots of methods might have done close to or as well last season. This season will be tougher.
Back to top
View user's profile Send private message
magicmerl



Joined: 30 Dec 2007
Posts: 27


PostPosted: Mon Mar 16, 2009 1:45 pm Post subject: Reply with quote
bastillon wrote:
Quote:
The second trio of columns are those same three columns vs teams with a .500 or better record. These teams are ones expecting to advance in the playoffs.

The third trio are teams record vs the teams who plan on advancing in the playoffs. There are only 4 'Elite' teams who can win consistently against the cream of the league.


Are these connected ? And what do you actually mean by "expecting to advance in the playoffs" ? over 0.500 record vs teams with record over .500 vs the entire league ?

Like... lets say Heat are .500+, but they are almost .400 vs teams like themselves(over .500), so they qualify to expected to make the Playoffs, but not to advance after they do make ?

Yes to all of your questions. Sorry for using different wording. Essentially, the second and third trios are team records vs teams with a winning record from the previous trio.
Back to top
View user's profile Send private message
bastillon



Joined: 04 Nov 2008
Posts: 34


PostPosted: Mon Mar 16, 2009 7:05 pm Post subject: Reply with quote
I wonder if points/efficiency differential wouldn't be a better indicator. For example C's lost some really close games(Ray Allen TO out of bounds vs Spurs, Fisher's fouling Ray last possession vs LA) and their differential would be much higher I think.
Back to top
View user's profile Send private message Send e-mail
bastillon



Joined: 04 Nov 2008
Posts: 34


PostPosted: Mon Mar 16, 2009 9:40 pm Post subject: Reply with quote
This actually can be a very good way to evaluate players' performances. Some guys step up in those kind of situations, some don't, while actually it's their real value. I think these 3 groups would show some very interesting stuff. It can be sort of a revolution in analysis...
Back to top
View user's profile Send private message Send e-mail
tarheeljks



Joined: 24 May 2008
Posts: 2


PostPosted: Fri Mar 20, 2009 3:20 am Post subject: Reply with quote
bastillon wrote:
I wonder if points/efficiency differential wouldn't be a better indicator. For example C's lost some really close games(Ray Allen TO out of bounds vs Spurs, Fisher's fouling Ray last possession vs LA) and their differential would be much higher I think.


it should be. it's the same thing as using point/eff differential instead of win/loss for the general standings, but instead you are taking a subset applying it on a subset. it would be as if you just pretended that the nba was a league w/4 or 5 teams comprised of the list of "contenders."

i had actually been wondering about this fairly recently b/c watching cleveland play other contenders they don't seem to be quite on their level. (it was nothing more than eye test but they seemed to regress more when playing another contender than other contenders.) i had been looking at their w/l profiles on bref b/c i am too lazy to do the differentials, but i agree that they are the way to go
Back to top
View user's profile Send private message AIM Address
Italian Stallion



Joined: 04 Mar 2009
Posts: 3


PostPosted: Sat Mar 21, 2009 2:40 pm Post subject: Reply with quote
I think this a very useful way to look at teams because superior consistency of play can produce extra wins without making the team a favorite to win it all.

If a second tier playoff team almost always plays its best it will rarely lose to any of the mediocre or bad teams. That will enable the team to accumulate a very good record. It might even produce a better record than a superior team that has more off nights and that occasionally loses to a really bad team. However, head to head, the better team is likely to win as long as it is reasonably consistent. So record against other playoff teams could easily expose something like that.


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 2:29 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
dogra



Joined: 25 Feb 2005
Posts: 5
Location: Brooklyn, NY

PostPosted: Fri Feb 25, 2005 5:11 pm Post subject: +/- Statistics: Golden Egg or Wooden Nickel? Reply with quote
Hi everyone. I am new to this forum. I used to read Dean's articles many years back when I was working for an Internet company and looking for interesting things to make my day go faster.

My interest in statistics is strong, but my math background is very limited. I am joining this forum to learn more.

I have a question. Lately, a lot of basketball fans I know seem to be in love with +/- statistics for a player. They claim that this tells us how valuable a player REALLY IS to their team. Because it shows how many points the team scores, and gives up, when said player is on the floor.

Am I missing something, or is this wildly simplistic? It would seem to be clear that there are so many other factors and complex relationships at play.

What don't I understand?

Last edited by dogra on Fri Feb 25, 2005 5:21 pm; edited 1 time in total
Back to top
View user's profile Send private message Visit poster's website
radio



Joined: 17 Feb 2005
Posts: 12
Location: New York

PostPosted: Fri Feb 25, 2005 5:18 pm Post subject: Reply with quote
I'm not sure you're missing anything. I like +/- to help figure out optimal starting 5s for teams with set rosters (and even there, they're fraught with selection errors and small sample sizes), but I don't know that many people that think the numbers are applicable across teams.
_________________
Ankur Desai
Amateur Hoops Junkie
Back to top
View user's profile Send private message
dogra



Joined: 25 Feb 2005
Posts: 5
Location: Brooklyn, NY

PostPosted: Fri Feb 25, 2005 5:26 pm Post subject: Reply with quote
A friend of mine was talking up Yao Ming's negative plus/minus rating. And I thought there could be an enormous number of reasons for that -- given the situation/current roster in Houston. And only one of those reasons was that Yao Ming, alone, was hurting the team.

Then I thought of the complexity of parsing out these different variables, and my head started spinning.

At which point, I began to think that these +/- ratings don't mean too much. That they're just an eyeball stat.

Maybe I'm overstating it, though.
Back to top
View user's profile Send private message Visit poster's website
Golabki



Joined: 25 Feb 2005
Posts: 5
Location: Boston

PostPosted: Fri Feb 25, 2005 5:38 pm Post subject: problems with +/- Reply with quote
One problem with +/- is the fact that bench players tend to play with other bench players and starters tend to play with other starters. You can see this by glancing threw a few game flows (see "linkage" thread). To my knowledge there has been no effort to deal with this problem (although I could be wrong as I am new here). Without an adjustment +/- seems seriously flawed.

A second issue is that these numbers may have too much noise in them to be useful. Looking at raw point differential is probably is too blunt a tool to rate players in any meaningful way (this is me being totally subjective and very possibly wrong). I'll bet someone has done a study of this, but I don't know where.
_________________
Rock over London, Rock on Chicago.

Last edited by Golabki on Fri Feb 25, 2005 5:41 pm; edited 1 time in total
Back to top
View user's profile Send private message
Mike G



Joined: 14 Jan 2005
Posts: 85
Location: Delphi, Indiana, USA

PostPosted: Fri Feb 25, 2005 5:40 pm Post subject: Reply with quote
I too was disappointed in the seeming randomness of +/-. The biggest single factor seems to be Who plays in your place when you're out. If you back up Nowitzki, you better be damn good.

Forum member Dan R is the one person I'm aware of who has mathematically sorted thru all the combinations. It's called "regression", I think. There's fewer obvious departures from common sense; and when you do find one, he blames it on "noise".

Confused

Carry on.
Back to top
View user's profile Send private message Send e-mail
dogra



Joined: 25 Feb 2005
Posts: 5
Location: Brooklyn, NY

PostPosted: Fri Feb 25, 2005 5:44 pm Post subject: Reply with quote
If Dan R. could post or point to something on this subject, I'd be greatful.

What do you guys mean when you use the term "noise" in this [stats] context?
Back to top
View user's profile Send private message Visit poster's website
dogra



Joined: 25 Feb 2005
Posts: 5
Location: Brooklyn, NY

PostPosted: Fri Feb 25, 2005 5:49 pm Post subject: Reply with quote
I also noted this quote from an article link posted by the Admin --

Quote:
In Seattle and around the NBA, statistics haven't made the same kind of impact on the sidelines as they have in front offices. That's to be expected. Coaching ultimately comes down to feel, and, unlike general managers, coaches need only evaluate players they watch on a daily basis, not hundreds throughout the NBA and countless more amateur prospects.

Still, there are two primary ways in which the Sonics use statistics: Evaluating their own players and lineup combinations with plus-minus statistics, and scouting upcoming opponents by looking at their statistics.

[emphasis mine]

I would hope these are adjusted plus/minus stats.

Otherwise I'm confused again.
Back to top
View user's profile Send private message Visit poster's website
HoopStudies



Joined: 30 Dec 2004
Posts: 80
Location: Bay Area, California

PostPosted: Fri Feb 25, 2005 6:18 pm Post subject: Reply with quote
Mike G wrote:
I too was disappointed in the seeming randomness of +/-. The biggest single factor seems to be Who plays in your place when you're out. If you back up Nowitzki, you better be damn good.

Forum member Dan R is the one person I'm aware of who has mathematically sorted thru all the combinations. It's called "regression", I think. There's fewer obvious departures from common sense; and when you do find one, he blames it on "noise".

Confused

Carry on.


Dan's base column is here:

http://www.uncg.edu/bae/people/rosenbaum/NBA/stats1.htm

I explain the concept in Basketball on Paper, too.

What it does is separate out the fact that starters tend to play against better guys, so if you're a starter but get moved to the bench, you could benefit in raw +/- by playing against weaker guys and by replacing a generally worse guy on your own team. Dan's method accounts for all of this.

But yes, there is noise. Noise means that Dan comes out with estimates of how far off he could be from "truth" (I hate using the word). So his estimate of one guy's value may be +6 points with an error of 8. That would mean that the guy could be adding 0 points or adding a lot more than +6. You see larger errors for players who don't play much or who play only with a small set of players.

Noise happens. People aren't consistent. Nor will numbers representing them be consistent. No one should make a decision based on a single number.

Raw +/- numbers are interesting, but again just one number. On their own, they're just an indicator. As with almost everything we do here, the story is more complex than one number can represent. I think we do a good job here of balancing the story by looking at several different numbers.
_________________
Dean Oliver
Consultant to the Seattle Supersonics
Author, Basketball on Paper
http://www.basketballonpaper.com
Back to top
View user's profile Send private message Visit poster's website
HoopStudies



Joined: 30 Dec 2004
Posts: 80
Location: Bay Area, California

PostPosted: Fri Feb 25, 2005 6:27 pm Post subject: Reply with quote
dogra wrote:
I also noted this quote from an article link posted by the Admin --

Quote:
Still, there are two primary ways in which the Sonics use statistics: Evaluating their own players and lineup combinations with plus-minus statistics, and scouting upcoming opponents by looking at their statistics.

[emphasis mine]

I would hope these are adjusted plus/minus stats.

Otherwise I'm confused again.


Not adjusted, just in context. Seattle uses +/- numbers as indicators, as part of the full story. As Nate said, he didn't get a big bonus for having the highest +/- when he was a player. It's just an indicator. Decisions (such as playing time) are much more complex than looking at the #, as he says in the article.

Do you hire people based solely on IQ? Or just on SAT/GRE score? No. Real life, which to the surprise of some people includes basketball, means decisions based on a broad variety of indicators. In Seattle, we actually choose to use those indicators, whereas the NBA as a whole has historically not done much with them (beyond ppg, rpg, apg). It's changing, though. We here in this discussion group are making a difference.
_________________
Dean Oliver
Consultant to the Seattle Supersonics
Author, Basketball on Paper
http://www.basketballonpaper.com
Back to top
View user's profile Send private message Visit poster's website
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 115
Location: Greensboro, North Carolina

PostPosted: Sat Feb 26, 2005 12:29 am Post subject: Reply with quote
The big advantage of adjusted plus/minus ratings are that they are the closest we can come to an "unbiased" measure of a player's effectiveness. By unbiasedness, I mean that if we could observe a player in an infinite number of games matched up with and against lots of combinations of players, adjusted plus/minus ratings would be a near perfect measure of a player's effectiveness.

Regular plus/minus statistics and any rating based upon traditional statistics is not unbiased. Even if we observed an infinite number of games these ratings would still be a good deal off from measuring a player's effectiveness.

So that means adjusted plus/minus ratings are the best rating system, right? Well, no, not necessarily. Plus/minus ratings, both adjusted and unadjusted, are very "noisy." What I mean by "noisy" is that if we measure a player's adjusted plus/minus ratings over two different 20 game stretches, there is a very good chance that the ratings will differ a lot - even if the player has not really gotten a lot better or worse. Ratings based upon traditional statistics would vary a lot less over these two 20 game stretches.

So adjusted plus/minus ratings are almost unbiased but have a high variance.

Ratings based upon traditional statistics (PER, TENDEX, nba.com efficiency) are biased but have a relatively small variance.

Regular plus/minus ratings are biased and have a high variance.

One way to measure the variance of a rating is to present its standard error. Suppose a player has a rating of 5.0 and a standard error of 3.5. For this player, the rating of 5.0 says that if we replaced 40 minutes of play by an average NBA player with the 5.0 player, the team would improve its performance by 5.0 points.

But standard errors are useful because about 95% of the time a "confidence interval" equal to 5.0 +/- 2*3.5 = (-2.0, 12.0) should include the "true" effectiveness of the player. In other words, we are 95% confident that this player is between 2.0 points per 40 minutes less and 12.0 points per 40 minutes more effective than an average NBA player.

So you see the standard error is pretty important. Suppose the standard error was 10.0, which is sometimes is for players that don't play much. In that case the 95% confidence interval would be (-15.0, 25.0), which is so huge that it includes pretty much any value that we might have thought was reasonable. What it tells us that we cannot really use these data to say anything useful about the effectiveness of this player.

Winston and Sagarin made a big deal about Mitchell Butler last year with Wizards, claiming he was the best player on the Wizards. But his 95% confidence interval was pretty large and basically said his effectiveness was somewhere between that of Kevin Garnett and that of a replacement player. In my opinion, such an estimate is pretty noisy and certainly not worthy of saying much about.

It takes work, but it is possible to compute standard errors for pretty much any rating. The standard errors for plus/minus ratings, both regular and adjusted, are quite large. Those for ratings based upon traditional statisitics tend to be much, much smaller. It generally takes about 2 to 4 times as many games for plus/minus-based ratings to have the same standard error as traditional statistics-based ratings.

Thus, in small samples the additional "noise" of adjusted plus/minus ratings may make them less useful than traditional-statistic-based ratings. It is the old bias versus variance tradeoff that is a recurring theme in statistics.

For this reason I have often argued that one needs to use a couple seasons to get reasonable results from plus/minus ratings. Also, I have tended to focus a lot of my time trying to see how traditional statistics translate into adjusted plus/minus statistics. I can then use these estimated relationships to devise a traditional statistic-based rating that is closer to being unbiased than other traditional statistic-based ratings. This hybrid ratings, IMO, combines the best of both worlds.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
gabefarkas



Joined: 31 Dec 2004
Posts: 93
Location: NYC

PostPosted: Sun Feb 27, 2005 12:16 am Post subject: Reply with quote
what about Roland Ratings, comparing +/- on-court to that when the player is off-court? i'm definitely a big fan of those.
_________________
If the statistical revolution won't be televised, I need to call my cable provider, pronto!
Back to top
View user's profile Send private message AIM Address
Golabki



Joined: 25 Feb 2005
Posts: 5
Location: Boston

PostPosted: Sun Feb 27, 2005 1:20 pm Post subject: Reply with quote
Dan R.

It's clear that relating traditional stats to adj. +/- is a fairly obvious next step. When you say, "I have tended to focus a lot of my time" what do you mean exactly? That would be interesting to see.
_________________
Rock over London, Rock on Chicago.
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 115
Location: Greensboro, North Carolina

PostPosted: Sun Feb 27, 2005 5:15 pm Post subject: Reply with quote
I would include Roland Ratings in the unadjusted plus/minus category - even though it does make a small adjustment from raw plus/minus ratings.

So far, I have talked about my results relating traditional statistics and adjusted plus/minus ratings in a variety of posts here and elsewhere. However, the most significant collection of comments on this are in my original piece on adjusted plus/minus ratings.

http://www.uncg.edu/bae/people/rosenbau ... inval2.htm


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 2:43 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
Ed Küpfer



Joined: 30 Dec 2004
Posts: 130
Location: Toronto

PostPosted: Fri Feb 04, 2005 8:52 pm Post subject: Playoff Experience Reply with quote
I posted this in the Raptors newsgroup where the question of the value ofplayoff experience came up. It's not the most rigourous study I've ever done, and among the sloppiest writing ever penned, but I thought it still might hold some interest.

I looked at every playoff series since 1974. I regressed the outcomes against the following variables using a logit model: Home court advantaged team Pythagorean record, away team Pythagorean record, series length (i.e. best-of-x), home team previous playoff experience in minutes, away team previous playoff experience. I also used weighted playoff experience instead of the last two, with minutes played in that playoff year as the coefficient.

Home team playoff experience turned out to be statistically significant at 5%, but away team experience was not significant at all. Very strange. [I've now given this quite a bit of thought, and I still have no answer.] Regression results below for posterity.

So how big an effect does playoff experience have on the probability one team beating another in a playoff series? To answer this, we'll assume two evenly matched .500 teams playing a 7-game series. If they both have 0 games playoff experience, the team with home court advantage (call them HOME) has a 61% chance of taking the series. Assume the AWAY team has 20 player-games playoff experience, which is the median number of games among AWAY teams (player-games defined here as 36 minutes played).

Code:
HOME experience Probability
in player-games of HOME win

0 58%
5 60%
10 61%
15 63%
20 64%
25 66%
30 67%
35 69%
40 70%


So every 10 extra player games of playoff experience -- 7.5 full team games -- raises the probability of a team win by 3%. Seven and a half game is roughly a series and a half. That seems like a lot of playoff experience for such a little amount of gain.

[Note: I did the regression twice, once using straight playoff experience per team in minutes, and once using the average playoff experience per team weighted by each player's minutes. The results were very similar, so I've only shown straight playoff experience below.]

Code:

[REGRESSION RESULTS]

Link Function: Logit

Response Information

Variable Value Count
HWin 1 313 (Event)
0 103
Total 416

Logistic Regression Table
Odds 95% CI
Predictor Coef SE Coef Z P Ratio Lower Upper
Constant 0.537 1.311 0.41 0.682
BestOf -0.0033 0.1113 -0.03 0.976 1.00 0.80 1.24
H_Pyth 8.363 1.822 4.59 0.000 4284.83 120.57 1.52E+05
A_Pyth -8.553 1.893 -4.52 0.000 0.00 0.00 0.01
H_Exp 0.0017062 0.0007922 2.15 0.031 1.00 1.00 1.00
A_Exp -0.0007275 0.0007920 -0.92 0.358 1.00 1.00 1.00

Log-Likelihood = -200.462
Test that all slopes are zero: G = 64.729, DF = 5, P-Value = 0.000

_________________
ed


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 2:43 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
d Küpfer



Joined: 30 Dec 2004
Posts: 130
Location: Toronto

PostPosted: Tue Feb 08, 2005 2:40 am Post subject: Comeback Probability Reply with quote
Here's a question: if a team is trailing in the latter part of a game, what strategy would be more likely to produce a comeback? There are lots of ways to answer that. Here's the simplest:

Assume that there's only two strategies available to the team:

* Chucking up three pointers quickly and fouling the opposing team as soon as they gain possession. Call this the 3 Point Strategy.
* Shooting two pointers quickly and fouling the opposing team as soon as they gain possession. Call this the 2 Point Strategy.

Neither of these are very realistic, but I didn't want to complicate matters. Of the two, most of you will probably guess that the comeback is most likely using the chuck-a-three. But it's fun to put some numbers to these types of questions. To that end, I created a spreadsheet which calculates the comeback probability, based on a few variables:

1. The number of points the team is trailing by.
2. The team's 3 point shooting accuracy.
3. The team's 2 point shooting accuracy.
4. The opposing team's free throw shooting accuracy.
5. Possessions remaining in the game. This variable depends in turn on three other variables:
1. The time remaining in the game.
2. The number of seconds into the shot clock at which point the team shoots.
3. The number of seconds into the shot clock at which point the team fouls.

The spreadsheet, which I'm calling the Comeback Probability Toy, assumes that both of the strategies require the team to shoot the ball at a given point in the shot clock, the same on each possession; and for the team to foul at a certain point, the same in each possession. The Toy allows you to change these values to whatever you want.

Example scenario: a team down 5 points with 30 seconds remaining. The team shoots 50% on 2-pointers, 33% on 3-pointers. Their opponents shoot 70% on free throws. Team will chuck a shot 5 seconds into the shot clock, and will foul one second after opponents gain possession. What is the probability of a comeback? The Toy says the 2 Point Strategy results in a comeback less than 1% of the time, while the 3 Point Strategy works about 4% of the time.

How about: down 2, 30 seconds left, same shooting percentages, same shooting and fouling times? The Toy says the 2 Point Strategy results in a comeback 10% of the time, while the 3 Point Strategy works 16% of the time.

Of course, reality is more complex, and nobody would ever opt to use an all-two- or all-three-point strategy. I have to make it clear that the probabilities produced by the Toy rely heavily on the assumptions built into it, and these assumptions bear little resemblance to what happens in an actual game. But I am not trying to model reality here. I am trying to coontrast two reductio ad absurdum strategies in order to have a baseline for comparison.

You can download the Comeback Probability Toy at my Yahoo page, in the files section.
_________________
ed


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 2:55 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
antcole



Joined: 12 Dec 2005
Posts: 74


PostPosted: Mon Jun 05, 2006 7:02 am Post subject: The equation for True Shooting Percentage is wrong. Reply with quote
I have an important announcement to make and it pertains to one of our favorite tools in the basketball analysis community. I'm talking about True Shooting percentage. I've just realized something about it: the formula is incorrect.

I discovered this when I looked at the stat and I realized that it wasn't adjusted for three pointers. As an example: Player A takes one three point attempt (his only shot attempt) and makes it. We use the TS% formula (Points/(2*(FGA+(FTA*.44)))). The equation will produce a result of 1.5 which is absurd. If a player has only attempted one field goal attempt and he makes the shot, it should only equal 1, because 1 represents 100%. Anything representing a percentage stat shouldn't go above 100%, and if it does then it can't possibly be accurate. I figured out the real true shooting percentage equation. I tested it to make sure it is accurate. Everytime I had a perfect shooting rate and I used this equation it produced a perfet rating:

Points/((1*FTA)+(2*2point FGA)+(3*3point FGA))

As we can see, the real True Shooting Percentage is Points divided by Free Throw Attempts multiplied by 1, plus two point field goal attempts multiplied by 2 plus three point field goal attempts multiplied by 3.
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Mon Jun 05, 2006 9:05 am Post subject: thinking about True Shooting Percentage Reply with quote
The TS% formula looks correct to me. See if this helps:

One three point shot is worth 3. One two point shot is worth 2. The TS% of going 1 for 1 on a threepoint shot should/ must be 50% more than going 1 for 1 on a 2 point shot. The TS% is denominated in terms of a two point shoot so 1 for 1 on a three just looks a little weird, objectionable formatting to you, but it is stilly convey an accurate representation of scoring rate per scoring posession used. For a large mix of shots the formula will look more normal and will accurately show the % of a 2 point basket earned per scoring attempt.

In your new shooting percentage you've changed the weighting for FTA.
If you go 1 for 1 on a three point shot and 2-2 on FTs, under the standard TS% that is 5pts/3.76. Under your formula it is 5pts/5.
If you want essentially points per scoring attempt or scoring possessions used, the .44 comes into play to adjust FTattempts properly, estimating roughly the number of scoring attempts/scoring possessions used that end up in freethrow attempts. Backing out technical fouls I believe.

If I have misstated anything, someone can correct me.

Last edited by Mark on Fri Jun 09, 2006 12:21 pm; edited 7 times in total
Back to top
View user's profile Send private message
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 978
Location: Seattle

PostPosted: Mon Jun 05, 2006 9:53 am Post subject: Re: The equation for True Shooting Percentage is wrong. Reply with quote
antcole wrote:
Anything representing a percentage stat shouldn't go above 100%, and if it does then it can't possibly be accurate.

I hope you don't ever read about slugging percentage; that's been "wrong" for several decades!
Back to top
View user's profile Send private message Send e-mail Visit poster's website
deepak



Joined: 26 Apr 2006
Posts: 664


PostPosted: Mon Jun 05, 2006 10:43 am Post subject: Reply with quote
TS% just takes two points as a baseline number for points that should be produced per scoring attempt. It makes some sense since most successful scoring attempts result in two points. But it shouldn't matter what you use as that baseline figure. I think you could use 3 or 4 if you're worried about TS% being higher than 100% in a give game. Everyone's TS% will just be scaled down.
Back to top
View user's profile Send private message
hpanic7342



Joined: 16 Feb 2005
Posts: 201


PostPosted: Mon Jun 05, 2006 10:47 am Post subject: Re: The equation for True Shooting Percentage is wrong. Reply with quote
admin wrote:
antcole wrote:
Anything representing a percentage stat shouldn't go above 100%, and if it does then it can't possibly be accurate.

I hope you don't ever read about slugging percentage; that's been "wrong" for several decades!


Word. True shooting percentage is not, and never has been, a percentage. No one on this list has ever pretended otherwise.
Back to top
View user's profile Send private message Send e-mail
Ed Küpfer



Joined: 30 Dec 2004
Posts: 785
Location: Toronto

PostPosted: Mon Jun 05, 2006 11:07 am Post subject: Re: The equation for True Shooting Percentage is wrong. Reply with quote
hpanic7342 wrote:
True shooting percentage is not, and never has been, a percentage. No one on this list has ever pretended otherwise.


We've been through this before, usually with me taking antcole's position, until I saw the light. Technically, TS% is a percentage, just a percentage not bounded at one. No big deal.

In any case, changing the formula would not be helpful. The current TS% has two desirable properties: it looks kinda like FG%, which everyone is familiar with, and it's been around forever, which means it would be a huge chore to change -- just to satisfy one definition of "percentage." I think our energies would be more productively exhausted in other pursuits.

If it were up to me, we'd go all the way back to the beginning of basketball, and start keeping track of Points Scored By Field Goal Attempts instead of FG%. This would allow for the later introduction of 3-pointers, and also provide an easy, and analogous transition to PTS/(FGA + .44*FTA), which is simply TS% * 2, but without the invitation every few years for someone (like me) to argue that the stat is not a percentage.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Mike G



Joined: 14 Jan 2005
Posts: 3565
Location: Hendersonville, NC

PostPosted: Mon Jun 05, 2006 4:50 pm Post subject: Reply with quote
Why would any use of 'percentage' need to have a ceiling of 1.000? This year, Lebron James scored 204% as many points as #2 Cav, Ilgauskas. He also scored 114% as many points as he scored last year.

There's nothing mysterious about numbers greater than 100%, any more than numbers greater than 1. They're the same exact thing.

TS% can be expressed as 'the percent of times that 2 points happened' on a scoring attempt, by a person, a team, a league, etc. On each attempt, we may have recorded 0, 1, 2, 3, or 4 points. The 'basket' is traditionally 2 points, and so that is the standard 'successful' possession, on offense.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
antcole



Joined: 12 Dec 2005
Posts: 74


PostPosted: Wed Jun 07, 2006 11:26 am Post subject: Reply with quote
Let me start off and say that I’m shocked that nobody on this thread sat down and considered what my equation was determining. If you look at it, the equation, it explicitly shows a particular player’s overall shooting accuracy. It simply gives the fractionof points scored relative to the amount of possible points, nothing more and nothing less. When I thought this up, I was keeping in mind Effective Field Goal Percentage. I adjusted for the values of a particular stat and I conducted the experiments to make sure that it is true.
Now let me look at the points I made:
-True Shooting Percentage is wrong because its final value can exceeds 1
-The formula I produced is the true stat we should use if we are trying to find a player’s actual equivalent shooting percentage.

Are my claims correct? Sure they are!

I didn’t say anything about Points per Shot attempt being incorrect or we should stop using it to gauge a player’s shooting effectiveness. I said that True Shooting Percentage as a percentage is wrong. This problem that I have with True Shooting Percentage all relates back to two critical subjects: Language and context.

I’m not being picky by speaking on this subject; I’m just stating the importance of both language and context within our work. As we all know as basketball statistical analysis we must work in context because stats that are not put into proper context are worthless. However, as basketball writers we also must exercise proper use of language. It is crucial for us to use language in its proper context in order for us to establish clarity between us and the reader. If we don’t put language in its correct context we run the risk of undermining our research and losing the trust of the readers.

Which brings me back to this thing we call “True Shooting Percentage”. I must ask everybody on this board a question. What does the words “True Shooting Percentage” stand for? Or take it one step further: What does True Shooting percentage suggest to a basketball fan that is just getting into the APBR movement? I’ll tell you what it means to me: the words “True Shooting Percentage” tells me that this is a stat that gauges a player’s overall shooting accuracy by relating his shots made to the total shots this particular player attempted. But does “True Shooting Percentage” tells this to the reader? No, it does not because TRUE SHOOTING PERCENTAGE is not a percentage stat.

A percentage equation is any formula that measures a section or part relative to the whole object. That means the maximum value should be 1 because that represents 100%- total value, meaning if the final value exceeds 1 then it is obviously not a percentage stat. As I’ve already stated and somebody else on this board agreed, “True Shooting Percentage” is not a percentage, thus I don’t think we should label it such.

I know people are saying well “Anthony it’s not that big of a deal. We shouldn’t pay attention to the name that is given to it, because it is insignificant. What truly manners is the information that the stat gives to us.” I personally don’t agree with this logic because it goes back to my earlier point about language and context. Percentages play a very important role in the minds of basketball fan and they think of it in the context of its original definition. When a fan reads or hears that Shaq went 13 for 20 from the field they know that means he made 65% of his shots and had a great night shooting from the field. Meanwhile if Steve Nash went say 6 for 7 from the free throw line fans will automatically know that he made 85.7 percent from the line, and he’s an assassin when he’s allowed to shoot from the line. In contrast, when the average fan hears that Ben Wallace averaged something like 25 percent from the free throw line in the playoffs they know he’s a horrendous free throw shooter.
Percentages go a long way in telling the fan what kind of caliber a shooter truly is because it is a measure of accuracy, error, and precision. But more importantly, shooting percentages is a part of basketball fan’s collective unconscious and the word produces an indelible image in the fan’s mind. This is why I’m no longer a fan of labeling true shooting percentage as a percentage. The label is not only a misnomer, but the language is also misleading and confusing to the fans.

Also I must ask everybody on this board another question: did anybody sit down and look at “True Shooting Percentage” in its current numerical form and ask yourselves what does this value truly mean? Seriously, for a career, Charles Barkley’s career TS% is 61.2%. So what does that truly mean? Percentage of what? Does it mean Barkley made 61.2% of his shot attempts? What is this telling us and more importantly the readers we are trying to educate? We all know that the stat truly means that Barkley averaged 1.224 points per shot attempt. OK, then why don’t we just use the point per shot attempt statistic instead of True Shooting percentage? We all know why.

We’re really just taking the PSA stat and dividing it by two (or multiplying by 50) in order for us to get a bastardized number that looks like a percentage. We use this method as a means for PSA to compete with Field Goal Percentage. I understand and respect the logic of the people who desire to use the True Shooting Percentage measure as a means for people to grasp the concept of a player’s shooting value. However, I think we should also consider the fact that while Points per Shot Attempt is related to Field Goal percentage we also should realize that these two stats serve two completely different functions. Points Per Shot Attempt is an average, a rate, and a measure of a player’s overall shooting effectiveness. However, as I’ve already mentioned Field Goal percentage is a measure of a player’s shooting accuracy, error, and precision. That is why I think it is a folly to try to express this PSA as a percentage in order to make it look superior to Field Goal Percentage because they aren’t measuring the same thing. They are both great stats in their own right but we shouldn’t, expect one stat to do the other’s job.

Also as I suggested earlier, by trying to convert PSA into a “pseudo”-percentage stat it loses much of its original meaning. If a player has a true shooting percentage of 53 percent, it isn’t telling you anything. We have to go out of our way to see what the stat truly means, that this particular player averaged 1.06 points per shot attempt. I personally think that we would help our community if we use PSA instead of TS%. The language produces absolute clarity.

I say all of this not to sound like a trouble maker. I say this because it strikes at the biggest problems that is hurting our community: Accessibility. It is no secret that a good percentage of the basketball community still looks down at our work. Yes we like to claim that it is all due to the fact that they are traditionalists and they believe in all those weak-ass clichés and conventional wisdom. But we’d be foolish to believe that traditionalism is the only problem. Part of the problem is with the writers themselves.

Let’s be honest, some of us have done a poor job communicating with the readers. Most basketball fans (and yes I do mean most basketball fans) think that we are just geeks with calculators who are using various meaningless formulas to prove something about basketball. At first when I heard these remarks I would roll my eyes, but now I see their point. The problem truly isn’t with these formulas that we use, but with the writers that use them. Some of our writers on occasions haven’t done a good job of showing how these stats relate back to the outcome of a game of basketball. This problem is always a result back to a misuse of language, context or in some cases pure laziness. That is why I think we should really consider the power of language in our articles. To pretend that the words that we use do not have a strong effect on the imaginations of our readers is being reckless.

Oh yeah and by the way Kevin, I do know the history of Slugging Average. The creators of Slugging average, Branch Rickey and Alex Roth, never called it “slugging percentage”. Rickey himself called the stat “extra base power” and he knew that the stat was an average. If you ever read Rickey’s “Life Magazine” article, he never once called “slugging average” a percentage. They were well aware that Slugging wasn’t a percentage.
The only reason why people began labeling slugging average as slugging percentage is because of a mistake by people reading the numbers. People looking at the numbers noticed that the numbers for every major league player were decimals and assumed that it was a percentage. It was all a mistake of accounting and personally I always call the stat “slugging average” because its really only telling you how many bases a hitter produces per official at bat and its not giving me a percentage of anything. I swear when I read somebody say “slugging percentage” my mind automatically thinks “percentage of what”? A real slugging percentage would have to calculate total bases divided by the maximum number of bases a hitter can produce per at bat. And in case you’re interested Kevin, here are the all-time leaders in Slugging Percentage:
Babe Ruth: .1725
Ted Williams: .1584
Lou Gehrig: .1581
Albert Pujols: .1552
Steroid Monster: .1527
Back to top
View user's profile Send private message
deepak



Joined: 26 Apr 2006
Posts: 664


PostPosted: Wed Jun 07, 2006 12:37 pm Post subject: Reply with quote
antcole wrote:
What does True Shooting percentage suggest to a basketball fan that is just getting into the APBR movement? I’ll tell you what it means to me: the words “True Shooting Percentage” tells me that this is a stat that gauges a player’s overall shooting accuracy by relating his shots made to the total shots this particular player attempted. But does “True Shooting Percentage” tells this to the reader? No, it does not because TRUE SHOOTING PERCENTAGE is not a percentage stat.


If anything, perhaps the name is a misnomer. But the formula itself, pts/(2*(fga+.44*fta)), is a valid and useful representation of scoring efficiency for a player. I think it's less useful for single games (since the .44*fta is an approximation), but over the course of several games it works well.

Suppose in one game, Billups takes 10 shots: 5 threes and 5 two pointers. Let's say he sinks all the threes, and misses all the two pointers. By the original formula, his TS% = 75%. By your formula it's 50%.

Now, suppose in the same game Rip takes 10 shots: 5 threes and 5 two pointers. Let's say he sinks all the twos, and misses all the threes. By the original formula, his TS% = 50%. By your formula, it's still 50%.

Your formula says both players were equally efficient. The original formula says Billups was more efficient. As I see it, the latter is obviously true. Using the same number of shot attempts in the Piston offense, Billups got more points for his team.

By the original formula, Billups had a TS% = 60.2% in 2006. Ben Wallace had a TS% = 50.1%.

By your formula, for the season Billups had a TS% = 39.3%. Ben Wallace had a TS% = 48.3%.

What's a more accurate statement: Billups is a more efficient scorer than Ben Wallace, or Ben Wallace is a more efficient scorer than Billups?

Perhaps True Shooting% doesn't quite capture the concept of efficiency which the original formula is trying to get at. Maybe a better name would be "Scoring Efficiency".

antcole wrote:
Seriously, for a career, Charles Barkley’s career TS% is 61.2%. So what does that truly mean? Percentage of what? Does it mean Barkley made 61.2% of his shot attempts? What is this telling us and more importantly the readers we are trying to educate? We all know that the stat truly means that Barkley averaged 1.224 points per shot attempt. OK, then why don’t we just use the point per shot attempt statistic instead of True Shooting percentage? We all know why.


The original formula is no less a percentage than yours. Both are measuring points/[expected points]. The difference is that the original formula assumes [expected points] is proportional to total scoring attempts -- and defines 2 points expected per scoring attempt. Your formula defines [expected points] as total points that would be scored if all shots went in.

They aren't really that different, conceptually. But I think it's important that the denominator is proportional to overall scoring attempts (approximately equal to fga + .44*fta), versus applying different weights for different shot types.
Back to top
View user's profile Send private message
antcole



Joined: 12 Dec 2005
Posts: 74


PostPosted: Wed Jun 07, 2006 1:45 pm Post subject: Reply with quote
deepak_e wrote:

By your formula, for the season Billups had a TS% = 39.3%. Ben Wallace had a TS% = 48.3%.

What's a more accurate statement: Billups is a more efficient scorer than Ben Wallace, or Ben Wallace is a more efficient scorer than Billups?

Perhaps True Shooting% doesn't quite capture the concept of efficiency which the original formula is trying to get at. Maybe a better name would be "Scoring Efficiency".



Actually Billups's shooting percentage by the formula I gave you is 50.3%, while Ben Wallace's percentage is 48.5%. Billups scored 1495 points in 05-06 season and he attempted 587 2 pointers, 425 3 pointers and 520 free throws. The total points that Chauncey could have scored if he made all of his shots were 2969. I divided his points total; 1495 by the points possible 2969 and I got 50.3%.

I also I totally agree with the idea for the name change to "scoring efficiency" or something along those lines. Kind of like PER or the NFL's QuarterBack rating or Baseball Prospectus' Equivalent Average. I've always loved the concept of PSA because it distinguishes between the effective shooters with the ineffective shooters. The stat has a real importance in our field of study.
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Wed Jun 07, 2006 2:13 pm Post subject: Reply with quote
I listened. I calculated examples with both equations. I tried to give it a fair hearing and some feedback based on how I heard it and what it means to me. Now based on your followup, I better understand what you presented and why. And will respond, to hopefully better explain my position.

If you had started the thread by saying "the label is incorrect" and "I have a different formula that does what I think the label true shooting percentage suggests" I would have phrased the beginning of my original reply differently (since revised) but essentially my main reponse would include the same. The TS% equation does what I want it to give me. Scoring efficency is indeed a better label and I have no opposition to that. It was however correctly weighted for three points as you incorrectly charged and as shown later I believe incorrectly fix with your equation and free throws and your equation is not in the context of a shooting percentage that gives scoring efficiency which is what I want.

The freethrows are adjusted in traditional TS% by dividing by .44 because each FTA represents about half a scoring possesion used and is a necessary step to maintain equal treatment of possessions, which is a fundamental tenet here. Your equation seeks and gives equality to the three types of shots but in doing so no longer the three underlining scoring possessions used. [/u] If you are indeed arguing that a departure from an equal scoring possession used basis of analysis is justified and your equation should replace TS%, I disagree with doing that since I am seeking scoring efficiency.

"It simply gives the fraction of points scored relative to the amount of possible points, nothing more and nothing less."

It actually gives the fraction of points scored relative to the amount of possible points after the shot is released and or a foul is called, Something different than the current equation (where every scoring possession could yield up to three points, and it was denominated in terms of 2 points but scores could exceed 100% up to 150% as already discussed in much detail and as mentioned that could be changed if it really was important- which I dont think it is.)

Now under your equation, 2 point shots and nearly all trips to the line can only yield 2 points (except the rare fouled on a three attempt). But a missed threepointer, in a large pool of data, will now add zero to the top of the equation but 3 to the bottom, not two. It is now bonus and bigger penalty for a 3 pt shot miss proposition instead of bonus points vs. just record a missed field goal, equal in weight to a missed 2 point shot. Again, this deviates from all scoring posessions used being treated equal. 2 missed threes would be equal in "cost" in the new equation to 3 missed 2 point shots, but only 2 zero point possesions were caused instead of 3.

(The equation result in the case of a single missed 3 point would be correct but that doesnt validate the equation for use in all cases.)

If I want a look at shooting accuracy the big three separate rates are easy enough to look at- FG%, 3pt%, and FT% and I can look at all three simultaneously in a "shooting triple crown" comparison among great shooters.

I will still most frequently use a possesion based TS% for checks of global scoring efficiency rates. And the separate shooting rates or eFG% when I want to consider just 1 or 2 parts of shooting/scoring.

Last edited by Mark on Fri Jun 09, 2006 12:24 pm; edited 22 times in total
Back to top
View user's profile Send private message
deepak



Joined: 26 Apr 2006
Posts: 664


PostPosted: Wed Jun 07, 2006 2:25 pm Post subject: Reply with quote
antcole wrote:

Actually Billups's shooting percentage by the formula I gave you is 50.3%, while Ben Wallace's percentage is 48.5%. Billups scored 1495 points in 05-06 season and he attempted 587 2 pointers, 425 3 pointers and 520 free throws. The total points that Chauncey could have scored if he made all of his shots were 2969. I divided his points total; 1495 by the points possible 2969 and I got 50.3%.


My mistake. Forgot to subtract 3fga from fga.

Quote:
I also I totally agree with the idea for the name change to "scoring efficiency" or something along those lines. Kind of like PER or the NFL's QuarterBack rating or Baseball Prospectus' Equivalent Average. I've always loved the concept of PSA because it distinguishes between the effective shooters with the ineffective shooters. The stat has a real importance in our field of study.


Ok. What I'm not quite seeing is how useful TS%, as you want to define it, would be. Wouldn't it just make more sense to look at FT%, 3p%, and 2p% separately, rather than combining them all with your formula? What does combining them really achieve?
Back to top
View user's profile Send private message
antcole



Joined: 12 Dec 2005
Posts: 74


PostPosted: Wed Jun 07, 2006 5:10 pm Post subject: Reply with quote
I don't have alot of time to elaborate to your points, Deepak_e and Mark, but I'll leave you three points to ponder:

1) By reading my posts, did I ever suggest that I thought that my formula was better than PSA?

2) Do you think I believe that my formula is better than PSA?

3) Tupac Shakur and Biggie aren't dead. Wink
Back to top
View user's profile Send private message
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Wed Jun 07, 2006 5:39 pm Post subject: Reply with quote
"1) By reading my posts, did I ever suggest that I thought that my formula was better than PSA?
2) Do you think I believe that my formula is better than PSA? "

Since to most people TS% and PSA are different forms of the same thing
(points per scoring attempt), on this basis it seemed to me that you were doing these two things.

(It is worth noting that points per FG attempt recorded as a FGA is different than points per scoring attempt. You included free throws so I assume you dont really have a problem with defining PSA to include points generated from FTs in it. )

PSA and TS% are not exactly the same to you because of label and format- strictly speaking you objected to true shooting percentage. But you attacked both the label and the formula and the formula is functionally equivalent to the equation in PSA, just a different format for expression of the result; so by translation you appeared to have attacked both to me and can't in my view reasonably claim to accept one formula and attack / want to replace the use of the other. The equation under TS% / PSA is perfectly solid to report scoring efficiency fairly and none of your criticisms of the formula changed that.

"We have to go out of our way to see what the stat truly means" is your opinion of TS%-so use PSA; but to me TS% gives me what I want and exactly the same thing as PSA and as valid but in a form that to me is the easiest to understand. That appears to the preference of most folks since that is by far more heavily used format.

You made good points about the label- if you had only done that, fine no problem from me. I probably woudnt have entered this thread. But your initial "announcement" was "the formula is incorrect". You did not prove TS% to be "incorrect" in equation at all as I or others use it. You dont like the label and format of that equation. Point taken, but then I assumed you set off to replace the TS% equation with one that did the same thing, to "correct" the supposedly "incorrect". You did not. No matter what you call it, if you are interested in total scoring efficency,as I am, the original TS% / PSA equation is correct for finding it. Call it total scoring efficiency, is the case closed? And if you say you accept PSA, was there ever a need for a case then against the formula that it shares with TS%? Just ditch the label.

So was there really any need for a new equation? Not to find scoring efficiency. But perhaps, trying to hear your case, to get at your concept of scoring accuracy. But that would be an add-on, not a replacement to scoring efficiency. Moving from calling the TS% equation incorrect to proposing a new equation that did something different threw many of us even though you indeed said you were trying to do so to get to a "true shooting percentage" of some kind as you wished to define it.

I felt your new formula had fundamental conflicts with possession based stats regarding the handling of free throws and three point shots because I was viewing it as a scoring efficiency measure. I thought, remove those conflicts I found and you would be right back where we started so why the fuss? That was my thinking. Different from yours.

But going beyond what I previously thought and said, I'll accept for a moment that you want to create a new measure closer to the old label- true shooting percentage. Based on my previous post I'd say if you used your fomula you have a shot "scoring potential realized" percentage. Is that what "true shooting percentage" means to you? It is clearly different from scoring efficiency, I think we can now agree about that rather than disagree. It is apparently a useful method to you. Will it be useful to me or others? I guess you shared an option and we will find out, decide for ourselves. Right now I'll stick to using a scoring efficiency formula. But I can make room on the shelf for the concept of a different equation on shooting, what I see as a shot "scoring potential realized" percentage (at this point the old label seems like it might be a bad idea for a new formula). After working at it, that is the best I can do, to say what I wanted to say about the method I prefer to use and you criticized and seemed to want to replace and then get around to understanding your position on your new equation- as different in purpose and therefore to me not really competitive with the measure of scoring efficiency whether called and formatted as PSA or TS%.

Last edited by Mark on Fri Jun 09, 2006 12:33 pm; edited 1 time in total
Back to top
View user's profile Send private message
Mike G



Joined: 14 Jan 2005
Posts: 3565
Location: Hendersonville, NC

PostPosted: Fri Jun 09, 2006 11:45 am Post subject: Reply with quote
Maybe the confusion regarding 'percentage' can be resolved after all. You can increase a quantity by more than 100%, but you cannot decrease it by more than 100%.
A solution of several constituents must be <100% composed of any one constituent; but that is because 100% is the upper limit. There's no theoretical upper limit to production or productivity.
If we compare player or team productivity to league averages, roughly half of player/team rates are above 100% of the average.

So it's just a 'special case' in the use of the term 'percentage' that has 100 as an absolute limit. The rest of the time, we may say 135% or 1.351. The latter notation has more information in equal line-space.
_________________
`
36% of all statistics are wrong

Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 491
Location: Greensboro, North Carolina

PostPosted: Fri Jun 09, 2006 12:11 pm Post subject: Reply with quote
For me the reason I like true shooting percentage rather than true shooting efficiency (or whatever we call it) is because I see it as an adjustment (and potential replacement) for field goal percentage. And because it is called a "shooting percentage," it immedidately tips off the newcomer that he/she might want to think of this as a substitute for field goal percentage. So in my opinion the name of this statistic is doing precisely what we want it to do.

Newcomers have some intuition about what a good field goal percentage is, but they have no idea what a good PSA is. So by naming this true shooting percentage, we borrow on that intuition. Anthony, your suggestion would ignore that intuition. I am not sure how that helps us become more accessible to newcomers, since I have always found it easier to teach students when I try to relate it something they know. True shooting percentage does that, while points per shot attempt does not.

As to what it is a percentage of, it is the percentage of shots made by a player if we replace all of his shot attempts by two point attempts and he scores the same number of points. That may be convoluted but it does explain how it can be interpeted as a percentage. And the fact that this "percentage" can be higher than 100% certainly is within the bounds described by the Wikipedia definition of percentage.

Wikipedia defines percentage as "a way of expressing a proportion, a ratio or a fraction as a whole number, by using 100 as the denominator."

http://en.wikipedia.org/wiki/Percentage

That definition is not nearly as rigid about what a percentage is as Anthony's posts would suggest. It appears that he has a sense of what a percentage is that is narrower than this definition.

Note also that this complaint about true shooting percentage would apply to effective field goal percentage as well.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
Mike G



Joined: 14 Jan 2005
Posts: 1428
Location: Delphi, Indiana

PostPosted: Sat Jun 10, 2006 11:06 am Post subject: Reply with quote
This forum is now 30 pages deep, but back in the beginning we sort of agreed that 'true shooting percentage' was the least-objectionable term for:

TS% = Pts/(2*(FGA+.44*FTA)

At the time, I had reservations about the word 'true'. What if something even truer should come along?

I've called my stat 'effective shooting percentage', and for years it was:

Eff% = Pts/(2*(FGA+.5*FTA)

I think this is what BobC uses, but he has the term 'FG' in the name of it. I've gone back to this formula (and my own term for it) because 'effectively' a missed FT is much less likely to be offensively rebounded than a missed FG. The difference in the multiplier (.50 vs .44) seems to almost exactly account for the rebounding differential -- in 'effectiveness'.

So Bob and I arrive in the same camp thru entirely different paths. None of these is 'true', per se. The term 'effective' cannot be reserved exclusively for another stat, any more than 'efficiency' or 'rate' can. Those 2 terms can mean almost anything.
_________________
40% of all statistics are wrong.
Back to top
View user's profile Send private message Send e-mail
asimpkins



Joined: 30 Apr 2006
Posts: 146


PostPosted: Sat Jun 10, 2006 1:00 pm Post subject: Reply with quote
antcole wrote:
Also I must ask everybody on this board another question: did anybody sit down and look at “True Shooting Percentage” in its current numerical form and ask yourselves what does this value truly mean? Seriously, for a career, Charles Barkley’s career TS% is 61.2%. So what does that truly mean?


To me it means that Barkley, after adjusting for free throws and three pointers, shot as efficiently as a 61.2% FG% shooter. I think of it as an adjusted percentage. The concept always came fairly intuitively to me. (In fact, I deduced a clumsy version of the formula myself before I stumbled upon the stats community.)

Sure, PSA is much more of a pure notion, but as Dan pointed out, it's much more practical to build off the FG% thinking that all fans already have in their heads.
Back to top
View user's profile Send private message
bchaikin



Joined: 27 Jan 2005
Posts: 497
Location: cleveland, ohio

PostPosted: Sat Jun 10, 2006 2:05 pm Post subject: Reply with quote
I think this is what BobC uses, but he has the term 'FG' in the name of it.

what i call Scoring FG% or ScFG% is:

(2pters + 1.5x3pters + FTM/2)/(FGA + FTA/2)

which is the same as:

Eff% = Pts/(2*(FGA+.5*FTA)

and this formula says that this % is what the player would have shot from 2pt range had the player scored the same number of points he did in fact score taking 2pters, 3pters, and FTs - on just 2pters. it is not possessions based which is what i believe:

TS% = Pts/(2*(FGA+.44*FTA)

attempts to be. they are for all intent and purpose measuring two different things...


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 3:06 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
replayhoops



Joined: 18 Feb 2005
Posts: 10
Location: Connecticut

PostPosted: Mon Feb 21, 2005 9:57 pm Post subject: Estimating Statistics Reply with quote
I was wondering if anyone had methods to estimating certain statistics (such as turnovers and offensive rebounds) for those years before such stats were kept by the NBA?

Thanks,

Dave
Back to top
View user's profile Send private message
Mike G



Joined: 14 Jan 2005
Posts: 1492
Location: Delphi, Indiana

PostPosted: Mon Feb 21, 2005 10:48 pm Post subject: Reply with quote
Here's my estimated turnover formula:

eTO = .08*Pts + .07*Reb + .16*Ast + .05*Stl + .10*Blk - .005*Min

I don't have real theories behind the correlations; I just tinkered with the factors until listing guys' TO/eTO ratio, I seemed to have all sorts of players (PG,PF, etc) in all parts of the list.

Of course, this formula doesn't help you before 1974; in that era, you further have to estimate steals and blocks. Since these are so variable, I haven't concentrated on anything very precise. But here they are:

eStl = sqrt(Ast) - .4

eBlk = sqrt(Reb) - 1.6

Some guys will come up with negative numbers, so do whatever you think. (SQRT is Excellese for square root.)

In 1974, players were turning it over much more than today. There's a theory that a faster game pace might naturally produce more TO. But I don't know. You still have to take care of the ball for some 24 minutes. Shooting quickly is one way to avoid turnovers.
Back to top
View user's profile Send private message Send e-mail
kjb



Joined: 03 Jan 2005
Posts: 660
Location: Washington, DC

PostPosted: Tue Feb 22, 2005 9:56 am Post subject: Reply with quote
Check appendix 4 in "Basketball On Paper" for Dean Oliver's method.
Back to top
View user's profile Send private message Visit poster's website AIM Address Yahoo Messenger
Mike G



Joined: 14 Jan 2005
Posts: 1492
Location: Delphi, Indiana

PostPosted: Tue Feb 22, 2005 12:10 pm Post subject: Reply with quote
This topic reminds me I haven't checked TO/eTO lately. It turns out, of 350 players with at least 215 minutes, there are 15 guys who turn it over at least 50% more than they should (according to the aforementioned formula):

TO/e player pos tm TO
1.97 Araujo,Rafael C Tor 2.4
1.82 Telfair,Sebastia PG Por 3.1
1.76 Livingston,Shaun PG LAC 3.1
1.70 Martin,Kevin SG Sac 1.8
1.69 Norris,Moochie PG NY 2.1
1.66 Rogers,Rodney SF NO 2.7
1.63 Diaw,Boris SF Atl 2.3
1.63 Barbosa,Leandro PG Phe 2.6
1.62 Williams,Aaron PF Tor 1.6
1.59 Brown,Kwame PF Was 2.4
1.56 Gill,Eddie PG Ind 1.9
1.56 Perkins,Kendrick PF Bos 2.4
1.55 Ivey,Royal SG Atl 2.2
1.50 Murray,Ronald SG Sea 2.3
1.50 Collier,Jason C Atl 2.1


Several rookies head the list; only 2 play over 20 mpg (Livingston, 22; and Rogers, 30). The number after the team name is normalized TO/36.

Now here are the best at taking care of the ball, relative to their production:

.28 Hoiberg,Fred SG Min .5
.35 Thomas,Billy SG NJ .4
.37 Varejao,Anderson SF Cle .7
.43 Marshall,Donyell PF Tor .9
.43 Laettner,Christi PF Mia .7
.47 Cook,Brian PF LAL .8
.48 Lafrentz,Raef C Bos 1.1
.50 Bowen,Bruce SF SA .7
.50 Daniels,Antonio PG Sea 1.2
.51 Doleac,Michael C Mia .7
.51 Battier,Shane SF Mem .9
.52 Bonner,Matt PF Tor .8
.54 Griffin,Eddie PF Min 1.1
.54 Marion,Shawn PF Phe 1.4
.54 Booth,Calvin C Dal .7
.55 Jones,Jumaine SF LAL .8
.55 Finley,Michael SG Dal 1.1
.55 Padgett,Scott PF Hou .9
.56 Miller,Brad C Sac 1.5
.57 Duncan,Tim PF SA 2.1

Until Duncan, all are very-low TO players, per-minute. And only 1 player listed as PG (Daniels, who is really just a "guard"). Suggesting that the high-risk Assist category may tend to have built-in turnovers. System turnovers?
Back to top
View user's profile Send private message Send e-mail
HoopStudies



Joined: 30 Dec 2004
Posts: 526
Location: Near Philadelphia, PA

PostPosted: Tue Feb 22, 2005 12:10 pm Post subject: Reply with quote
WizardsKev wrote:
Check appendix 4 in "Basketball On Paper" for Dean Oliver's method.


Note that this is for teams. I'd probably do teams first then estimate individuals so that the total is roughly equivalent to teams. But I haven't spent much time on this.
_________________
Dean Oliver
Author, Basketball on Paper
http://www.basketballonpaper.com
Back to top
View user's profile Send private message Visit poster's website
replayhoops



Joined: 18 Feb 2005
Posts: 10
Location: Connecticut

PostPosted: Tue Feb 22, 2005 12:24 pm Post subject: Reply with quote
Thanks for the replies, guys. I just bought Dean's book and will take a look!

Dave


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 3:08 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
John Quincy



Joined: 01 Feb 2005
Posts: 14


PostPosted: Fri Feb 25, 2005 6:29 pm Post subject: Defensive plus/minus Reply with quote
http://maverickblog.blogspot.com/2005/0 ... -stat.html

Quote:
One of the statistics I have been looking at a lot lately has been the on court off court defensive stats at the 82 games site. Many believe figuring out who are good defenders and who are not is almost impossible. I think its difficult, but not impossible. I believe the following on court off court defensive statistic is one of the best barrometers I have found to giving a fairly accurate account as to who can play defense and who can't. What the following stat shows is who gave up more points against the other team when they were on the court as compared to their own teamates when they were off the court. I have calculated the top 100 or so players in the NBA plus anyone who recieved more then two votes in the ALL NBA defensive team voting in any of the last 3 seasons. I think its a complete list of the best players and defenders. The list is for the last 3 years starting with 3 years ago and finishing with their average for the last 3 seasons combined.
Back to top
View user's profile Send private message
nate33



Joined: 04 Feb 2005
Posts: 1


PostPosted: Fri Feb 25, 2005 7:30 pm Post subject: Reply with quote
Are those the "top 100 or so" best defensive players, or the top 100 players? How are the top 100 defined?

The reason I ask is that I noted the glaring omission of Brendan Haywood. The Wizards have allowed 8.1 fewer points per 100 possessions with Haywood on the floor this year. Last year, it was 5.5 points fewer. The year before, 2.9 fewer. Surely he ranks higher than the likes of Peja Stojakovic and Carmelo Anthony. I guess he didn't make it past whatever filter you used to weed out certain players.
Back to top
View user's profile Send private message
Minstrel



Joined: 24 Jan 2005
Posts: 1
Location: California, USA

PostPosted: Fri Feb 25, 2005 9:50 pm Post subject: Reply with quote
Quote:
I believe the following on court off court defensive statistic is one of the best barrometers I have found to giving a fairly accurate account as to who can play defense and who can't.


What are you comparing it to, to determine accuracy? That is, what's your measure of accuracy?

Is it just who you expect to see in such a list?
Back to top
View user's profile Send private message
John Quincy



Joined: 01 Feb 2005
Posts: 14


PostPosted: Sat Feb 26, 2005 11:51 am Post subject: Reply with quote
Sorry if it wasn't clear, but I'm not the author of that article. I just found it interesting so I posted the link to it.
Back to top
View user's profile Send private message
WizardsKev



Joined: 03 Jan 2005
Posts: 104
Location: Washington, DC

PostPosted: Sun Feb 27, 2005 9:30 am Post subject: Reply with quote
Maybe someone can fill in the blanks on his method. What is he calculating?
Back to top
View user's profile Send private message Visit poster's website AIM Address Yahoo Messenger
KD



Joined: 30 Jan 2005
Posts: 11


PostPosted: Sun Feb 27, 2005 4:10 pm Post subject: Reply with quote
I was about to rip the author's equation for having Reggie Miller make the top 20, but then I remebered Miller getting a solid defensive PER in the last Forecast. Looking it up, (3.69), Hollinger attributes his ranking to playing alongside and switching matchups with Ron Artest. Obviously, this hasn't been the case this season, as they've yet to play a game together.

How do most you regard Miller's defensive prowess?
Back to top
View user's profile Send private message
admin
Site Admin


Joined: 30 Dec 2004
Posts: 158
Location: Seattle

PostPosted: Sun Feb 27, 2005 5:41 pm Post subject: Reply with quote
He's averaging the defensive on-court/off-court difference over the last three years per 82games.com. It's more of an organization of that data than a new way to rate things. ...
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address Yahoo Messenger
WizardsKev



Joined: 03 Jan 2005
Posts: 104
Location: Washington, DC

PostPosted: Mon Feb 28, 2005 7:53 am Post subject: Reply with quote
Tracking this over 3 seasons makes this kinda interesting, but there needs to be some other kind of control here when it comes to identifying individual defensive achievement. Using the player pair data might be a good start. An example from this season's Wizards:

The Wizards are 3.4 pts per 100 possessions better defensively when Jamison is on the court. Wow -- Jamison must be a decent defender, right? Well, not really. Through 2/15, Jamison had played 59% of his minutes with Brendan Haywood -- and the team is 8.7 pts per 100 possessions better when Haywood is on the court. The player pair data reveals that the team gives up 7.6 pts per 48 minutes MORE when Jamison is on the court without Haywood.

Basically, I think this list is kinda interesting, but I think the method needs some work to make it worthwhile.

By the way: I don't think this is worth a new thread, but I've been tracking Wizards individual defense the past 20+ games. Against the Wizards, the Bulls shot 35-86 for the game. They were 0-13 on shots Haywood contested. The 0-13 vs. Haywood lowered the Bulls FG% from .479 (what they shot vs. all the other Wizards combined) to .407.

(I'll now duck the roundhouse headslap coming from North Carolina for my use of fg% instead of efg. Sorry, Dan, I haven't had a chance to do the calculations.)

Bonus factoid: The Wizards gave up 27 open or wide open shots vs. the Bulls. Chicago made 18 of them, including 7 3s.
Back to top
View user's profile Send private message Visit poster's website AIM Address Yahoo Messenger
Keven



Joined: 28 Feb 2005
Posts: 1


PostPosted: Mon Feb 28, 2005 10:49 pm Post subject: Reply with quote
Thanks for checking out my site. The stat looks at the top 100 overall players in the NBA[I used my Ball Theory list] plus any player that recieved more then 2 votes for the all defensive team in any of the last 3 seasons. That added another 7 players[Grant, Snow etc.]. So I feel the list is a comprehensive list of the top NBA players and top NBA defenders. Haywood may be an under rated defender[I like him], but he didn't get the all defensive votes and he came up short on my Ball Theory top 100. I didn't do every player in the NBA because when you get to players with small minutes the numbers get eratic. So it was more an idea of who can play defense among the best players in the NBA then who was the best defenders. So the Haywoods and Watsons of the world got left out.

The list was not a new stat, but just a 3 year look at the on court off court defensive numbers from 82games. There is no doubt there are weaknesses in the stat. Who you play with. Who replaces you ussually in the lineup. Tempo. Stat aberations like Reggie Millers from 2 seasons ago when he had a great number, but it was in very few games so likely very misleading. Etc, etc, etc. However, its just another defensive stat to look at and I felt many of the kinks of the stat got ironed out over a 3 year look. Maybe not all though.

Thanks again.
Back to top
View user's profile Send private message
WizardsKev



Joined: 03 Jan 2005
Posts: 104
Location: Washington, DC

PostPosted: Tue Mar 01, 2005 4:59 am Post subject: Reply with quote
Keven: I think the effort was worthwhile -- I think it can be improved by finding a way to incorporate other data (such as player pair info).

By the way: I think we've reached "...Kev..." saturation in basketball sabermetrics. If another one comes in, we may have to send him to hockey.


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 3:13 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
Doc319



Joined: 02 Feb 2005
Posts: 4


PostPosted: Thu Mar 03, 2005 4:13 am Post subject: ABA Reunion/Will A-Train Ever Arrive at HoF? Reply with quote
Although this is a stats oriented group, some readers may be interested in my first-hand account of the ABA Reunion that Fatty Taylor organized in Denver during NBA All-Star Weekend:

http://hoopshype.com/articles/aba_friedman.htm

The ABA Reunion and the 29th anniversary of the ABA Slam Dunk Contest bring to mind the career of Artis Gilmore. This year's Hall of Fame Finalists were announced during All-Star Weekend and Gilmore failed to make the cut. He produced 24,041 points and 16,330 rebounds during his ABA-NBA regular season career. He made the ABA or NBA All-Star team in 11 of his 17 seasons and led the Kentucky Colonels to the 1975 ABA Championship. I know that DeanO has said that he is more interested in using his formulas to evaluate current teams and to devise coaching strategies, but I would be interested to hear how Gilmore's career rates under the various formulas that have been discussed here. Perhaps DanR and others can weigh in on this subject.
Back to top
View user's profile Send private message Send e-mail Visit poster's website
KD



Joined: 30 Jan 2005
Posts: 9


PostPosted: Thu Mar 03, 2005 9:52 am Post subject: Reply with quote
Though I can name a half-dozen basketball books that I would classify as "better" or more important, I've probably read "Loose Balls" more than any other hoops tome out there. It's the perfect bathroom book, and with remembertheaba.com updating as little as it does, I needed the ABA fix. Great article.
Back to top
View user's profile Send private message
GreggGeth



Joined: 16 Feb 2005
Posts: 8


PostPosted: Thu Mar 03, 2005 10:14 am Post subject: Loose Balls Reply with quote
There's also a great chapter in there about Hubie Brown using stuff like the pythagorean model and scoring differential for his Kentucky Colonels.
Back to top
View user's profile Send private message
Mike G



Joined: 14 Jan 2005
Posts: 74
Location: Delphi, Indiana, USA

PostPosted: Thu Mar 03, 2005 11:56 am Post subject: Reply with quote
I've got Artis ranked #27 alltime. He was as high as #10, in the mid-80s.
Back to top
View user's profile Send private message Send e-mail
KD



Joined: 30 Jan 2005
Posts: 9


PostPosted: Thu Mar 03, 2005 2:17 pm Post subject: Re: Loose Balls Reply with quote
GreggGeth wrote:
There's also a great chapter in there about Hubie Brown using stuff like the pythagorean model and scoring differential for his Kentucky Colonels.


All I got was his insistence on scoring differential. Hubie and Costello were brilliant, but let's not give the mid-70s (and those lapels) that much credit.
Back to top
View user's profile Send private message
Doc319



Joined: 02 Feb 2005
Posts: 4


PostPosted: Fri Mar 04, 2005 12:33 am Post subject: Another ABA article Reply with quote
For KD and anyone else who has an ABA fix that can only be helped by reading more about the league with the tri-color ball, click on this link:

http://www.suite101.com/welcome.cfm/bas ... _spotlight

You will find the first installment of my three part series about James Silas.
Parts II and III, plus other ABA material (and NBA material) are forthcoming.


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 3:23 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
superbox



Joined: 29 Mar 2007
Posts: 4


PostPosted: Thu Mar 29, 2007 11:23 pm Post subject: How to explain Jason Collins Reply with quote
Jason Collins must be a freak of nature..
Or at the very least a freak of statistics.

His stats for this year

23.3 min, 2.4 pts, 4.1 reb, 0.6 ast, FG 37.5%, FT 47.5%, 0.45 blk, 0.97 TO, 3.5 PF

So statistically he is not only below average,
but near the bottom of the league in every single statistic!
(adjusted for his playing time & position)

He can't shoot, score, rebound, pass or block.
He fouls frequently, and has a lot of turnovers compared to possesions.

He has the worst PER for a regular starter in 20 years!

Yet he has started every single game for the Nets except one,
and amazingly has an on/off point differential of +5.5 which is second highest
on his team, higher than (admittedly overrated) Vince Carter.

Some might think this is a fluke but he has had a high on/off +- for nearly
every year in the league. (according to 82games.com)

Okay, so supposedly he is good at some things.
He's known as a good man to man defender, (without getting any steals or blocks)
sets good screens, gets loose balls.. whatever.
But it's not like he's the best defender in the league like Bruce Bowen
Ben Wallace, Ron Artest or even close to that.
And it's not that other guys don't set screens or don't get any loose balls,
it doesn't really require a lot of skill.

So he must be amazingly good at these "things that don't show up on the box score"
to be a plus on the Nets, when his stats are historically low.

It could be that his backup was even worse than him making is on/off +- look good,
but I can't really imagine a guy significantly worse than Jason Collins playing
in the league.

So I guess he is the best NBA player ever at doing "the little things".

Do you guys have another explanation?
Back to top
View user's profile Send private message
davis21wylie2121



Joined: 13 Oct 2005
Posts: 535
Location: Atlanta, GA

PostPosted: Thu Mar 29, 2007 11:49 pm Post subject: Reply with quote
It looks like Collins has just a ridiculous impact on NJ's defense. This year, he has shaved 9.2 points off of their DRtg while he's on the floor; last year, it was 5.6 points; in 2005, 13.2 (!!) points; in 2004, 6.0 points; and in 2003, 3.7 pts/48 off their defensive scoring average (on/off DRtg not available that year). Looking at the numbers, a typical year for the Ben Wallaces and Ron Artests of the world doesn't even come close to making as much of a defensive impact as a typical Collins season. He also does well throughout his career in Dean's boxscore DRtg. I don't know what he's doing out there, because like you said, he's not stealing or blocking shots that much, but whatever it is, it's making a big difference on the Nets' D.

Last edited by davis21wylie2121 on Fri Mar 30, 2007 12:03 am; edited 1 time in total
Back to top
View user's profile Send private message Visit poster's website
superbox



Joined: 29 Mar 2007
Posts: 4


PostPosted: Thu Mar 29, 2007 11:54 pm Post subject: I guess.. Reply with quote
Thanks for the insight.

So I guess he has an incredible Basketball IQ on defense..
that he somehow can't translate to offense,
because he's not athletic and can't shoot.

Yet his defensive rating is only slightly better than the
league average, so it seems it's really difficult to gauge his value.
Back to top
View user's profile Send private message
davis21wylie2121



Joined: 13 Oct 2005
Posts: 535
Location: Atlanta, GA

PostPosted: Fri Mar 30, 2007 12:02 am Post subject: Reply with quote
Sports Illustrated ran a piece on Dan Rosenbaum and Dean Oliver's methods in their 2005-06 NBA preview, and they highlighted Collins as one the "hidden gems" revealed by Dan's adjusted plus-minus system:

"According to Rosenbaum's calculations, Collins is not a stiff at all but one of the NBA's premier defensive centers... Over the last three seasons the Nets have been remarkably more effective at the defensive end with Collins in the lineup; they foul less, allow fewer free throws, rebound better, and allow fewer points. 'He's very consistent and consistently very good,' says Rosenbaum, 'meaning he's either the luckiest center alive and [opposing] teams just fall apart when he's on the court, or he's doing something.'"
Back to top
View user's profile Send private message Visit poster's website
superbox



Joined: 29 Mar 2007
Posts: 4


PostPosted: Fri Mar 30, 2007 12:05 am Post subject: Well.. Reply with quote
Well.. he could be a great trash talker. Smile

And it's funny that the team fouls less and rebounds better,
when Collins fouls a lot, and is a bad rebounder.

I guess that's the importance of boxing out.

And since he doesn't get many offensive rebounds,
I guess he gets back on defense real quick which helps the defense.
Back to top
View user's profile Send private message
davis21wylie2121



Joined: 13 Oct 2005
Posts: 535
Location: Atlanta, GA

PostPosted: Fri Mar 30, 2007 12:10 am Post subject: Re: I guess.. Reply with quote
superbox wrote:
Yet his defensive rating is only slightly better than the
league average, so it seems it's really difficult to gauge his value.


A career rating of 101 in leagues where the average is 105 is actually pretty good. If he were merely average on offense, his defense alone would propel his player winning percentage to .633 -- good for 52 wins over the course of 82 games (which is, perhaps not coincidentally, in the neighborhood of NJ's win totals over the past five years). I suppose the Nets plan is simple, then: never let Collins shoot (so that his abysmal ORtg never affects the team's efficiency), and just let him do his thing on defense. In fact, upon further inspection, Jason Collins is probably the unsung hero of New Jersey's recent run of success...
Back to top
View user's profile Send private message Visit poster's website
superbox



Joined: 29 Mar 2007
Posts: 4


PostPosted: Fri Mar 30, 2007 12:20 am Post subject: So.. Reply with quote
I guess playing only 4 players on offense doesn't affect the game that much as teams starting Rodman & Wallace have won the title.
But at least these guys were excellent offensive rebounders.

Maybe if Collins could rebound, the Nets could have won a championship.
I really don't think you could win a championship starting Collins,
unless you can substitute an offensive player for Collins on
virtually every possesion.
Back to top
View user's profile Send private message
deepak_e



Joined: 26 Apr 2006
Posts: 373


PostPosted: Fri Mar 30, 2007 1:07 am Post subject: Reply with quote
I don't think you can take too much from Jason Collin's Defensive Rating. Clearly, the most important things he's doing on the defensive end isn't really evident in the box scores. Outside of that, I think his Defensive Rating just reflects New Jersey's overall defensive rating which has generally been pretty good over the years.
Back to top
View user's profile Send private message
TrueHoop



Joined: 21 Aug 2006
Posts: 2


PostPosted: Tue Apr 03, 2007 2:40 pm Post subject: Theory Reply with quote
I have heard NBA players say Collins is a little bit, umm, extra fierce. If that's true, and he's someone players fear might injure them, that could have a big effect on shot selection without showing up in any traditional statistics. It would be interesting to see opponents' shot charts when he is in vs. out of the game. Maybe he leads the league in opponent layups never attempted.
_________________
Henry Abbott
http://www.truehoop.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
JNichols42887



Joined: 18 Aug 2005
Posts: 77


PostPosted: Tue Apr 03, 2007 10:07 pm Post subject: Reply with quote
He's definitely useful in certain situations. In the Nets' last two playoff series against the Heat, he played tremendous defense on Shaq.
Back to top
View user's profile Send private message AIM Address
John Hollinger



Joined: 14 Feb 2005
Posts: 107


PostPosted: Tue Apr 10, 2007 2:47 pm Post subject: Reply with quote
Collins is a vastly underrated defender -- he's an outstanding position defender and draws lots of offensive fouls especially against post players, and probably guards Shaq better than anyone in the league. But the "terrible backup effect" may also play a role -- Jersey hasn't had a backup C who could guard anyone since they let Mutombo go; the past couple years they've used a well-past-prime Cliffy and the waifish Krstic to replace him.


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 3:24 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
HoopStudies



Joined: 30 Dec 2004
Posts: 333
Location: Near Philadelphia, PA

PostPosted: Mon Mar 14, 2005 3:27 pm Post subject: Power of Context Reply with quote
There is this great book that was recommended to me by a friend in sports management. It's called the Tipping Point. Lots I could talk about in it, but I thought I'd highlight one of the basketball references:

Quote:

The mistake we make in thinking of character as something unified and all-encompassing is very similar to a kind of blind spot in the way we process information. Psychologists call this tendency the Fundamental Attribution Error (FAE), which is a fancy way of saying that when it comes to interpreting other people's behavior, human beings invariably make the mistake of overestimating the importance of fundamental character traits and underestimating the importance of the situation and context. We will always reach for a "dispositional" explanation for events, as opposed to a contextual explanation. In one experiment, for instance, a group of people are told to watch two sets of similarly talented basketball players, the first of whom are shooting baskets in a well-lighted gym and the second of whom are shooting baskets in a badly lighted gym (and obviously missing a lot of shots). Then they are asked to judge how good the players were. The players in the well-lighted gym were considered superior.


(emphasis is mine)

I found that experiment amazing. Sheesh. My god, I'd probably do the same thing even though I know how light affects my shooting.

The power of context is very big, which is why I say often in BoP to evaluate the team first, individual second. But it is easy, very easy to forget how teammates (context) affect individual performance. I never really thought that psychologists already know how easy we can be fooled to forgetting that. Anyway, this is a different subject, but I thought it an interesting point worthy of raising here.

ps -- it's a really interesting book.
_________________
Dean Oliver
Consultant to the Seattle Supersonics
Author, Basketball on Paper
http://www.basketballonpaper.com
Back to top
View user's profile Send private message Visit poster's website
WizardsKev



Joined: 03 Jan 2005
Posts: 409
Location: Washington, DC

PostPosted: Mon Mar 14, 2005 3:57 pm Post subject: Reply with quote
Great post, Dean. In some ways, it seems an obvious point -- that part of evaluation should include considering the context. In practice, though...
_________________
If you can't explain it simply, you don't understand it well enough.

-- Albert Einstein
Back to top
View user's profile Send private message Visit poster's website AIM Address Yahoo Messenger
Ed Küpfer



Joined: 30 Dec 2004
Posts: 479
Location: Toronto

PostPosted: Mon Mar 14, 2005 4:30 pm Post subject: Reply with quote
Dean, have you seen this paper?

Frank, Mark G., and Gilovich, Thomas, "The Dark Side of Self- and Social Perception: Black Uniforms and Aggression in Professional Sports", Journal of Personality and Social Psychology, 1988, 54:1:74-85

The authours found that black uniforms and aggression were linked in the minds of their subjects, and that football teams whose uniforms were black were called for more penalties than expected. I wonder to what extent this applies in basketball.

Further discussion can be found here.
_________________
ed
Back to top
View user's profile Send private message
gabefarkas



Joined: 31 Dec 2004
Posts: 401
Location: NYC

PostPosted: Mon Mar 14, 2005 5:08 pm Post subject: Reply with quote
The same guy wrote "Blink" which is also a pretty good read...
_________________
Statistics are like a woman's bikini. What it reveals can be fascinating, but what it conceals is ultimately critical!
Back to top
View user's profile Send private message Send e-mail AIM Address
bchaikin



Joined: 27 Jan 2005
Posts: 344
Location: cleveland, ohio

PostPosted: Mon Mar 14, 2005 5:16 pm Post subject: Reply with quote
well then it's plainly obvious why shooting percentages were lower in the NBA in the 1950s and early 1960s - everyone knows those gyms/arenas had poorer lighting Smile
Back to top
View user's profile Send private message Send e-mail Visit poster's website
admin
Site Admin


Joined: 30 Dec 2004
Posts: 514
Location: Seattle

PostPosted: Wed Mar 16, 2005 12:52 am Post subject: Reply with quote
I came across this today at the Management of Baseball blog. It's not precisely what we're talking about here, but something similar and interesting nonetheless.

The discussion is of a Bill James essay in the new SABR research journal about "the fog" and how analysts have considered randomness in a statistic to be proof it is not a skill, when it is not sufficient evidence of that.

Quote:
We ran astray because we have been assuming that random data is proof of nothingness, when in reality random data proves nothing. In essence, starting with Dick Cramer's article, Cramer argued that "I did an analysis which should have identified clutch hitters, if clutch hitting exists. I got random data; therefore, clutch hitters don't exist.


The best basketball application might be plus-minus data; while the noise inherent in plus-minus is good reason to take it with an enormous grain of salt, it isn't reason to dismiss it.
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
jeffpotts77



Joined: 18 Feb 2005
Posts: 69
Location: Cambridge, MA

PostPosted: Wed Mar 16, 2005 3:00 pm Post subject: Reply with quote
Dean,

I think for a shining example of the importance of context, and how teammates affect individual performance we need to look no further than Chris Webber's ugly 10 games with Philly so far (particularly his 5 point effort last night).

Webber was by no means the model of efficiency while in Sacramento, but he was very productive accross the board. Not only is he affected by having new teammates, but also by having a new coach who may not know how to, or be willing to call plays that optimize his talents.
Back to top
View user's profile Send private message Yahoo Messenger
HoopStudies



Joined: 30 Dec 2004
Posts: 333
Location: Near Philadelphia, PA

PostPosted: Wed Mar 16, 2005 3:19 pm Post subject: Reply with quote
admin wrote:
I came across this today at the Management of Baseball blog. It's not precisely what we're talking about here, but something similar and interesting nonetheless.

The discussion is of a Bill James essay in the new SABR research journal about "the fog" and how analysts have considered randomness in a statistic to be proof it is not a skill, when it is not sufficient evidence of that.

Quote:
We ran astray because we have been assuming that random data is proof of nothingness, when in reality random data proves nothing. In essence, starting with Dick Cramer's article, Cramer argued that "I did an analysis which should have identified clutch hitters, if clutch hitting exists. I got random data; therefore, clutch hitters don't exist.


The best basketball application might be plus-minus data; while the noise inherent in plus-minus is good reason to take it with an enormous grain of salt, it isn't reason to dismiss it.


This is absolutely correct. Absence of statistically significant results only means that evidence in the form you were looking doesn't exist. Ruling out all evidence requires an infinite number of null results (or probably just a lot). Similarly, I think there is a lot of value in the people who know when to take an 85% significant result as the truth -- through incorporating some sort of instinctive knowledge of what wasn't in the study. And equally likewise, knowing when a supposedly significant result is not answering the question you want is useful. My formal stat training didn't include a lot of high end courses, but I learned the lies and truths told by stats early on.
_________________
Dean Oliver
Consultant to the Seattle Supersonics
Author, Basketball on Paper
http://www.basketballonpaper.com


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 3:28 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
Ed Küpfer



Joined: 30 Dec 2004
Posts: 617
Location: Toronto

PostPosted: Tue Jan 18, 2005 5:48 pm Post subject: A Yahoo Group to store files Reply with quote
Over the last few weeks I've had the oportunity to exchange data files with members here. Unfortunately for me, I don't really have much webspace available to store these files, and I'm not a big fan of exchaging them by email, since the files aren't then available publically. So I've created a Yahoo Group where files can be stored. I'm not sure if you have to sign up for Yahoo membership or not, but it shouldn't be a hassle. I've turned off all the feaures of the group that I could except for the files section. If you want to share a spreadsheet with this group, you can upload it there, and announce it here.


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 3:30 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
Dennis_D



Joined: 28 Jul 2006
Posts: 2


PostPosted: Wed Aug 08, 2007 10:16 pm Post subject: Ranking of GM's drafting records (LONG!!) Reply with quote
Here is a system I have developed from evaluating the drafting of 17 GM's. Please give me some feedback.

Here's my GM rating (best to worst):
+4.1 - Rod Thorn of New Jersey
+4.0 - Geoff Petrie of Sacramento
+3.9 - Randy Pfund of Miami
+3.8 - John Paxson of Chicago
+3.1 - Bryan Colangelo of Phoenix
+3.0 - Don and Donnie Nelson of Dallas
+2.6 - Mitch Kupchak of the LA Lakers
+2.1 - Joe Dumars of Detroit
+1.6 - Indiana
+1.1 - Rick Sund of Seattle
-0.4 - Kevin McHale of Minnesota
-0.7 - RC Buford of San Antonio
-1.6 - Jerry West of Memphis
-2.1 - Carroll Dawson of Houston
-5.2 - Kevin O'Connor of Utah
-8.3 - Elgin Baylor of LA Clippers

I expect for each pick in the draft order to net a certain level of talent. If a GM drafts better talent with lottery picks than another GM does with late first round picks, that GM isn't necessarily doing a better job. So, I defined what level of talent I expect with each pick, then I will give GM credit (positive points) or blame (negative points) if he the player he drafts turns out better or worse than I expected. Here is the scoring chart:
Code:
----------------------------------------------------------------------
| | | | | | Deep | Never |
Pick |All-Star| Star |Starter |Rotation| Bench | Bench | in NBA |
======|========|========|========|========|========|========|========|
1-2 | 1 | 0 | -1 | -2 | -3 | -4 | -5 |
3-4 | 1.5 | 0.5 | -0.5 | -1.5 | -2.5 | -3.5 | -4.5 |
5-8 | 2 | 1 | 0 | -1 | -2 | -3 | -4 |
9-12 | 2.5 | 1.5 | 0.5 | -0.5 | -1.5 | -2.5 | -3.5 |
13-16 | 3 | 2 | 1 | 0 | -1 | -2 | -3 |
17-20 | 3.5 | 2.5 | 1.5 | 0.5 | -0.5 | -1.5 | -2.5 |
21-24 | 4 | 3 | 2 | 1 | 0 | -1 | -2 |
25-28 | 4.5 | 3.5 | 2.5 | 1.5 | 0.5 | -0.5 | -1.5 |
29-32 | 5 | 4 | 3 | 2 | 1 | 0 | -0.6 |
33-40 | 5.5 | 4.5 | 3.5 | 2.5 | 1.5 | 0.5 | -0.2 |
41+ | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
----------------------------------------------------------------------

The columns mean the following:

* All-Star - Selected as an All-Star
* Star - Not an All-Star, but one of the team's two best players
* Starter - Not a star, but starts when not injured
* Rotation - Plays when not injured, gets 15 - 24 mpg
* Bench - Plays 10 - 15 mpg, occasionally has a DNP-CD
* Deep Bench - Gets erratic minutes, will generally play in the NBA three seasons or less
* Never in NBA - Not made it on to to his team's roster or played just a handful of games


My assessment is based upon the player's current year unless he has played more than 4 seasons in the NBA, in which case it is based upon his fourth year. What matters is how the player did on the team that drafted him. The fact that George Lynch went on to be a starter on other teams doesn't change the fact that he didn't do well as a Laker draft pick. If a team trades a player, I assess his value in the trade (which will be based mostly on his performance to date) and then don't evaluate him further. If a team lets a draft pick walk, the draft pick gets a "Deep Bench" rating for the rest of his career. If a player who walked made the rotation for a season for his team before walking, I am going to give the GM a +0.5 bonus. The fourth year cut off is because most players are what they are after their fourth season and the chance of trade, career-ending injury, being beat out by a newer draft pick, etc. greatly increases after that.

Second round picks that make the bench for a year or two don't really matter for a team's fortunes. Also, a second round pick can get playing time one season because of injuries, shortness of bench, etc. and that shouldn't inflate a GM's score. To reflect that, here are some special rules for players picked 33 or later:
1. If the player plays only one season and then is waived, the GM gets a 0.1
2. If the player never played more than 500 minutes, then the GM gets a 0.1
3. If the player is at least a Bench player during his fourth season, the GM gets the full score
4. If the player is at least a Rotation player for two seasons, then GM gets the full score
5. If the player is at least a Starter player, then GM gets the full score
6. If the player doesn't meet #3, #4 or #5, then the GM gets 0.3 if the player is a Bench player and 0.5 if the player is a Rotation player

Alternatives considered
Assessing players for years 2-4 of their career
I tried this but realized it didn't work after I did Atlanta to Indiana. This alternative is considering the years 2-4 of a draft picks career, giving more weight to the most recent years. A draft pick's second year would have a weight of 1, third year a weight of 2, and fourth year a weight of 3. Then, the GM would be credited the weighted rating for the player. For most players, this didn't move the rating that much. For some players, this alternative didn't work. What to do about Eddie Griffin, the #7 pick who was a rotation player his second season, and then released after the season? What to do about Mehmet Okur, who was a rotation player his second season and then left Detroit through free agency? Isn't Orien Greene, who played 15.4 mpg in 80 games his rookie year before being cut a better late second round pick than someone who never played a game for the team?

Rating players how they did regardless of their team
Mehmet Okur was an absolute steal at #38. However, Detroit didn't recognize that and let him go as a free agent. Does his stellar play in Utah redeem the pick? Some think yes, but I don't

Adjusting scores for C's, PF' and PG's
Centers are extremely hard to draft. SG's and SF's are relatively easily. I played with the idea of adding a point for Centers that earn at least a Bench rating and 0.5 points for PF's and PG's that make at least a Bench rating. However, I don't know the positions of all NBA players and didn't know what to do with PF/C's and combo guards.

Seeing how each pick rated out of the available talent
For each draft pick, assess all of the players taken after that pick and how well the player ranked relative to the other players. For example, Chris Duhon would have better a pick that Sasha, but Sasha was the second best option. To do this would require evaluate and rank every player drafted for 4 years, which is more work than I want to do.

Known Flaws to This System
No adjustment for depth in draft
The 2003 draft has produced 5 All-Stars so far, so drafting one then was much than in the 1997 draft, which produced 2.

GM's of bad, young teams are overrated
Philidelphia, Boston and Atlanta score very high in these rankings because those teams have gotten rid of most of their veteran players and are playing most of their draft picks regardless if their are NBA quality or not. For example, of the 12 players who played the most minutes for Boston last season, 7 were drafted in the last 3 years and one more was drafted the year before. As a consequnce, players like Salim Stoudamire, Ryan Gomes Green and Willie Green get large positive ratings when they aren't NBA quality players. I have moved the rankings for those three teams out of the overall rankings.

Tenures to short to evaluate
Bernie Bickerstaff of Charlotte (3 seasons)
Danny Ferry of Cleveland (2 seasons)
Mark Warkentien of Denver (1 season)
Chris Mullin of Golden State (3 seasons)
Larry Harris of Milwaukee (3 seasons)
Jeff Bower of New Orleans (2 seasons)
Isiah Thomas of New York (3 seasons)
Otis Smith of Orlando (2 seasons)
Kevin Pritchard of Portland (New)
Bryan Colangelo of Toronto (1 season)
Ernie Grunfeld of Washington (3 seasons)

Ratings pulled because they are for bad, young teams
+12.0 - Billy King of Philadephia
+10.8 - Danny Ainge of Boston
+3.7 - Billy Knight of Atlanta

Billy Knight's picks (Atlanta) and my scoring of them:
2001
Wasn't GM yet
2002
Wasn't GM yet
2003
Boris Diaw (21) - Rotation then traded => 1
Travis Hansen (37) - Never in NBA (-0.2)
2004
Josh Childress (6) - Rotation (-1)
Josh Smith (17) - Star (2.5)
Donta Smith (34) - Deep Bench (0.1)
Royal Ivey (37) - Bench (0.3)
Viktor Sanikidze (42) - Never in NBA (0)
2005
Marvin Williams (2) - Starter (-1)
Salim Stoudamire (31) - Rotation (2)
Cenk Ayol (59) - Never in NBA (0)
Overall Score
+3.7

Danny Ainge's picks (Boston) and my scoring of them:
2001
Wasn't GM yet
2002
Wasn't GM yet
2003
Marcus Banks (13) - Bench then traded => -1
Kendrick Perkins (27) - Rotation (1.5)
Brandon Hunter (56) - Deep Bench (0.1)
2004
Al Jefferson (15) - Star (2)
Delonte West (24) - Starter then traded => 2
Tony Allen (25) - Rotation (1.5)
Justin Reed (40) - Deep Bench then traded => 0.1
2005
Gerald Green (18) - Rotation (0.5)
Ryan Gomes (50) - Starter (4)
Orien Greene (53) - Deep Bench => 0.1
Overall Score
+10.8

John Paxson's picks (Chicago) and my scoring of them:
2001
Wasn't GM yet
2002
Wasn't GM yet
2003
Kirk Hinrich (7) - Starter (0)
Mario Austin (36) - Never played in NBA (-0.2)
Tommy Smith (53) - Never played in NBA (0)
2004
Ben Gordon (3) - Star (0.5)
Luol Deng (7) - Star (1)
Chris Duhon (39) - Rotation (2.5)
2005
No picks
Overall Score
+3.8

Don and Donnie Nelson's picks (Dallas) and my scoring of them:
2001
Kyle Hill (43) - Never in NBA (0)
2002
Mladen Sekularac (54) - Never in NBA (0)
2003
Josh Howard (29) - All-Star (5)
2004
Pavel Podkolzin (21) - Never in NBA (-2) Note: Played 1 game
2005
No picks
Overall Score
+3.0

Joe Dumars' picks (Detroit) and my scoring of them:
2001
Rodney White (9) - Bench then traded => -1.5
Mehmet Okur (38) - Rotation, lost to FA because of CBA, became Star => 4.5
2002
Tayshaun Prince (23) - Starter (2)
2003
Darko Milicic (2) - Deep Bench then traded => -4
Carlos Delfino (25) - Not signed for a year, Rotation (0.5)
Andreas Glyniadakis (58) - Never in NBA (0)
2004
Rickey Paulding (54) - Never in NBA (0)
2005
Jason Maxiell (26) - Bench (0.5)
Amir Johnson (56) - Deep Bench (0.1)
Alex Acker (60) - Deep Bench (0)
Overall Score
+2.1

Carroll Dawson's picks (Houston) and my scoring of them:
Note: Daryl Morey is now GM for Houston
2001
Eddie Griffin (7) - Rotation then released => -2.5
2002
Yao Ming (1) - All-Star (0.5)
Bostjan Nachbar (15) - Bench then traded => -1
Tito Maddox (38) - Deep Bench (-0.1)
2003
Malick Badiane (44) - Never in NBA (0) => 0
2004
Vassilis Spanoulis (50) - Took until '06-'07 to sign
2005
Luther Head (24) - Rotation (1)
Overall Score
-2.1

Indiana Pacer's picks and my scoring of them:
2001
Jamison Brewer (40) - Deep Bench (0.1)
2002
Fred Jones (14) - Rotation (0)
2003
James Jones (49) - Rotation then traded for 2nd round pick => 1.5
2004
David Harrison (29) - Deep Bench (0)
Rashad Wright (59) - Never in NBA (0)
Overall Score
+1.6

Elgin Baylor's picks (Clippers) and my scoring of them:
2001
Traded away draft rights to Tyson Chandler
2002
Chris Wilcox (8) - Bench then traded for Vladimir Radmanovic => -2
Melvin Ely (12) - Deep Bench then traded => -2.5
Traded away rights to Mario Kasun (41)
2003
Chris Kaman (6) - Starter (0)
Sofoklis Schortsanitis (34) - Never in NBA (-0.2)
2004
Shaun Livingston (4) - Injury prone all career. Estimate: Rotation (-1.5)
Lionel Chalmers (33) - Deep Bench (0.1)
2005
Yaroslav Korolev (12) - Deep Bench (-2.5)
Daniel Ewing (32) - Bench (0.3)
Overall Score
-8.3

Mitch Kupchak's picks (Lakers) and my scoring of them:
2001
No picks
2002
Kareem Rush (20) - Traded for 2 high second round picks => -1
2003
Brian Cook (24) - Bench (0)
Luke Walton (32) - Starter (3)
2004
Sasha Vujacic (27) - Bench (0.5)
Marcus Douthit (56) - Never in NBA (0)
2005
Andrew Bynum (10) - Rotation (-0.5)
Ronny Turiaf (37) - Rotation (0.5)
Von Wafer (39) - Deep Bench (0.1)
Overall Score
+2.6

Jerry West's picks (Memphis) and my scoring of them:
Note: Chris Wallace is now the GM of Memphis
2001
Wasn't GM yet
2002
Drew Gooden (4) - Difficult to evaluate because traded mid-rookie season in a multi-player deal. Estimate: Starter (-0.5)
Robert Archibald (32) - Deep Bench (0)
2003
Troy Bell (16) - Deep Bench (-2) Note: Acquired draft day trade
Dahntay Jones (20) - Rotation (0.5) Note: Acquired draft day trade
2004
Andre Emmett (35) - Never in NBA (-0.2) Note: Acquired draft day trade and played 8 games
Antonio Burks (36) - Deep Bench (0.1) Note: Acquired draft day trade
Segei Lishouk (49) - Never in NBA (0)
2005
Hakim Warrick (19) - Rotation (0.5)
Overall Score
-1.6

Randy Pfund's picks (Miami) and my scoring of them:
2001
Ken Johnson (49) - Deep Bench (0.1)
2002
Caron Butler (10) - Difficult to evaluate because traded twice. Estimate: Starter (0.5)
Rasual Butler (53) - Bench then traded => 2
2003
Dwayne Wade (5) - All-Star (2)
Jerome Beasley (33) - Never in NBA (-0.2) Note: Actually played 2 games
2004
Dorell Wright (19) - Bench (-0.5)
Albert Miralles (39) - Never in NBA (0) Note: Acquired draft day trade and rights were later traded
Matt Freije (53) - Never in NBA (0) Note: Was waived by Heat before playing
2005
Wayne Simien (29) - Deep Bench (0)
Overall Score
+3.9

Kevin McHale's picks (Minnesota) and my scoring of them:
2001
No draft picks
2002
Marcus Taylor (52) - Never in NBA (0)
2003
Ndudi Ebi (26) - Deep Bench (-0.5)
Rick Rickert (55) - Never in NBA (0)
2004
Blake Stepp (58) - Never in NBA (0)
2005
Rashad McCants (14) - Rotation, had microfracture knee surgery => 0
Bracey Wright (47) - Deep Bench (0.1)
Overall Score
-0.4

Rod Thorn's picks (New Jersey) and my scoring of them:
2001
Richard Jefferson (13) - Star (2)
Jason Collins (18) - Starter (1.5)
Brandon Armstrong (23) - Deep Bench (-1)
Brian Scalabrine (35) - Bench (1.5)
2002
Nenad Krstic (24) - Starter (2)
Tamar Slay (54) - Deep Bench (0.1)
2003
Zoran Planinic (22) - Deep Bench (-1)
Traded rights to Kyle Korver (51)
2004
Traded rights to Viktor Khryapa (22)
Christian Drejer (51) - Never in NBA (0)
2005
Antoine Wright (15) - Bench (-1)
Mile Ilic (43) - Not signed until '06-'07 season
Overall Score
+4.1

Billy King's picks (Philadephia) and my scoring of them:
2001
Samuel Dalembert (26) - Starter (2.5)
Damone Brown (37) - Deep Bench (0.1)
Alvin Jones (57) - Deep Bench (0.1)
2002
Traded rights to Jiri Welsch
John Salmons (26) - Rotation (1.5)
Sam Clancy (44) - Never in NBA (0)
Randy Holcomb (56) - Never in NBA (0)
2003
Willie Green (41) - Rotation (3)
Traded rights to Paccelis Morlende (50)
Kyle Korver (51) - Rotation (3)
2004
Andre Iguodala (9) - Star (1.5)
2005
Louis Williams (45) - Bench (0.3)
Overall Score
+12.0

Bryan Colangelo's picks (Phoenix) and my scoring of them:
Note: Colangelo is now the GM of Toronto
2001
Alton Ford (51) - Deep Bench (0.1)
2002
Amare Stoudemire (9) - All-Star (2.5)
Casey Jacobsen (22) - Rotation then traded => 1
2003
Zarko Cabarkapa (17) - Deep Bench then traded for 2 second round picks => -1
Leandrinho Barbosa (28) - Rotation (1.5) Note: Acquired draft day trade
2004
Jackson Vroman (31) - Deep Bench then traded => 0
2005
Dijon Thompson (24) - Deep Bench (-1)
Overall Score
+3.1

Geoff Petrie's picks (Sacramento) and my scoring of them:
2001
Gerald Wallace (25) - Deep Bench then selected in expansion draft => -0.5
Maurice Jeffers (55) - Never in NBA (0)
2002
Corsley Edwards (58) - Never in NBA (0) Note: played 10 games for Hornets in '04-'05 season
2003
None
2004
Kevin Martin (26) - Star (3.5)
Ricky Minard (47) - Never in NBA (0)
2005
Francisco Garcia (23) - Rotation (1)
Overall Score
+4.0

RC Buford's picks (San Antonio) and my scoring of them:
2001
Wasn't GM yet
2002
Wasn't GM yet
2003
No Draft Picks
2004
Beno Udrih (28) - Bench (0.5)
Romain Sato (52) - Never in NBA (0)
Sergei Karaulov (57) - Never in NBA (0)
2005
Ian Mahinmi (28) - Never in NBA (-1.2)
Overall Score
-0.7

Rick Sund's picks (Seattle) and my scoring of them:
2001
Vladimir Radmanovic (12) - Rotation (-0.5)
Earl Watson (40) - Bench then let go as FA => 0.1
2002
Peter Fehse (49) - Never in NBA (0)
2003
Nick Collison (12) - Starter (0.5)
Luke Ridnour (14) - Starter (1)
Paccelis Morlende (50) - Never in NBA (0) Note: Acquired draft day trade
2004
Robert Swift (12) - Bench (-1.5)
David Young (41) - Never in NBA (0)
[2005]
Johan Petro (25) - Rotation (1.5)
Mickael Gelabale (48) - Didn't play until '06-'07 season
Overall Score
+1.1

Kevin O'Connor's picks (Utah) and my scoring of them:
2001
Raul Lopez (24) - Two years to sign, Deep Bench, then traded => -1
Jarron Collins (53) - Rotation (3)
2002
Curtis Borchardt (18) - Deep Bench then traded => -1.5 Note: Acquired draft day trade
2003
Aleksandar Pavlovic (19) - Bench then taken in expansion draft => -1.5
Maurice Williams (47) - Bench then lost to FA => 0.1
2004
Kris Humphries (14) - Deep Bench then traded => -2
Kirk Snyder (16) - Deep Bench then traded => -2
2005
Deron Williams (3) - Starter (-0.5)
CJ Miles (34) - Deep Bench (0.1)
Robert Whaley (51) - Deep Bench then traded => 0.1
Overall Score
-5.2

Edit: Corrected the rating for Dahntay Jones from Deep Bench to Rotation, which increased Jerry West's rating from -3.6 to -1.6.

Edit: I didn't realize that Dumars couldn't have re-signed Okur because of the CBA. Changed the rating of Okur from 1.0 to 4.5, increasing Dumars rating from -1.4 to +2.1

Last edited by Dennis_D on Sat Aug 18, 2007 9:55 pm; edited 1 time in total
Back to top
View user's profile Send private message
Mountain



Joined: 13 Mar 2007
Posts: 279


PostPosted: Thu Aug 09, 2007 10:12 am Post subject: Reply with quote
The sliding credit conceptual framework makes sense but here is one opinion about some particulars:

Getting an all-star is still an achievement even very early so I might make that at least a +2 at the start of the table, maybe even +3. (And shift at least some of the others accordingly.)

There will always be cases of players who fall between categories but I'd suggest below average starters get credit halfway between a solid starter and a rotation guy. I think that is the spot where splitting hairs further would be most useful.

After rotation moving to the right, I might chop the size of the credit steps to 1/2 point. To me it matters less after rotation and is often a lot about limited opportunity and less about major gradations in talent level.

I might look at reducing the change for moving down the table a row to .33 or even .25, or perhaps start at .5 and slowly shifts to lower or perhaps reduce the number of rows. I feel picking late gems or even just better than expected is maybe getting too much credit compared to just picking a gem period or even solid value for the slot, which is still hard. Two all-stars from 5-8 equal one star at 30? I'd take the former. It is both a far better circumstance but, apart from that, I think it is also a more powerful, better performance. A bench player at 41+ equal to a star at 13-16? I'd take the later. I'd consider disrupting the consistency of the steps further beyond the above suggestions if subjectively I felt it was warranted to get different picks of equal credit closer to feeling equal across the many spots on the table.

Welcome, thanks for sharing the ratings and hope the comments that come help.

Last edited by Mountain on Thu Aug 09, 2007 12:54 pm; edited 1 time in total
Back to top
View user's profile Send private message
Rasta978



Joined: 26 Mar 2007
Posts: 46
Location: Orlando, FL

PostPosted: Thu Aug 09, 2007 10:57 am Post subject: Reply with quote
Dennis,

Very interesting analysis. You've obviously put a lot of time and thought into this. Great contribution to the board.

Any chance you could score John Gabriel, gm of the Magic from 1993 through 2003? I always considered him a terrible gm who wasted draft picks on the likes of Gertt Hammink, Brooks Thompson and Rodney Dent. I'd be curious to see how you rate him.

Code:
2003 NBA 1 15 Reece Gaines University of Louisville
2 13 Zaza Pachulia
2002 NBA 1 18 Curtis Borchardt Stanford University
2001 NBA 1 15 Steven Hunter DePaul University
1 22 Jeryl Sasser Southern Methodist University
2 3 Omar Cook St. John's University
2000 NBA 1 5 Mike Miller University of Florida
1999 NBA 2 9 Laron Profit University of Maryland
1998 NBA 1 12 Michael Doleac University of Utah
1 13 Keon Clark University of Nevada, Las Vegas
1 15 Matt Harpring Georgia Institute of Technology
2 13 Miles Simon University of Arizona
1997 NBA 1 17 Johnny Taylor University of Tennessee at Chattanooga
2 18 Eric Washington University of Alabama
1996 NBA 1 27 Brian Evans Indiana University
2 20 Amal McCaskill Marquette University
1995 NBA 1 25 David Vaughn University of Memphis
1994 NBA 1 27 Brooks Thompson Oklahoma State University
2 4 Rodney Dent University of Kentucky
1993 NBA 1 1 Chris Webber University of Michigan
1 26 Geert Hammink Louisiana State University


A few draft day trades to note:
Webber for Hardaway
Eric Washington for Jason Lawson
Traded Horace Grant for Corey Maggette in 1999 draft
Curtis Borchard for Ryan Humphrey
Laron Profits' draft rights traded on draft night
Back to top
View user's profile Send private message
THWilson



Joined: 19 Jul 2005
Posts: 142
Location: phoenix

PostPosted: Thu Aug 09, 2007 12:18 pm Post subject: Reply with quote
I enjoyed this, like the concept.

I would re-jigger some of the player rankings. Barbosa is a high contributing sixth man on a very good team, Gomes, Collison and Marvin Williams start on very bad teams. If the rankings were more rooted in contribution, rather than minutes, you could weight these players more appropriately.

Win Shares seem like a perfect measure since it combines minutes with contribution.
Back to top
View user's profile Send private message
Ben



Joined: 13 Jan 2005
Posts: 211
Location: Iowa City

PostPosted: Thu Aug 09, 2007 2:16 pm Post subject: Reply with quote
THWilson wrote:
I enjoyed this, like the concept.

I would re-jigger some of the player rankings. Barbosa is a high contributing sixth man on a very good team, Gomes, Collison and Marvin Williams start on very bad teams. If the rankings were more rooted in contribution, rather than minutes, you could weight these players more appropriately.

Win Shares seem like a perfect measure since it combines minutes with contribution.


I was thinking about comparing the nth draft pick to the nth best player in win shares. Thanks to Justin, this would be pretty easy to do.
Back to top
View user's profile Send private message
Jakedfw



Joined: 27 May 2006
Posts: 26


PostPosted: Fri Aug 10, 2007 11:34 pm Post subject: Need to include free agents Reply with quote
I think that you really need to add a scoring system for non-drafted free agents. For example, Dallas signed Marquis Daniels as a non-drafted free agent, and he was second team all-rookie his rookie year. Miami signed Udonis Haslem as an undrafted free agent, and he went on to start on their NBA title team.
Back to top
View user's profile Send private message Send e-mail
LanierFan



Joined: 12 Aug 2007
Posts: 1


PostPosted: Sun Aug 12, 2007 11:54 pm Post subject: Reply with quote
The scoring classification seems a little mushy to me. All Stars can be easier or harder to draft depending on the conference you're in, and the "Star" category punishes GMs for doing the rest of their job well -- i.e., creating a lineup of above-average veterans via trade or free agency.

One other point: excluding Mehmet Okur is a head-scratcher for me. You say that Detroit "didn't recognize" him to be a steal, which is ridiculous. What they recognized was that, prior to the closing of the Arenas loophole, they couldn't keep Memo and the player who'd brought them the 2004 championship, Rasheed Wallace. That was an unusual situation, but one can control for it in far more reasonable ways than arbitrarily deducting 60% of his value.
Back to top
View user's profile Send private message
Dennis_D



Joined: 28 Jul 2006
Posts: 2


PostPosted: Sat Aug 18, 2007 9:57 pm Post subject: Reply with quote
Corrected the ratings of Okur and Dahntay Jones
Back to top
View user's profile Send private message
Ben



Joined: 13 Jan 2005
Posts: 211
Location: Iowa City

PostPosted: Thu Aug 23, 2007 4:49 pm Post subject: Reply with quote
I decided to try out my proposed method of comparing the nth draft pick to the nth best player in the draft as rated by career win shares. Note that this has the odd effect of a GM never getting positive credit when they have the number one pick.

To illustrate with an example, take Joe Dumars with the number 2 pick in the NBA draft, we compare his pick of Darko (15 win shares) to Dwyane Wade (111 win shares) to get a score of -96.

When a player was traded before he ever played, I decided to grade the GM of the franchise where the player actually ended up. Thus, Gabriel is graded on Penny rather than on Webber. In this particular instance, the value is zero for both cases, so it didn't have any effect on Gabriel's rating.



Code:

2003 Reece Gaines -27
2003 Zaza Pachulia +32
2002 Ryan Humphrey -15
2001 Steven Hunter -24
2001 Jeryl Sasser -37
2001 Omar Cook -7
2000 Mike Miller +20
1999 Corey Maggette 0
1998 Michael Doleac -55
1998 Keon Clark -28
1998 Matt Harpring +33
1998 Miles Simon -3
1997 Johnny Taylor -55
1997 J. Lawson +1
1996 Brian Evans -17
1996 Amal McCaskill +4
1995 David Vaughn -15
1994 Brooks Thompson -4
1994 Rodney Dent -7
1993 Penny Hardaway 0
1993 Geert Hammink -6


All in all, a pretty sorry record. If the drafts were "done over" and everybody drafted by win shares, then Orlando would gain 210 win shares.



Rasta978 wrote:
Dennis,

Very interesting analysis. You've obviously put a lot of time and thought into this. Great contribution to the board.

Any chance you could score John Gabriel, gm of the Magic from 1993 through 2003? I always considered him a terrible gm who wasted draft picks on the likes of Gertt Hammink, Brooks Thompson and Rodney Dent. I'd be curious to see how you rate him.


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 3:36 am 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
supersub15



Joined: 21 Sep 2006
Posts: 78


PostPosted: Thu Sep 13, 2007 6:20 am Post subject: A way to define 3-pt specialists Reply with quote
Is Kyle Korver a 3-pt specialist? How about Bruce Bowen (other than his defensive role) or Antoine Walker?

I've always wondered how media people pigeon-holed certain players just because they were proficient at shooting the 3. So, I wanted a more quantitative way of measuring this role.

My initial thought is to see how many threes a player shot out his total FGA, but maybe a better way is to tabulate it as such: 3PA/(FGA-3PA) to get the number of attempted 3s for every attempted 2.

My problem is two-fold:
1. The league average is 0.2692 3PA for every 2PA. How do I set the cut-off for nominating a player as a 3-pt specialist? Is it 0.75 for 1, 1 for 1? Do I set it arbitrarily or is there a more scientific way to do it?

2. Do I integrate 3P shooting proficiency into this, i.e. is the fact that Walker shoots .86 threes for every 2PA make him a 3-pt specialist even though he only shoots at a .275 clip? What's the best way to integrate 3FG%?
Back to top
View user's profile Send private message
Analyze This



Joined: 17 May 2005
Posts: 354
Location: Belgium

PostPosted: Thu Sep 13, 2007 7:05 am Post subject: Reply with quote
For me a 3p specialist is somenone who 1) shoots a high efficiency from behind the 3p line and 2) who also shoots a lot of 3p attempts. Which % of his total attempts need to be 3's and which % does he need to make before you call him a 3p specialist? That will differ from person to person.
_________________
Where There's a WilT There's a Way
Back to top
View user's profile Send private message
admin
Site Admin


Joined: 30 Dec 2004
Posts: 677
Location: Seattle

PostPosted: Thu Sep 13, 2007 11:49 am Post subject: Reply with quote
To me, a three-point specialist should also use relatively few possessions. There are exceptions to this (the WNBA's Shanna Crossley, who has never been shy about pulling the trigger, comes to mind), but I don't think of someone like Walker or Baron Davis as a three-point specialist, even though they do attempt a relatively high percentage of threes.
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
THWilson



Joined: 19 Jul 2005
Posts: 124
Location: phoenix

PostPosted: Thu Sep 13, 2007 12:07 pm Post subject: Reply with quote
How about finding the 90th percentile for whatever stat you choose as a somewhat less arbitrary threshold?

admin wrote:
To me, a three-point specialist should also use relatively few possessions. There are exceptions to this (the WNBA's Shanna Crossley, who has never been shy about pulling the trigger, comes to mind), but I don't think of someone like Walker or Baron Davis as a three-point specialist, even though they do attempt a relatively high percentage of threes.


Interesting thought. What about Ray Allen?
Back to top
View user's profile Send private message
admin
Site Admin


Joined: 30 Dec 2004
Posts: 677
Location: Seattle

PostPosted: Thu Sep 13, 2007 12:18 pm Post subject: Reply with quote
THWilson wrote:
Interesting thought. What about Ray Allen?

That makes me realize I chose two examples who shoot a relatively low percentage and shoot a lot of threes. Allen is different, but I still wouldn't consider him in that group. To me, specialist implies it's about all they do on offense.
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
admin
Site Admin


Joined: 30 Dec 2004
Posts: 677
Location: Seattle

PostPosted: Thu Sep 13, 2007 12:31 pm Post subject: Reply with quote
So here's the list (minimum 100 FGA) of highest three attempt percentage in 06-07:

Code:
Player 3A%
----------------
Barry .675
D. Jones .651
Vujacic .601
Posey .600
Horry .594
J.R. Smith .593
Diawara .582
Scalabrine .576
Head .573
D. Johnson .751

I actually already have a junk number which measures what I was talking about: it's 3A% squared divided by possession percentage squared. Here's that list:

Code:
Player 3Rat
----------------
Scalabrine 26.6
Diawara 22.6
Battier 22.0
Posey 21.4
Bowen 21.1
Horry 18.8
Barry 18.3
D. Jones 16.4
D. Gibson 16.2
Vujacic 14.9

The biggest difference there is that J.R. Smith, who shot a ton of threes but was also a fairly big part of the Denver offense, moves off. Shane Battier, who almost never shot in Houston, moves way up and the positions slide a little. Not a huge difference either way. Now if you factor in three-point percentage, that's a different story.
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
BadgerCane



Joined: 05 Sep 2007
Posts: 3
Location: University of Miami-Florida

PostPosted: Thu Sep 13, 2007 12:51 pm Post subject: Reply with quote
I had thought the standard for the "3-Point Specialist" was 3PMade/FGA. That way you integrate shooting percentage. This past year the top of the list looks like this: B Barry, L Head, D Jones, S Battier, JR Smith, B Scalabrine, B Nachbar, D Gibson, J Posey[/code]
Back to top
View user's profile Send private message Send e-mail AIM Address
Charles



Joined: 16 May 2005
Posts: 58


PostPosted: Thu Sep 13, 2007 4:35 pm Post subject: Reply with quote
admin wrote:
To me, a three-point specialist should also use relatively few possessions. There are exceptions to this (the WNBA's Shanna Crossley, who has never been shy about pulling the trigger, comes to mind), but I don't think of someone like Walker or Baron Davis as a three-point specialist, even though they do attempt a relatively high percentage of threes.


I agree that Baron Davis and Antoine Walker are not three-point specialists. However, my reasoning would be different.

In my mind, a specialist is someone who has "special" skill in a particular area. It is not implicit that the skill should be exclusive of other skills. Therefore, I don't see using "few possessions" as relevant.

A person can have more than one specialty. Jason Kapono's only specialty is three-point shooting (although strangely he doesn't make your lists.) However, Steve Nash is a three point specialist -- and also a play-making specialist. Anthony Parker is a three-point specialist -- and also a defensive specialist. Brent Barry and Ray Allen are certainly established three point specialists with fairly balanced offensive skills.

However, Antoine Walker is not a three point specialist. Doing something frequently doesn't make you a specialist -- you have to be good at it. It wouldn't matter whether 'toine shot a trey on every possession. If he is shooting .275 he is not a three-point specialist.

Here are a couple of attempts to quantify this.

In terms of efficiency I would just look at: (3PM * 3) / (3PA * lgPPA) (2006-07 league points per scoring attempt = 1.083) This is obviously just points per 3PA compared to an "average" attempt that season, but I like it because it makes is easy to compare the "extra" produced by one individual compared to another.

3-Point Efficiency (minimum 100 3PA)
1.42 Jason Kapono
1.28 Walter Herrmann
1.26 Steve Nash
1.24 Brent Barry
1.22 Luther Head
1.22 Anthony Parker
1.21 Jason Terry
1.20 Leandro Barbosa
1.20 Al Harrington
1.19 Kyle Korver

Jason Kapono scored an "extra" 42% per attempt -- the best by quite a large margin. Antoine Walker scored 0.89 or 11% under the league average points per attempt on his 3PAs.

However, it's often more meaningful to see these kinds of numbers in terms of production, rather than ratios or percentages. You can get the "extra" points a player produces with three-point shots per game with: ((3PM * 3) - (3PA * lgPPA)) / GP. These methods get rid of many of the the marginal players who creep into some lists.

3-Point "extra" Production per Game (minimum 50 GP)
2.62 Jason Kapono
2.15 Brent Barry
2.10 Luther Head
1.80 Eddie House
1.76 Leandro Barbosa
1.73 Steve Nash
1.54 Bostjan Nachbar
1.44 Jason Terry
1.33 Kyle Korver
1.28 Raja Bell

Again, some high volume 3-point shooters such as Walker (-1.00), Baron Davis (-0.77) Arenas (-0.24) and Billups (-0.22) actually cost their team points each game. No matter how much they may shoot threes, they are not what I would call specialists (or, at least, they didn't perform like they were that year.)
Back to top
View user's profile Send private message
tiongkiat



Joined: 30 Apr 2007
Posts: 2


PostPosted: Thu Sep 13, 2007 10:49 pm Post subject: Reply with quote
Charles wrote:
Again, some high volume 3-point shooters such as Walker (-1.00), Baron Davis (-0.77) Arenas (-0.24) and Billups (-0.22) actually cost their team points each game. No matter how much they may shoot threes, they are not what I would call specialists (or, at least, they didn't perform like they were that year.)


It might be enlightening to remove both FGM and FGA at end-of-shot-clock or end-of-period bail-out shots, perhaps with 2 seconds or less left. Players who take many possession bailout shots would be punished for being responsible for the offensive set's failure or the circumstances of the clock, unless the rest of their 3-pt shooting is even worse...
Back to top
View user's profile Send private message
94by50



Joined: 01 Jan 2006
Posts: 403
Location: Phoenix

PostPosted: Thu Sep 13, 2007 10:54 pm Post subject: Reply with quote
BadgerCane wrote:
I had thought the standard for the "3-Point Specialist" was 3PMade/FGA. That way you integrate shooting percentage. This past year the top of the list looks like this: B Barry, L Head, D Jones, S Battier, JR Smith, B Scalabrine, B Nachbar, D Gibson, J Posey[/code]

Good idea. That list is a good start. Guys who make as many threes as possible, but shoot as little as possible, also...
Back to top
View user's profile Send private message
thref23



Joined: 13 Aug 2007
Posts: 18


PostPosted: Fri Sep 14, 2007 12:03 am Post subject: Reply with quote
When putting together composite scores, I ranked players by 3 pt FG%, punished players who didn't take so many 3 point attempts (that way Elton Brand wasn't tops in the league), and weighted the results slightly on a per team basis (players on teams with a better offense were more likely to get open looks, so they were slightly punished).

For the best 10 3 point shooters in the league I came up with:

kapono,jason
herrmann,walter
korver,kyle
head,luther
barry,brent
udoka,ime
parker,anthony
szczerbiak,wally
house,eddie
ray,allan


If one were to take each player's 3 point ranking and divide it by their overall ranking (to find guys who's 3 pt shooting ability makes up the highest percentage of their overall ability), filter ,players who scored below 20 (i.e. players like Primo Brezec Adam Morrison and David Noel), take the top 20 and rank the top 20 by 3 pt shooting %, the results are:

korver,kyle
szczerbiak,wally
ray,allan
gibson,daniel
scalabrine,brian
bogans,keith
azubuike,kelenna
webster,martell
stoudamire,salim
butler,rasual

If you place the cutoff at 40 however, which is really what seems to be the minimum score for a good role player, the results become:

kapono,jason
herrmann,walter
korver,kyle
head,luther
udoka,ime
szczerbiak,wally
house,eddie
mobley,cuttino
nachbar,bostjan
murphy,troy

And I think thats your best list. Spots 11-20, btw, are as follows (Bruce Bowen does end up top 20 by this measure, but I think these guys are a step below "specialists"):

giricek,gordan
redick,j.j.
harrington,al
pargo,jannero
wilkins,damien
jones,damon
bowen,bruce
cook,brian
james,mike
nocioni,andres

Alternatively, I tried simply multiplying each player's overall score by the percentage of 3 pt shooting to overall score, filtered out players with scores higher than 70 (too good to simply be a specialist), and came up with the following top 20:

kapono,jason
herrmann,walter
korver,kyle
head,luther
udoka,ime
szczerbiak,wally
house,eddie
ray,allan
gibson,daniel
mobley,cuttino
nachbar,bostjan
carroll,matt
scalabrine,brian
pavlovic,sasha
murphy,troy
giricek,gordan
turkoglu,hedo
redick,j.j.
bogans,keith
harrington,al

That looks more accurate to me, actually. If you filter out players that scored above 60 overall, Turkoglu is replaced by Jannero Pargo.

Bowen, by this measure, would be #40, or #34 if players scoring above 60 overall are filtered out.

Last edited by thref23 on Fri Sep 14, 2007 12:22 am; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail
basketballvalue



Joined: 07 Mar 2006
Posts: 51


PostPosted: Fri Sep 14, 2007 12:21 am Post subject: Re: A way to define 3-pt specialists Reply with quote
supersub15 wrote:

I've always wondered how media people pigeon-holed certain players just because they were proficient at shooting the 3. So, I wanted a more quantitative way of measuring this role.


It's semantics, but I think the original post captures what a specialist is. I would contrast Steve Kerr and Jason Kapono with Steve Nash. Charles' stats are valuable, but I don't think they capture specialization (associated with concentration in a particular area). I think Charles' stats capture talent (expertise). Particulary given supersub's original question, I don't think Steve Nash should be pigeon-holed into the Kerr/Kapono role because Charles shows he shoots the three almost comparably.

I think of the analogy in medicine, people aren't both superstar caridologists and orthopedic surgeons. They are specialists in one field or the other. As a player with multiple talents, Steve Nash is closer to a superstar primary care physicians, perhaps the surgeon general.

Similarly, I'd say that someone in the Walker category who concentrates on shooting the 3 on offense but doesn't do it as well is a specialist that is not as talented as others. Someone out there is the worst cardiologist in the US, but they're still a specialist.


Thanks,
Aaron

http://www.basketballvalue.com
http://www.82games.com/barzilai1.htm
Back to top
View user's profile Send private message
tmansback



Joined: 12 Aug 2005
Posts: 120


PostPosted: Fri Sep 14, 2007 12:36 am Post subject: Reply with quote
I would consider a 3 point specialist someone that would not be in the league if not for his ability to make the 3 point shot. I wouldn't consider guys like Brent Barry, Troy Murphy, or Brian Cook specialist. Only guys I would consider specialist are players that enter the game with the sole purpose to make 3 pointers. Steve Novak is a specialist. Anytime he is in the game its only to make 3 pointers or long jump shots. JJ Redick is pretty close to a specialist. Salim Stoudamire and Damon Jones are pretty much specialist only there to shoot. Kapano for most of his career I would consider specialist. Perhaps he has improved on other aspects enough you can say he a little more now.

If I was to look for who the 3 point specialist are I would take the players who play limited minutes who shoot high number of 3 point shots. Most likely these players have pretty low PERs and don't provide anything defensively.
Back to top
View user's profile Send private message
thref23



Joined: 13 Aug 2007
Posts: 18


PostPosted: Fri Sep 14, 2007 12:40 am Post subject: Reply with quote
BadgerCane wrote:
I had thought the standard for the "3-Point Specialist" was 3PMade/FGA. That way you integrate shooting percentage. This past year the top of the list looks like this: B Barry, L Head, D Jones, S Battier, JR Smith, B Scalabrine, B Nachbar, D Gibson, J Posey[/code]


I like this more than using 3 point attempts. A question, IMO, is whether a "3 point specialist" is allowed to specialize in multiple categories. Is Shane Battier "a specialist" or is he a really good player that shoots threes well (or both).

Perhaps a "fake 3 pt specialist" category would be interesting to compute. Meaning players that put up 3 pt attempts like they are specialists, but shoot too poorly to be specialists. Antoine Walker might top the league in that category.
Back to top
View user's profile Send private message Send e-mail
thref23



Joined: 13 Aug 2007
Posts: 18


PostPosted: Fri Sep 14, 2007 12:46 am Post subject: Reply with quote
tmansback wrote:
I would consider a 3 point specialist someone that would not be in the league if not for his ability to make the 3 point shot.


You posted this at the same time I was thinking it (and posting something similar). Personally, I would simply be inclined to view a 3 pt specialist as a player who's 3 pt shooting ability accounts for a certain percentage of his overall ability.

supersub15



Joined: 21 Sep 2006
Posts: 273


PostPosted: Fri Sep 14, 2007 6:35 am Post subject: Reply with quote
Charles wrote:
However, it's often more meaningful to see these kinds of numbers in terms of production, rather than ratios or percentages. You can get the "extra" points a player produces with three-point shots per game with: ((3PM * 3) - (3PA * lgPPA)) / GP. These methods get rid of many of the the marginal players who creep into some lists.

3-Point "extra" Production per Game (minimum 50 GP)
2.62 Jason Kapono
2.15 Brent Barry
2.10 Luther Head
1.80 Eddie House
1.76 Leandro Barbosa
1.73 Steve Nash
1.54 Bostjan Nachbar
1.44 Jason Terry
1.33 Kyle Korver
1.28 Raja Bell


I got the correct results from the first formula, but I got different results with the second one:
Jason Kapono 1.441
Steve Nash 1.270
Luther Head 1.209
Leandro Barbosa 1.196
Jason Terry 1.053
Raja Bell 0.998
Brent Barry 0.976
Mike Miller 0.952
Kyle Korver 0.858
Anthony Parker 0.854

I checked and double-checked my spreadsheet. Is something missing?

Edit: How about integrating some sort of 3PA qualifier, something to the effect that a player has to shoot an x number of 3PA out of his total FGA to qualify (instead of using the minimum 100 3PA)?
Back to top
View user's profile Send private message
Charles



Joined: 16 May 2005
Posts: 134


PostPosted: Fri Sep 14, 2007 8:49 am Post subject: Reply with quote
Those numbers are "per 48 minutes", not "per game." Nice catch, supersub15.
Back to top
View user's profile Send private message
Charles



Joined: 16 May 2005
Posts: 134


PostPosted: Fri Sep 14, 2007 9:02 am Post subject: Reply with quote
basketballvalue wrote:
supersub15 wrote:

I've always wondered how media people pigeon-holed certain players just because they were proficient at shooting the 3. So, I wanted a more quantitative way of measuring this role.


It's semantics, but I think the original post captures what a specialist is. I would contrast Steve Kerr and Jason Kapono with Steve Nash. Charles' stats are valuable, but I don't think they capture specialization (associated with concentration in a particular area). I think Charles' stats capture talent (expertise). Particulary given supersub's original question, I don't think Steve Nash should be pigeon-holed into the Kerr/Kapono role because Charles shows he shoots the three almost comparably.

I think of the analogy in medicine, people aren't both superstar caridologists and orthopedic surgeons. They are specialists in one field or the other. As a player with multiple talents, Steve Nash is closer to a superstar primary care physicians, perhaps the surgeon general.

Similarly, I'd say that someone in the Walker category who concentrates on shooting the 3 on offense but doesn't do it as well is a specialist that is not as talented as others. Someone out there is the worst cardiologist in the US, but they're still a specialist.


Thanks,
Aaron

http://www.basketballvalue.com
http://www.82games.com/barzilai1.htm


I have to disagree with part of this, Aaron. If you specialize in one area and later obtain certification in a second area, you are then a specialist in both areas. The second specialization does not negate the first and return you to being a generalist. I certainly know doctors and academics who have more than one specialty.

Nevertheless, your point about semantics is well taken. Sports has its own parlance, so you can make an argument that "specialist" means whatever basketball fans think it means. But, dictionary definitions aside, is there any value in ranking by "Three point specialists who take relatively few two-point attempts"?

This summer, when Bryan Colangelo wanted to add an outside shooter to the Raptor's roster, he might have consulted a list of available free agents ranked by some combination of three-point efficiency and three-point scoring volume. (It makes sense to determine exactly how that list might best be calculated or presented.) But, why would Bryan want to lower a player's ranking simply because he was also a good passer or defender or two-point scorer? Frankly, a list like Kevin's seems arbitrary at best.

After all, is it reasonable for a list of "Three point specialists" to rank Diawara (who shot .288) second, while ignoring Kapono and Herrmann and Nash? Semantics aside, a method that rewards players for their limitations rather than their contributions just seems self-defeating.
Back to top
View user's profile Send private message
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 976
Location: Seattle

PostPosted: Fri Sep 14, 2007 12:13 pm Post subject: Reply with quote
Something that ought to be kept in mind in this discussion is that three-point specialists are not like other specialists (say, rebounding or shot-blocking) in that their primary statistic is much more variable than most.

For example, let's take the case of Fred Hoiberg. Hoiberg was generally considered one of the best shooters in the NBA, and during his two seasons in Minnesota he shot 44.2% and 48.3% from downtown. I think you'd find a pretty strong consensus that Hoiberg is a three-point specialist.

However, during his last two seasons in Chicago, Hoiberg shot just 26.1% and 23.8% on threes. Was he not a specialist then?

Charles, in response to your question I'll point out that there is value to descriptive statistics even if they don't relate to a player's value. Sometimes, as in projecting future development/performance, it may be as important to know how a player plays as how well a player plays.
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Charles



Joined: 16 May 2005
Posts: 134


PostPosted: Fri Sep 14, 2007 1:06 pm Post subject: Reply with quote
admin wrote:
Something that ought to be kept in mind in this discussion is that three-point specialists are not like other specialists (say, rebounding or shot-blocking) in that their primary statistic is much more variable than most.

For example, let's take the case of Fred Hoiberg. Hoiberg was generally considered one of the best shooters in the NBA, and during his two seasons in Minnesota he shot 44.2% and 48.3% from downtown. I think you'd find a pretty strong consensus that Hoiberg is a three-point specialist.

However, during his last two seasons in Chicago, Hoiberg shot just 26.1% and 23.8% on threes. Was he not a specialist then?

Charles, in response to your question I'll point out that there is value to descriptive statistics even if they don't relate to a player's value. Sometimes, as in projecting future development/performance, it may be as important to know how a player plays as how well a player plays.


Sure. I have no problem with categorizing players based on their limitations, rather than on their production and efficiency. The problem is compiling a list of "specialists" on that basis.

I call on a specialist when I need his or her special skills or knowledge. What other special skills the individual may or may not have is irrelevant. As long as they are a specialist in the matter at hand, that's what count.

I realize that sports lingo can be idiomatic, but there is no way I am counting on Diawara as a "three-point specialist" when he shoots .288.
Back to top
View user's profile Send private message
gabefarkas



Joined: 31 Dec 2004
Posts: 1311
Location: Durham, NC

PostPosted: Sat Sep 15, 2007 10:05 am Post subject: Reply with quote
thref23 wrote:
When putting together composite scores, I ranked players by 3 pt FG%, punished players who didn't take so many 3 point attempts (that way Elton Brand wasn't tops in the league), and weighted the results slightly on a per team basis (players on teams with a better offense were more likely to get open looks, so they were slightly punished).


It seems like you guys are bemoaning using something like 3PA/Poss because it ranks guys like Antoine Walker, who take and miss a lot of 3's, too highly. And understandably, he's not really a specialist if he doesn't do it well.

So, how about instead using 3PM/Poss? This rewards guys for both taking and making 3-point shots.
Back to top
View user's profile Send private message Send e-mail AIM Address
gabefarkas



Joined: 31 Dec 2004
Posts: 1311
Location: Durham, NC

PostPosted: Sat Sep 15, 2007 10:15 am Post subject: Reply with quote
Charles wrote:
This summer, when Bryan Colangelo wanted to add an outside shooter to the Raptor's roster, he might have consulted a list of available free agents ranked by some combination of three-point efficiency and three-point scoring volume. (It makes sense to determine exactly how that list might best be calculated or presented.) But, why would Bryan want to lower a player's ranking simply because he was also a good passer or defender or two-point scorer? Frankly, a list like Kevin's seems arbitrary at best.


I think there's a few reasons. The first one that comes to mind is that if a player has a few other specialties, they might be inclined to want to use them, along with using their 3-point shooting skills. For example, I would bet Steve Nash is somewhere near the top of the list that you mentioned. If Nash happened to be a free agent this summer, should he consider coming to the Raptors to be a 3-point specialist, and in the process accept a lower salary than he could otherwise command as a full-service PG who can also drain the 3?

My point is that any such list needs to be tempered with an assessment of the player's overall abilities and tendencies, in order to properly determine who would be the best fit for the team's available need.
Back to top
View user's profile Send private message Send e-mail AIM Address
Charles



Joined: 16 May 2005
Posts: 134


PostPosted: Mon Sep 17, 2007 12:03 am Post subject: Reply with quote
gabefarkas wrote:
Charles wrote:
This summer, when Bryan Colangelo wanted to add an outside shooter to the Raptor's roster, he might have consulted a list of available free agents ranked by some combination of three-point efficiency and three-point scoring volume. (It makes sense to determine exactly how that list might best be calculated or presented.) But, why would Bryan want to lower a player's ranking simply because he was also a good passer or defender or two-point scorer? Frankly, a list like Kevin's seems arbitrary at best.


I think there's a few reasons. The first one that comes to mind is that if a player has a few other specialties, they might be inclined to want to use them, along with using their 3-point shooting skills. For example, I would bet Steve Nash is somewhere near the top of the list that you mentioned. If Nash happened to be a free agent this summer, should he consider coming to the Raptors to be a 3-point specialist, and in the process accept a lower salary than he could otherwise command as a full-service PG who can also drain the 3?


Sure, you're right. Steve Nash would be foolish to use his ability to shoot threes as his primary factor in choosing a team to play for. But, the question is not whether Nash would consider going to the Raptors as a three-point specialist, but whether the Raptors would be well served having Nash as a three-point specialist.

Perhaps the issue would be clearer if we looked at a less extreme example.

The previous summer Colangelo was in a similar situation and picked up Anthony Parker partially for his ability to provide a three-point threat (Parker made 115 threes at a .441 clip.) However, the fact that Parker also penetrates well means that he does not rate particularly well on a method, such as the one Kevin presented, which counts two-point shots against you.

If Colangelo looked for Parker on Kevin's 3PA/FGA list , he would find him way down in 53rd spot, well behind many inferior 3P shooters. Is that list providing Colangelo with useful information? I think calling such a list "Three point specialists" would be misleading.

Quote:

My point is that any such list needs to be tempered with an assessment of the player's overall abilities and tendencies, in order to properly determine who would be the best fit for the team's available need.


I agree. In deciding how to use a player you would, of course, consider all the available information, not just his ability in one area. But, that doesn't justify a "3-point specialist" method that rates Diawara and Vujacic ahead of Kapono and Herrmann.

The fact that Diawara can not make two-pointers is just not sufficient reason to call him a three point specialist.

Do you really think it makes sense for Jason Kapono, who led the NBA in 3-point percentage and won the All-Star 3-point contest, to rank 63rd among three point specialists simply because he can also knock down a mid-range two?
Back to top
View user's profile Send private message
mtamada



Joined: 28 Jan 2005
Posts: 375


PostPosted: Mon Sep 17, 2007 4:50 am Post subject: Reply with quote
Charles wrote:
The fact that Diawara can not make two-pointers is just not sufficient reason to call him a three point specialist.


Depending on what we're doing with the terminology, it's on its way to being sufficient. If a player in fact IS good at 2-point shooting as well as 3-point shooting, then I would say that ipso facto we do NOT have a specialist, what we have is a GOOD SHOOTER period. Or you can call him a "deadeye", or "unconscious", or a "marskman" -- but he's not a specialist.

Now I admit that what that really does is tell us who is NOT a specialist, i.e. it gives us a necessary condition for being a specialist, not a sufficient one. But the general drift is: yeah, if you shoot 3-pointers better than you shoot 2-pointers, you're on you're way to meeting the definition of being a "specialist".

Quote:
Do you really think it makes sense for Jason Kapono, who led the NBA in 3-point percentage and won the All-Star 3-point contest, to rank 63rd among three point specialists simply because he can also knock down a mid-range two?


Again, it depends on what the term "specialist" is being used for. You seem to want it to mean "someone who's really good at shooting 3-pointers". I call those guys good shooters, long range marksmen, etc. -- but not specialists. The purpose of the ranking, at least as I see it, is not to show who's BEST at 3-point shooting, but who's most SPECIALIZED in it.
Back to top
View user's profile Send private message
Charles



Joined: 16 May 2005
Posts: 134


PostPosted: Mon Sep 17, 2007 9:46 am Post subject: Reply with quote
mtamada wrote:
Charles wrote:
The fact that Diawara can not make two-pointers is just not sufficient reason to call him a three point specialist.

Depending on what we're doing with the terminology, it's on its way to being sufficient. If a player in fact IS good at 2-point shooting as well as 3-point shooting, then I would say that ipso facto we do NOT have a specialist, what we have is a GOOD SHOOTER period. Or you can call him a "deadeye", or "unconscious", or a "marskman" -- but he's not a specialist.


The Oxford dictionary defines a specialist as "a person who is highly skilled or knowledgeable in a particular field." His ability or lack of ability in other areas is not considered. That is the definition I am accustomed to.

If I need a particular type of open heart surgery and my family doctor shows me a list indicating that Cardiologist A has performed this operation 250 times with a 98.5% success rate and Cardiologist B has performed the surgery 180 times with a 94.0% success rate, I am not going to care whether this procedure makes up a larger percentage of Cardiologist B's case load. I already have the information I need.

It is a question of semantics:

The traditional definition of a specialist demands that: 1) the individual demonstrate that he is "highly skilled" - Diawara is off the list, but that 2) an individual may have more than one specialty -- Kapono, Parker and Nash are on the list.

On the other hand, you could adopt an idiomatic sports media definition which says that 1) if a player only does one thing he is a specialist -- regardless of demonstrated skill, but that 2) a specialist may not have balancing skills.

I prefer the former definition because, in my opinion, a "three-point specialist" formula that ranks Diawara (49 three-pointers at .288) a page and a half ahead of Jason Kapono (108 three-pointers at .514) is dangerously close to being numbers for the sake of numbers -- at best, of limited utility and, at worst, downright misleading. But, hey, knock yourself out.
Back to top
View user's profile Send private message
thref23



Joined: 13 Aug 2007
Posts: 90


PostPosted: Mon Sep 17, 2007 1:38 pm Post subject: Reply with quote
Charles wrote:


If I need a particular type of open heart surgery and my family doctor shows me a list indicating that Cardiologist A has performed this operation 250 times with a 98.5% success rate and Cardiologist B has performed the surgery 180 times with a 94.0% success rate, I am not going to care whether this procedure makes up a larger percentage of Cardiologist B's case load. I already have the information I need.


What about Cardiologist C? He is more than willing to conduct open heart surgery for you at a lower cost, and appears to have a decent success rate. But he's not a licensed practioner and when you dig deep his track record is less than reassuring.

Does that make him a bad specialist? Or does it make him a con artist? I say he's a con artist, and not a specialist.

Also, if Cardiologist A is a do it all and this type of operation makes up a small percentage of his professional responsibilities, chances are either you are not going to be able to book an appointment with him until its too late, and/or you may have to pay an exorbitant fee for his services, because you will also have to pay for expertise in other areas which do not matter to you.

In other words, Cardiologist A can be a specialist when/if he wants to as he has the required skill set, but he arguably does not act as a specialist.

I could listen to an argument that Shane Battier is both a defensive specialist and a three point specialist - kind of like a dual major. Perhaps I could say the same perhaps about Anthony Parker, Sasha Vujacic, or Bruce Bowen. I have trouble considering Steve Nash to be a 3 pt specialist.

And I'm not necessarily right or wrong IMO, its just how I view the subject.
Back to top
View user's profile Send private message Send e-mail
basketballvalue



Joined: 07 Mar 2006
Posts: 204


PostPosted: Mon Sep 17, 2007 2:06 pm Post subject: Reply with quote
Charles,

I agree with most of your points. I should probably quote them to emphasize that fact, but my apologies for brevity here. In particular, I think your comments on Diawara make a lot of sense, and agree that there are some MDs who have multiple specialties and that does not make them a PCP.

However, I thought I'd highlight a couple of points where I disagree. After all, that's often more interesting.

Charles wrote:

If I need a particular type of open heart surgery and my family doctor shows me a list indicating that Cardiologist A has performed this operation 250 times with a 98.5% success rate and Cardiologist B has performed the surgery 180 times with a 94.0% success rate, I am not going to care whether this procedure makes up a larger percentage of Cardiologist B's case load. I already have the information I need.


I actually think you might. What if I told you that Cardiologist A used to perform this procedure regularly but hasn't in the last five years as he switched to focusing on stents? Meanwhile, all of Cardiologist B's procedures have been performed in the last two years, when open heart surgery has been reserved for sicker patients.

We're probably getting off-topic, but thought I'd mention that. Perhaps this falls in the category of "just-one-number-isn't-meaningful" that we're hearing on the PER threads.

Charles wrote:

Nevertheless, your point about semantics is well taken. Sports has its own parlance, so you can make an argument that "specialist" means whatever basketball fans think it means. But, dictionary definitions aside, is there any value in ranking by "Three point specialists who take relatively few two-point attempts"?


This could be valuable to front offices as they try to fill their rosters as efficiently as possible. There's no point in a team paying for Kapono's 2 point shooting skills (or Nash's passing per another post) if you know that his role is only going to let him shoot from 3 point range, right?

I think an interesting question that gets back to supersub's original question is "Given that Jason Kapono demonstrated great skill at three-point shooting this season (e.g. Charles' stats) but also can shoot two-pointers effectively, is he unfairly pigeonholed as a player with a single talent, three-point shooting?" The answer might be "he has a single talent, shooting, and is not unfairly pigeonholed as a shooter of jump shots and free throws".

Thanks,
Aaron

http://www.basketballvalue.com
http://www.82games.com/barzilai1.htm
Back to top
View user's profile Send private message
tmansback



Joined: 12 Aug 2005
Posts: 129


PostPosted: Mon Sep 17, 2007 4:22 pm Post subject: Reply with quote
Every time I've heard the word specialist used in basketball it has been to refer to a player that sole purpose in the game is to provide a certain skill. When u put it in context with the five players on the floor sometimes u have certain players that sole reason for playing time is shooting. I would not consider players with multi-dimensions in there game as specialist. Specialist is not really a positive term. Specialist are paid for 1 skill.

In the draft its often said u don't take specialist till the second round. A player like Paul Milsap was considered a rebounding specialist. Bobby Jones was a defensive specialist. Steve Novak was a 3 point specialist. I criticized the pick of JJ Redick because why would u take a 3 point specialist so high in the draft.

I've always felt that 1 great skill your a specialist. 2 great skills your in the 8 man rotation and potential starter. 3 great skills your a potential all-star.
Back to top
View user's profile Send private message
mtamada



Joined: 28 Jan 2005
Posts: 375


PostPosted: Mon Sep 17, 2007 6:20 pm Post subject: Reply with quote
Charles wrote:
The Oxford dictionary defines a specialist as "a person who is highly skilled or knowledgeable in a particular field." His ability or lack of ability in other areas is not considered. That is the definition I am accustomed to.


Definitely not the definition that I (and evidently many others in this thread) are accustomed to. The Oxford dictionary can be idiosyncratic, possibly because it attempts to cover the entire English language, rather than the "American" language.

Here's a sampling from other dictionaries, these are the first four listed in dictionary.com:

Random House Unabridged Dictionary: 1. a person who devotes himself or herself to one subject or to one particular branch of a subject or pursuit. (Italics added)

American Heritage Dictionary: 1. One who is devoted to a particular occupation or branch of study or research:

WordNet® 3.0: 1. an expert who is devoted to one occupation or branch of learning

Kernerman English Multilingual Dictionary (Beta Version): a person who makes a very deep study of one branch of a subject or field


Those are the definitions (or really definition, they all pretty much agree with each other) that I am accustomed to. I literally have never heard anyone use the usage that your cite from Oxford, except of course for the connotation that a specialist will indeed by "highly skilled or knowledgable". There's a definite connotation there -- note however that it is not part of the other dictionaries' denotation. One can be "devoted" or make a "deep study" without being very good.

As for the Oxford definition that you cite, there's only one word that jumps to my mind with that definition, and the word is not "specialist".

The word is "expert".
Back to top
View user's profile Send private message
mtamada



Joined: 28 Jan 2005
Posts: 375


PostPosted: Mon Sep 17, 2007 10:21 pm Post subject: Reply with quote
I should add though that much as AaronB says, there is a potential area of agreement between all of us: we could define, for the purposes of studying 3-point shooters, a "specialist" as someone who is both better at 3-pointers than he is at 2-pointers (i.e. utilizing the concept that most of us are advocating) AND who is also GOOD at shooting 3-pointers (utilizing the concept that Charles is advocating, and that I think we all recognize is connoted by the word "specialist").

That does raise the issue however of which attribute to put more emphasis on: the degree of specialization, or the degree of "specialness".

That BTW may be at the crux of this semantic disagreement: most of us define a "specialist" as some who has "specialized", i.e. become "specific". Charles seems to define a "specialist" as someone who is "special", i.e. especially good.

supersub15



Joined: 21 Sep 2006
Posts: 273


PostPosted: Tue Sep 18, 2007 5:57 am Post subject: Reply with quote
How about this method: 3PM/(FGM-3PM)

The problem is deciding the cutoff for considering a player a 3-pt specialist:

(minimum 100 3PA):
Damon Jones 1.854
Brent Barry 1.730
Luther Head 1.372
Brian Scalabrine 1.333
Sasha Vujacic 1.333
Roger Mason 1.269
James Jones 1.250
Robert Horry 1.250
Shane Battier 1.163
Keith Bogans 1.152
Daniel Gibson 1.130
J.R. Smith 1.104
James Posey 1.090
Rafer Alston 1.005
Raja Bell 1.000
Bostjan Nachbar 0.982
Yakhouba Diawara 0.961
Allan Ray 0.923
Morris Peterson 0.914
Bruce Bowen 0.890
Back to top
View user's profile Send private message
Statman



Joined: 20 Feb 2005
Posts: 242
Location: Arlington, Texas

PostPosted: Tue Sep 18, 2007 11:04 am Post subject: Reply with quote
supersub15 wrote:
How about this method: 3PM/(FGM-3PM)

The problem is deciding the cutoff for considering a player a 3-pt specialist:

(minimum 100 3PA):
Damon Jones 1.854
Brent Barry 1.730
Luther Head 1.372
Brian Scalabrine 1.333
Sasha Vujacic 1.333
Roger Mason 1.269
James Jones 1.250
Robert Horry 1.250
Shane Battier 1.163
Keith Bogans 1.152
Daniel Gibson 1.130
J.R. Smith 1.104
James Posey 1.090
Rafer Alston 1.005
Raja Bell 1.000
Bostjan Nachbar 0.982
Yakhouba Diawara 0.961
Allan Ray 0.923
Morris Peterson 0.914
Bruce Bowen 0.890


This is what I threw together for the heck of it:

(3pt/3ptA)*(3pt/min*48 )*(3pt*3/pts)

So, it takes into account how good they are at hitting them (3pt%) plus how often they hit them (makes per 48 ) plus how big a percentage of points they score are solely from threes. I multiply all three together, since I figure all are probably pretty vital for one to be considered a "specialist".

Here's my list of players with over 424 minutes (so Peja makes it as reference):

Code:
1 J.R. Smith 1.04
2 Luther Head 1.03
3 Brent Barry 1.02
4 Damon Jones 0.95
5 Eddie House 0.78
6 Raja Bell 0.75
7 Bostjan Nachbar 0.72
8 Peja Stojakovic 0.69
9 Mike Miller 0.67
10 Jason Kapono 0.67
11 Allan Ray 0.67
12 James Jones 0.66
13 Sasha Vujacic 0.66
14 Roger Mason 0.62
15 Morris Peterson 0.61
16 Shane Battier 0.60
17 Daniel Gibson 0.60
18 Leandro Barbosa 0.60
19 Tim Thomas 0.59
20 Donyell Marshall 0.59
21 Rafer Alston 0.58


League average is 0.08
_________________
Dan

My current national college player rankings (and other stuff):
http://www.pointguardu.com/f136/statman ... post355594
Back to top
View user's profile Send private message Send e-mail
Mountain



Joined: 13 Mar 2007
Posts: 1527


PostPosted: Tue Sep 18, 2007 12:11 pm Post subject: Reply with quote
There are several other possible elements to this discussion: competency and high enough frequency from multiple areas of the 3 pt arc and more importantly frequency and accuracy of 3 pt shooting in crunch and clutch time- a big part of the reason for putting a 3 pt specialist on the court and distinguishing him from others who shoot the shot to a lesser extent, especially the open 3.


Top
 Profile  
 
PostPosted: Fri Apr 22, 2011 2:05 pm 
Offline

Joined: Thu Apr 14, 2011 11:10 pm
Posts: 2441
partial thread

supersub15



Joined: 21 Sep 2006
Posts: 273


PostPosted: Fri Sep 07, 2007 8:46 am Post subject: Winning a championship and having an All-NBA player Reply with quote
I wanted to see whether there was a correlation between winning an NBA championship and having players voted All-NBA First Team. I started with the 3-point era, i.e. from 1979 until now and found the following: In the 29 championships, only 6 teams did not have a player named to the All-NBA First Team, i.e. 80% of the teams have won when they had one of the top 5 players in the NBA. These 6 teams are:

2003-04 Detroit Pistons
1994-95: Houston Rockets
1989-90: Detroit Pistons
1988-89: Detroit Pistons
1981-82: Los Angeles Lakers
1978-79: Seattle Supersonics

Of those 6 teams, only the 1988-89 Pistons and the 1978-79 Sonics didn't have a player in one of the 3 all-NBA teams (mind you, the third all-NBA team was introduced for the 1988-89 season and the Sonics had Gus Williams, Jack Sickma, and Dennis Johnson, 3 legitimate candidates to the third All-NBA team, had it existed at the time).

Consequently, the rate of winning a championship goes up to 93% if you have a player among the top 15 players in the league.

I then went all the way back to the BAA and the start of the NBA, 1946-1947 to be exact, and tried to make the same correlation. In 61 championships, the winning team had at least one player in the All-NBA First Team 45 times (74%). However, if we account for teams having at least one player in the 2 or 3 All-NBA Teams, only 3 teams didn't have a player in the top 10 or 15, the aforementioned Sonics and Pistons teams and the 1978-1979 Washington Bullets team. That's 95% of the time.

Just something that intrigued me. Thoughts?
Back to top
View user's profile Send private message
basketballvalue



Joined: 07 Mar 2006
Posts: 204


PostPosted: Fri Sep 07, 2007 12:18 pm Post subject: Re: Winning a championship and having an All-NBA player Reply with quote
supersub15 wrote:


Consequently, the rate of winning a championship goes up to 93% if you have a player among the top 15 players in the league.



Supersub,

You mean that 93% of championship teams have a player among the top 15, right?

If it's not too hard to do, I'd be curious to know what the rate is of winning a championship given that you have at least one first-team All-NBA player (or at least one first through third team) on your roster. Might be interested to separate out the probabilities from one vs. two.

Thanks,
Aaron


http://www.82games.com/barzilai1.htm
http://www.basketballvalue.com
Back to top
View user's profile Send private message
supersub15



Joined: 21 Sep 2006
Posts: 273


PostPosted: Fri Sep 07, 2007 12:37 pm Post subject: Re: Winning a championship and having an All-NBA player Reply with quote
basketballvalue wrote:
supersub15 wrote:


Consequently, the rate of winning a championship goes up to 93% if you have a player among the top 15 players in the league.



Supersub,

You mean that 93% of championship teams have a player among the top 15, right?

If it's not too hard to do, I'd be curious to know what the rate is of winning a championship given that you have at least one first-team All-NBA player (or at least one first through third team) on your roster. Might be interested to separate out the probabilities from one vs. two.

Thanks,
Aaron


http://www.82games.com/barzilai1.htm
http://www.basketballvalue.com


You're right about the sentence. Sorry for the confusion.

Of the 61 championship teams, 35 had multiple All-NBAers (first, second, and third teams), 31 teams had only All-NBAer (first, second, and third teams), and 3 had none.

Did I understand your question correctly? Or do you want the breakdown differently?
Back to top
View user's profile Send private message
Mountain



Joined: 13 Mar 2007
Posts: 1527


PostPosted: Fri Sep 07, 2007 1:28 pm Post subject: Reply with quote
Checking over last 10 years I see 4 teams with 2 first team all-NBA players and 2 of the 4 won championship or 50% chance. Too small a sample of course but maybe supersub will want to do it for a longer span. Phoenix missed this past season and 03-04 Lakers did also.

Chance of winning with 1 all-NBA player isnt far from 1 in 5.

Chance with one player on any of the 3 all-NBA teams is only about 1 in 12-13, at least from looking at last few years.

This was a strong article by Dennis Gallagher on the topic from 18 months ago. A similar article or post based on who teams have today and what their key players have achieved in top ranks would be a good read.

http://www.82games.com/dennis.htm

Of the 7 one-Star champions all were top 6 on defensive efficiency. The average on offense was about 8.5 but 3 were near or actually below league average. If you are one-star you better be top level on defense. Teams see the advantages of 2 stars but a number of them I think are mistaken in thinking their second star is good enough (not that commmon by these standards) especially if both on offensive stars. One star and a top defense seems more feasible and cheaper than 2 stars. It may give you a legit shot but still only a modest chance with probably at least 6 other legit contenders a year?

I think 7 teams (Spurs, Suns, Rockets, Celtics, Cavs, Nuggets, Heat and Pistons) have a tandem that meets these qualifications. Chicago and Mavs join the contenders for having one star and a top 6 defense. Jazz is very close Boozer almost top 8 on PER and Kirilenko was a top defender but with his game sketchy I don't feel like giving them the slight break. I don't think Nets or anyone else makes this top player historical based cut but I'll add one spot for any dark horse and you have essentially a 10 horse race. Don't need a study to make that kind of statement but that's where I end up with a quick updating of the study.

To win a championship takes some combination of top players and top team efficiency on offense and defense or differential. It might be illuminating to construct a study that boiled in that down to:

Yes or no?
1. Top player (by Gallagher standards or otherwise)?
2. Qualifying second star?
3. A qualifying sidekick? (perhaps with a bit looser definition)
4. Top ten on offensive efficiency?
5. Top ten on defensive efficiency?
6. Top ten on point differential?

How many of last x champions had 4, 5, 6 of these?
How many others did each year? Were they any champions with less than 3 yeses?

Looks like 07 Spurs were first team with 6 since Bulls dynasty. All other champions in that timespan have had 5.

Based on performance last season I think (moving quickly) the teams score this way: Spurs 6, Suns 4 (possibly 5 if sidekick criteria is stretched), Rockets 4 (5 if Mutumbo still counts or Battier), Celtics n/a, Cavs 4, Nuggets 6, Heat 3 Chicago 4 and Mavs 4 (possibly 5 if sidekick criteria is stretched), Detroit 5.

Spurs a leader in championship construction, no surprise. Denver might be a surprise but they are top 10 on all 3 team measures and have Iverson, Anthony and Camby. The 6 criteria system is still first cut and you have to have the right stars. Denver with more time together to develop chemistry will get at least one more chance.

I'd add a 7th criteria on principle- a top 10 coach but I don't think it would change anything looking at last season, except maybe with Mike Brown of Cavs but too early to say on him and hard to say independent of team results.

Moving from last season numbers to next season, who improved? Celtics of course, probably become a 5. Chicago might become a 5 if Deng advances up top player list. Maybe Mavs or Cavs do that too- maybe. That about all I see in terms of major changes. Spurs best positioned. I don't think there is a clear #2 or 3. Pretty wide open. If a solid by the book 5 is needed next year then the list of top contenders might end up at 6-7 instead of near 10.

(If you counted a top flight 1st or 2nd team awarded 2 way player as 2 stars instead of just one some of these team scores would go up throughout the prior discussion.)

Last edited by Mountain on Fri Sep 07, 2007 6:21 pm; edited 18 times in total
Back to top
View user's profile Send private message
KeeneKaufmanWheeler



Joined: 02 Aug 2006
Posts: 72


PostPosted: Fri Sep 07, 2007 2:20 pm Post subject: Reply with quote
Isn't this a study that might lose some validity as the composition of the league has changed (through expansion and the evolution of the salary cap)?

20 years ago there were fewer teams and therefore a higher percentage of the league was all-NBA players.
Back to top
View user's profile Send private message Visit poster's website
Flint



Joined: 25 Mar 2007
Posts: 112


PostPosted: Fri Sep 07, 2007 3:33 pm Post subject: Reply with quote
A few thoughts. That Supersonics team and the Rockets, were the third and fourth worst teams by differential to win an NBA title. Only the Celtics and Bullets in the 1970's were worse.

So you have two teams that simply weren't very good.

The 81-82 Lakers team had Magic Johnson. His season that year was possibly the best season ever in the three point era. His not being first team all - nba was as close to cosmic injustice as you are likely to experience in the field of basketball statistics. Gus Williams!?!

http://www.basketball-reference.com/fc/ ... 02&y3=1982

That leaves the three Pistons teams. They had Ben Wallace and Dennis Rodman on them, players unlikely to ever receive consideration from the the First Team All NBA committee. I don't know that those players were top five players in those years, but if not they were very very close.

I think what this leaves you with is the observation that you very rarely win a championship without an actual top five player, and that the All NBA team doesn't always feature the top five players.

Lol, ok, Captain Obvious, signing off
Back to top
View user's profile Send private message
Mike G



Joined: 14 Jan 2005
Posts: 3496
Location: Hendersonville, NC

PostPosted: Fri Sep 07, 2007 3:53 pm Post subject: Reply with quote
Flint,

Why don't you just admit that you are Dave Berri?

1982 was far from Magic Johnson's best year. It was just his best rebounding year.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
asimpkins



Joined: 30 Apr 2006
Posts: 244
Location: Pleasanton, CA

PostPosted: Fri Sep 07, 2007 4:18 pm Post subject: Reply with quote
Great illustration of the differences in thought here. If we throw out the two seasons he was injured and played less than half the season, his 81-82 season ranks 10th out of 11 according to PER. This is because that year he was:

10th out of 11 in Usage Rate. 10th out of 11 in TS%. 10th out of 11 in Assist Rate. 7th out of 11 in Turnover Rate.

But yes, 1st out of 11 in Rebounding Rate.
Back to top
View user's profile Send private message
bchaikin



Joined: 27 Jan 2005
Posts: 678
Location: cleveland, ohio

PostPosted: Fri Sep 07, 2007 4:44 pm Post subject: Reply with quote
The 81-82 Lakers team had Magic Johnson. His season that year was possibly the best season ever in the three point era. His not being first team all - nba was as close to cosmic injustice as you are likely to experience in the field of basketball statistics. Gus Williams!?!

fwiw simulation shows magic johnson's 81-82 season as the 2nd best in terms of wins generated by a PG - on a 40 min/g and 82 game basis - over the past 30 years (since 77-78 and when turnovers were first tracked), bettered by only his 80-81 season...

however gus williams in 81-82 was also very good, generating wins at a rate of just about 4-5 less wins than magic's 81-82 season on a 40 min/g and 82 game basis...

Last edited by bchaikin on Fri Sep 07, 2007 4:44 pm; edited 1 time in total
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Flint



Joined: 25 Mar 2007
Posts: 112


PostPosted: Fri Sep 07, 2007 4:44 pm Post subject: Reply with quote
Well, thanks Mike G, thats something of a compliment from my view, but I am definitely not David Berri. Just a fan of his and the Knicks. Do you honestly think I would be asking how to calculate team TS% if I were an economics professor?

Also, even if that wasn't Magic's best season, for the purposes of this discussion, I think it's pretty clear he was a better player than both of the guards voted first team all nba ahead of him. Or did Gus Williams have a higher PER than Magic that year?
Back to top
View user's profile Send private message
Flint



Joined: 25 Mar 2007
Posts: 112


PostPosted: Fri Sep 07, 2007 4:45 pm Post subject: Reply with quote
Bchaikin - Are you Dave Berri? C'mon, fess up. We know it's you.

Actually I made a mistake. That was not Magic's best season according to Berri. That would be 88-89 where he had a WP48 of .538 and produced 32.4 wins

http://www.wagesofwins.com/MagicCareer.html

But what you wrote is very interesting though. Thanks for vaguely agreeing with me. That would possibly be a first on this site. What do your numbers show for Gervin actually? And for that matter what do they show for Rodman and Wallace?
Back to top
View user's profile Send private message
bchaikin



Joined: 27 Jan 2005
Posts: 678
Location: cleveland, ohio

PostPosted: Fri Sep 07, 2007 9:37 pm Post subject: Reply with quote
Are you Dave Berri? C'mon, fess up. We know it's you...

ya' got me...

actually i clicked on your link and he has a stats called PAWS. i've got cats for pets so anyone creating a stats called PAWS is ok by me...

Actually I made a mistake. That was not Magic's best season according to Berri. That would be 88-89 where he had a WP48 of .538 and produced 32.4 wins...

let's compare magic's 88-89 and 80-81 stats:

year------min---pts/g---ScFG%---reb---ast----st----to---bs----g
8081---1371---21.6----.570-----320---317--127--143--27---37
8889---2886---22.5----.610-----607---988--138--312--22---77

to compare apples to apples let's normalize to minutes played (multiply his 80-81 stats by 2886/1371):

year------min---pts/g---ScFG%---reb---ast----st----to---bs----g
8081---2886---21.6----.570-----674---667--267--301--57---77
8889---2886---22.5----.610-----607---988--138--312--22---77

and if you want to be a bit more technical, adjust for pace (80-81 lakers got 103.9 poss/g, the 88-89 lakers 101.6 poss/g, so multiply his 80-81 stats by 101.6/103.9):

year------min---pts/g---ScFG%---reb---ast----st----to---bs----g
8081---2886---21.1----.570-----659---652--261--294--56---77
8889---2886---22.5----.610-----607---988--138--312--22---77

what i see here is a player who got 261 - 138 = 123 more steals, 312 - 294 = 18 less turnovers, and 34 more blocks in 80-81 vs 88-89. estimating say 55% of all blocked shots are retrieved by the defense as defensive rebounds, that'd be 123 + 18 + 34x0.55 = 123 + 18 + 19 = 160 less zero point possessions for magic in 80-81 vs 88-89 (not to mention he got more rebounds in 80-81)...

so the question is for magic's 88-89 season to generate more wins than his 80-81 season, how does he make up for this difference of 160 less zero point team possessions in 77 games or 2.1 zero point possessions per game. after all a steal is the vast majority of the time the end of your opponent's possession, and thus they do not score, and a turnover for your team most often results on a team possession where your team did not score...

to make up for 160 zero point team possessions - not to mention 52 less rebounds - he would have to have accounted for, say, more defensive stops outside of steals, such as by forcing more misses or forcing more turnovers that are not steals. a forced miss is not a defensive stop unless your team gets the defensive rebound, so approximately 160/0.66 = 242 more forced misses would have to have been forced by magic in 88-89 vs 80-81 (with no more forced turnovers), or some combination of less forced misses and more forced turnovers (the 80-81 lakers got 64% def rebs and the 88-89 lakers got 69% def rebs so i used 66% as an average)...

you could point out that magic shot better overall in 88-89, but just because magic didn't score on a certain team possession (and isn't personally responsible for a team zero point possession) doesn't mean his team didn't score on that possession. plus you could point out that magic in 88-89 got 988 - 652 = 336 more assists than in 80-81. both surely make up for some of those 160 zero point possessions and 52 less rebounds. the question is how much?...

however a PG can have significantly less assists and even shoot worse than another PG but still generate as many if not more wins through his other stats. this was pointed out in another thread:

In Brandon's career year of '96, he was apparently as good as Stockton was in any year.

some will point out that magic's better shooting and more assists in 88-89 more than make up for these zero point possessions, but simulation shows that on a 40 min/g and 82 game basis, magic's 80-81 season generates slightly more wins than his 88-89 season, 2 to 3 more wins depending which team you run the simulation on. both seasons however are among the very best in generating wins by a PG in the last 30 years...

But what you wrote is very interesting though. Thanks for vaguely agreeing with me.

you are vaguely welcome...

What do your numbers show for Gervin actually?

that same 81-82 season george gervin scored 32 pts/g for the spurs. simulation shows that - on a 40 min/g and 82 game basis - he generated wins at the rate of about 12-13 more than some of the worst win generating SGs in the league that year (like chris ford of the celtics, phil smith of the clippers/sonics, and jim brogan of the clippers), and that year i have gervin rated as a poor defender outside of steals, blocks, and defensive rebounding. but that season simulation also shows sidney moncrief as generating 4-5 more wins than gervin did on a 40 min/g and 82 game basis, and about 16-17 more wins than the worst win generating SGs. that year i have moncrief rated as an excellent defender outside of steals, blocks, and defensive rebounding...

And for that matter what do they show for Rodman and Wallace?

ben wallace's best year statistically was probably his 01-02 season (his first DPOY award). simulation shows - on a 40 min/g and 82 game basis - wallace generating as many wins as shaquille o'neal did that year (when shaq scored 27 pts/g and wallace managed less than 8 pts/g). that season wallace got 13 reb/g, 1.7 st/g, and 3.5 bs/g, but while committing less than 1 to/g and shooting a ScFG% right at the league average...

possibly rodman's best year statistically was 91-92. that year he generated - on a 40 min/g and 82 game basis - as many wins as both karl malone and horace grant...
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Flint



Joined: 25 Mar 2007
Posts: 112


PostPosted: Fri Sep 07, 2007 11:51 pm Post subject: Reply with quote
Thanks for the response. And I did appreciate your response.

I am not a WOW expert. And I don't know exactly how he calculates wins produced. But I understand his logic I think.

For simplicity sake, here are the numbers I am looking at. Per 48.

http://www.basketball-reference.com/fc/ ... 02&y3=1989

In Berri's book, a block, assist, ft or personal foul is worth .5 a rebound, point, steal, or shot attempt.

Sometimes I mess up my numbers, I don't have a spreadsheet set up like everyone has here. Have to figure out how to do that.

In non-scoring stats, Young Magic has an advantage of

+ 1.1 rebounds
+ 2.1 steals
+.25 blocks
+.2 turnovers
- .3 pfs
- 2.65 assists

+ .7


On the scoring end, Young Magic scores 27.9 points but he takes 20.6 shots and 7.9 fta's. So dividing the ft's by 2, this works out to a deduction of 24.55. His net from his points is 27.9 - 24.55, or 3.35

On the Scoring End Old Magic scores 28.8 points on 23.6 shots, for a net of 5.2

So Old Magic has a scoring advantage of 1.85

Overall, the 88-89 version has a 1.15 advantage, per 48.

Now, that is just how it works out with win score. REally, thats some gunky analyis. As I said, I don't know how Wins Produced works exactly. From the end notes of his book it seems like a fairly involved process. But I think the basic logic I employed above probably approximates his reasoning for why the later Magic was better. I think Wins Produced is adjusted for pace, and there is a team defensive adjustment in there as well, so that might make a difference, but I think that's the gist of it.

A much shorter way of saying it would have been that a 4% difference in TS% is a big difference in shooting efficiency. And 5 more assists per 48 is a big difference. Those things more than make up for Young Magic's non-scoring stat superiority.

I had no idea Gervin or Williams were that good. Both were before my time. I thought Gervin was Iverson-esque, but clearly I was wrong. But, it still seems like Magic was the better player than them in 81-82. So, I feel my original point stands, basically, that sort of is one less team that won without a 1st team All-NBA player.

Re Wallace -

I don't know what the exact numbers for Wallace are, but I would think he was a top five player at least one of those years 2000-5. Your model sees to be a lot more in line with Berri's conclusions than others I have read about here. I am not familiar with what you do. Do you have a link to a thread explaining it?

And re Rodman

I don't know what his best year was. Berri says in the book he was the best power forward in the league in 93-94, when, incidentally, he averaged 8.4 offensive rebounds per 48. He seems to have been pretty good in 95-96 on the best team of all time also. I think Berri rates him at a .405 that year, although he didn't play 82 games that year, and that's just off the top of my head.

But who knows, Berri probably doesn't have it right. According to his PER of 13.6 he was a below average player and the seventh best player on that team, despite averaging 18 rebounds per 40.

Lol, shouldn't say stuff like that around here Smile Thanks for the post, good night.
Back to top
View user's profile Send private message
gabefarkas



Joined: 31 Dec 2004
Posts: 1311
Location: Durham, NC

PostPosted: Sun Sep 09, 2007 3:56 pm Post subject: Reply with quote
KeeneKaufmanWheeler wrote:
Isn't this a study that might lose some validity as the composition of the league has changed (through expansion and the evolution of the salary cap)?

20 years ago there were fewer teams and therefore a higher percentage of the league was all-NBA players.


There were also less all-NBA teams back then, too.
Back to top
View user's profile Send private message Send e-mail AIM Address
HoopStudies



Joined: 30 Dec 2004
Posts: 705
Location: Near Philadelphia, PA

PostPosted: Mon Sep 10, 2007 11:08 am Post subject: Reply with quote
Flint wrote:

And re Rodman

I don't know what his best year was. Berri says in the book he was the best power forward in the league in 93-94, when, incidentally, he averaged 8.4 offensive rebounds per 48. He seems to have been pretty good in 95-96 on the best team of all time also. I think Berri rates him at a .405 that year, although he didn't play 82 games that year, and that's just off the top of my head.

But who knows, Berri probably doesn't have it right. According to his PER of 13.6 he was a below average player and the seventh best player on that team, despite averaging 18 rebounds per 40.


I never knew PER devalued Rodman so much. His career PER is below average. He has to be one of the more statistically uncertain players -- if you look at different stat methods, his value would range a lot.

Justin gives him a 36% chance at the HOF. I can't argue much with that because I really don't know how posterity will view him. He definitely has some chance by being so good at a couple things.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.

Flint



Joined: 25 Mar 2007
Posts: 112


PostPosted: Mon Sep 10, 2007 12:57 pm Post subject: Reply with quote
Hi Dean - I love BOP, great book.

PER is a pretty bizarre statistic. I really don't understand it. Rodman had a 17 PER in 91-92, which according to Bchaikin was his best season overall. As a Knicks fan, I happen to know that Eddy Curry also posted a 17 this year, while finishing second in the league in turnovers.

I look at this comp, and it makes my head spin.

http://www.basketball-reference.com/fc/ ... 01&y2=2007

How does that work? Rodman averaged 22 rebounds per 48 that year and committed only 2 turnovers for a +20 in gaining and maintaining possession.. Curry was at +4.6. So the 15 points Curry outscored him by is less than the difference between them in their impact on possession.

How could they possibly have the same PER? I just don't get it.
Back to top
View user's profile Send private message
asimpkins



Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA

PostPosted: Mon Sep 10, 2007 2:30 pm Post subject: Reply with quote
Flint wrote:
How does that work? Rodman averaged 22 rebounds per 48 that year and committed only 2 turnovers for a +20 in gaining and maintaining possession.. Curry was at +4.6. So the 15 points Curry outscored him by is less than the difference between them in their impact on possession.


1. You suggest that Rodman was gaining +20 possessions per 48. At face value, this is as good as you claim. The reason PER doesn't rate Rodman off the charts, however, is that it doesn't consider securing a Defensive Rebound as equivalent to gaining a possession -- like you suggested. Yes, it is typically the final act in gaining the possession, but before it can happen a lot of defense is played. The entire team contributes in an effort to deny easy shots and ultimately put up a low percentage shot that will miss. And even after that everyone attempts to box out their man so that someone on the defensive team can get the rebound.

By crediting Rodman (or any player) with the full value of a possession for securing a defensive rebound you are basically rating him as single-handedly stopping an entire team. If Rodman really was one man defensive force on 16 or so plays per 48 minutes, then he would deserve the kind of credit you want to give him. I think that it is obvious that he only played a part.

2. The other difference in assumptions that PER makes is that Usage Rating matters. You point out that he only turned the ball over 2 times per 48 minutes. But that's only because his teammates didn't dare ever throw him the ball. In many of his great rebounding years his turnover rate was over 20%. His Usage Rate was down around 10% -- half of his share. When he put up a rare shot his efficiency was often terrible. Basically, any team with Rodman on the court had to play 4 on 5 on offense. PER thinks that matters, and it penalizes him for it.


My understanding is that WoW thinks otherwise. Grabbing a defensive rebound is the same as single-handedly stopping the opposing team on offense. And it doesn't matter if you are almost completely useless on offense, as long as you stay out of the way. By that criteria, then yes, Rodman would be one of the best.
Back to top
View user's profile Send private message
Mountain



Joined: 13 Mar 2007
Posts: 1527


PostPosted: Mon Sep 10, 2007 2:39 pm Post subject: Reply with quote
The 6 point star & team ranking exercise I posted earlier in thread is just a rough guide on contender status. A higher number doesnt ensure victory of course. But all champs in last 10 years were at least a 5 and all finalists except the Indiana Pacers were at least a 4. The main advantage of this system it is shows stars alone don't determine but also team strength without enough star power backbone might not be enough in the playoffs on its own either. It takes an adequate combination. There may be more wiggle room about what is enough than in this presentation but the wiggle room probably isnt that large. The Pacers did better than their score because they had 4 Factor hammers, 1 major (own FG%) and at least 1 minor (FG% allowed) as described in team factor wins thread. The third team criteria of point differential does a good job of indicating when an offensive or defensive efficiency weakness is being adequate offset by the other strength.
Back to top
View user's profile Send private message
bchaikin



Joined: 27 Jan 2005
Posts: 685
Location: cleveland, ohio

PostPosted: Mon Sep 10, 2007 8:55 pm Post subject: Reply with quote
A much shorter way of saying it would have been that a 4% difference in TS% is a big difference in shooting efficiency.

for a simple calculation, magic shot 50.9% in 88-89 or 579 FGM on 1137 FGAs. 4% less would have been 46.9% or 533 FGM. that's 579 - 533 = 46 FGM over 77 games, or (46 x 2)/77 = 1.2 more pts/g (not counting 3pters). from a team perspective with 46 more misses that'd be about 1/3 of those misses rebounded by the offense with some subsequent scores. so that 1.2 pts/g would be in reality closer to about 1 pt/g (perhaps slightly more or less). that makes up partly for 2.1 zero point team possessions per game since a team possession is on average worth close to about 1.0 point...

And 5 more assists per 48 is a big difference.

i don't see it - in terms of wins generated or how a player's stats contribute towards his team's wins. there are many players who have rung up high numbers of assists (or ast/48min) but have contributed little to their team's fortunes (kevin porter, brevin knight), and team rates of ast/fgm do not correlate well with wins...

Rodman had a 17 PER in 91-92, which according to Bchaikin was his best season overall.

possibly his best season overall, i'd have to simulate all (or most) of his seasons to be sure...

I look at this comp (Curry/Rodman), and it makes my head spin.

fwiw when i simulate the 06-07 knicks with curry for 40 min/g and 82 games, then replace curry with rodman 91-92, the team wins on average 10-11 more games per average 82 game season with rodman...

reverse this and simulate rodman on the 91-92 pistons for 40 min/g and 82 games, then replace him with curry 06-07, the pistons win 14-15 more games with rodman per average 82 game season...

those are some huge differences...

in 06-07 curry got few rebounds, few steals, few blocks, and a ton of turnovers. as a matter of fact, the only starting C last season with less (REB+ST+BS)/40min than curry (9.1/40min) was the nets' jason collins (8.5/40min). the league average for a C (with >= 20 games started) was 13.4/40min...

also curry shot quite well in 06-07 at 57.6% (585/1016), but had 295 TOs. from a team perspective a TO is basically the same as a missed FGA rebounded by the defense (a zero point team possession). since knicks' opponents got 69% defensive rebounds, that'd be like curry committing say just 125 turnovers (like mehmet okur in 06-07) but missing an additional (295-125)/0.69 = 246 FGAs, i.e. scoring 19.5 pts/g but shooting just 585/(1016+246) = 46.4% (league average C shot 51%) with few rebs, st, and bs. who'd want that?....

By crediting Rodman (or any player) with the full value of a possession for securing a defensive rebound you are basically rating him as single-handedly stopping an entire team.

correct, however....

If Rodman really was one man defensive force on 16 or so plays per 48 minutes, then he would deserve the kind of credit you want to give him. I think that it is obvious that he only played a part.

he was all-D 1st team 7 times in 8 seasons, and DPOY in 89-90 and 90-91, and not because of his rebounding. he was quite possibly the best defensive player over a stretch of a decade (late 1980s to late 1990s) who wasn't a premier shot blocker. when playing for the spurs/bulls in the mid-1990s he often guarded shaq rather than david robinson or a bulls C...

The other difference in assumptions that PER makes is that Usage Rating matters. You point out that he (rodman) only turned the ball over 2 times per 48 minutes. But that's only because his teammates didn't dare ever throw him the ball. In many of his great rebounding years his turnover rate was over 20%.

in curry's and rodman's first 6 seasons in the league (curry has played just six years), curry's rate of turnovers per touch were much higher than rodman's. curry turned the ball over with 12% of his touches (1 turnover for every 8-9 touches on offense) and rodman turned the ball over with 9% of his touches (1 turnover for every 11 touches on offense). there is very little a player can do on offense to help his team (i.e. generate wins) when he turns the ball over on a very high rate of 1 out of every 8-9 touches (1 out of every 7-8 touches curry's last 2 seasons), unless he limits his touches or contributes on offense in a major way without the ball, such as with excellent offensive rebounding (or by getting to the line a ton and hitting his FTs), but curry has been a poor offensive rebounder (and while he gets to the line a ton has a career FT% of less than 65%)...
Back to top
View user's profile Send private message Send e-mail Visit poster's website
asimpkins



Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA

PostPosted: Mon Sep 10, 2007 9:11 pm Post subject: Reply with quote
bchaikin wrote:
he was all-D 1st team 7 times in 8 seasons, and DPOY in 89-90 and 90-91, and not because of his rebounding. he was quite possibly the best defensive player over a stretch of a decade (late 1980s to late 1990s) who wasn't a premier shot blocker. when playing for the spurs/bulls in the mid-1990s he often guarded shaq rather than david robinson or a bulls C...


No argument there. PER does not measure defensive (or any other) contributions that don't show up in the box score, and it has never claimed to do so. PER is a summary of box score accomplishments, and it is on that level that it ranks Rodman to be an average player. Rodman was almost certainly a much more valuable player because of his defensive abilities, but that is beyond the scope of what PER sets out to measure.
Back to top
View user's profile Send private message
gabefarkas



Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC

PostPosted: Mon Sep 10, 2007 9:19 pm Post subject: Reply with quote
Flint wrote:

How does that work? Rodman averaged 22 rebounds per 48 that year and committed only 2 turnovers for a +20 in gaining and maintaining possession.. Curry was at +4.6. So the 15 points Curry outscored him by is less than the difference between them in their impact on possession.

How could they possibly have the same PER? I just don't get it.


Curry also scored more efficiently (at a higher TS%), and did so in 5 less MPG. Remember that PER is a per-minute stat, so stretch out Curry's production by 40/35 MPG difference.
Back to top
View user's profile Send private message Send e-mail AIM Address
Flint



Joined: 25 Mar 2007
Posts: 112


PostPosted: Mon Sep 10, 2007 10:15 pm Post subject: Reply with quote
Bchaikin - Are you including ft's in your analysis of Magic? The 4% difference in TS% reflects the difference there. The 88-89 version shot 91% from the line. Young Magic was at 76%. If you included ft's rather than just fg's, does that make an impact?

Re assists, I don't diagree that assists are less important than turnovers, rebounds, attempts, and points. But it is an advantage for Old Magic.

Your simulations are interesting, and absolutely in line with how I would view Rodman's value relative to Curry's. I just find it amazing, ridiculous really, that PER rates Curry's contributions last year as being equal to Rodman at his absolute prime. Rodman averaged just two less offensive rebounds per 48 than Curry did overall. Curry last year was one of only two centers in the three point era to average less than ten rebounds per 48 and five turnovers. (Darryl Dawkins is the other.) And his assist numbers were horrible. He was basically without historic precedent, which is why the Curry Line feature over at Yahoo, a basketball version of the Mendoza line, was so apt. And yet, because he scores well his PER was well above average. This doesn't make sense.

Btw- do you have a link to a post in which you have described the simulation process?

ASimpkins - re the WOW crediting players with team defensive accomplishments, I just don't really buy it. I know it is the big bone of contention, but I don't care to argue the point. I can only say that I saw Eddy Curry allow so many second chance opportunities last year, it killed me. There was one game against Detroit, a three overtime game we actually won, where Curry allowed Mohammed three offensive boards that led to thre crucial second chance buckets. Especially for a center, I think those d-boards matter a lot.

And re Rodman, the basic point to me here is that Rodman's stats were much better in every way than Curry other than a small difference in scoring efficiency, (that year anyway), and a big difference in points scored. How much difference can those extra points Curry scores really make?

Having Rodman out there does force his teammates to take more shots, but thats what most ballers do, they take shots. Yes, they did so less efficiently than Curry did last year, for sure. But Rodman was hardly useless on offense. He generated 4.3 more shots per 48 than Curry did on the offensive end. In my mind, there is no way a 60% ts% can make up that difference.

Rodman's team will take the 22.6 odd shots and free throws Curry took at a lower efficiency, granted. But say that we assume something ridiculous, that even with Rodman's contribution of 10 points on 10.2 net shots (adding up fga and .5fta) at 57.4%, those 22.6 shots only go in at a 50% ts% rate. that means his team scores 22.6 points, four less than Curry. However Rodman's team takes 4.3 more shots. So their net point total, if you carry through the 50% ts%, is 26.9, or .3 more than Curry.

And we haven't even assessed the damage Curry does to the Knicks with his turnovers. Rodman commits two. And in the course of taking the extra 12.x shots he doesn't take that Curry does, his teammates commit turnovers. But I really doubt they commit three more turnovers taking those extra shots. Overall, there are a lot less turnovers being committed by Rodman's team.

Many people seem to think Rodman was an offensive liability. I think the numbers show Rodman had to be a much much more valuable player to have on the court on the offensive end than Curry. Offensive rebounds are more important than scoring from the center position in my mind than shooting. Even a very efficient and prolific scoring center is going to have a hard time keeping up with a center who shoots almost as well as he does on lower volume, but outrebounds him offensively by a good margin. The extra shots his teams take are just a huge source of value.

I think its ridiculous that Rodman isn't going to the Hall of Fame. Somehow people lump him in with Horry I think, as a role player who hooked up with great teammates. I think Berri's view is much more accurate. Rodman was the best player on the Piston Championship teams, and as important to the Bulls teams when he was on the court as any player other than Jordan.
Back to top
View user's profile Send private message
Statman



Joined: 20 Feb 2005
Posts: 241
Location: Arlington, Texas

PostPosted: Mon Sep 10, 2007 11:05 pm Post subject: Reply with quote
Flint wrote:
and as important to the Bulls teams when he was on the court as any player other than Jordan.


I assume you mean other than Jordan OR Pippen.
_________________
Dan

My current national college player rankings (and other stuff):
http://www.pointguardu.com/f136/statman ... post355594
Back to top
View user's profile Send private message Send e-mail
Flint



Joined: 25 Mar 2007
Posts: 112


PostPosted: Mon Sep 10, 2007 11:22 pm Post subject: Reply with quote
i don't know that Pippen was better than Rodman. Considering how well he rebounded that year, and especially his 8.2 orpg, I would say no. But he clearly took a lot of rebounds from his teammates defensively. I don't know, see p. 144 of the WOW and let me know what you think. Personally, I preferred Rodman as a player, but I may be alone in my love of rebounders. I think my own playing style (in high school) makes me view the game a bit differently than most.
Back to top
View user's profile Send private message
Statman



Joined: 20 Feb 2005
Posts: 241
Location: Arlington, Texas

PostPosted: Tue Sep 11, 2007 6:01 am Post subject: Reply with quote
Flint wrote:
i don't know that Pippen was better than Rodman. Considering how well he rebounded that year, and especially his 8.2 orpg, I would say no. But he clearly took a lot of rebounds from his teammates defensively. I don't know, see p. 144 of the WOW and let me know what you think. Personally, I preferred Rodman as a player, but I may be alone in my love of rebounders. I think my own playing style (in high school) makes me view the game a bit differently than most.


I guess I just don't get how you look at things. You just made a point of talking about the huge disparity between Rodman & Curry - since Rodman was better (or MUCH better in case of rebounding) statistically at everything other than TS%, usage, & scoring.

Well - Pippen was better than Rodman at everything statistically other than rebounding (obviously) - MUCH better in scoring, assists, usage, steals, blocks, A/TO, TS%, win shares (including PW%), & PER.

It sounds like to me the ONLY thing that seems to matter to you is if the guy is a great rebounder & possibly a good defender. He can be pretty much a complete non entity offensively (outside of offensive rebounds) - and you don't seem to think that affects his team negatively whatsoever.

Whether you want to believe it or not - usage matters.

As for giving credit to role players (Rodman is the most extreme epitome of that - his role was solely rebounding & defense) as being the backbone to a good team- that is all fine & dandy. However, to seemingly ignore the role of higher usage players (especially an ALL AROUND great player like Pippen) as being vital in a lineup so that a good role player or two can see court time despite their limitations in certain areas (usually offensively) seems short sighted.

I wonder what the final score would be between a team full of Scottie Pippens and a team full of Dennis Rodmans? Those Bulls teams couldn't win a championship without Pippen imo. They still would have won championships (although maybe not at impressively) without Rodman.

Dennis Rodman NEEDS players like Jordan, Pippen, David Robinson, Thomas, Dumars, etc. in order for his team to succeed because of his own obvious limitations. He can then help make them & his team greater. Those players don't NEED a player like Rodman in the lineup to succeed - although it doesn't hurt.
_________________
Dan

My current national college player rankings (and other stuff):
http://www.pointguardu.com/f136/statman ... post355594
Back to top
View user's profile Send private message Send e-mail
Mike G



Joined: 14 Jan 2005
Posts: 3564
Location: Hendersonville, NC

PostPosted: Tue Sep 11, 2007 7:59 am Post subject: Reply with quote
I am pretty sure Berri ranked Rodamn as the best player in the NBA thru 1999 or 2000, by which time he was truly a bit player. He had outlived his usefulness, there was no niche role in which he was considered valuable, and yet Berri thought he was the best player around.

Flint drifted from a reasonable argument that peak (1992) Rodman was more valuable than 2007 Curry, to 2nd-best player on the greatest team ever (Bulls), to best player on the champion (1989-90) Pistons.

Rodman won't get to the HOF anytime soon. Working against him are emotional/social problems that sabotaged some of his teams (SA '95, notably). Bulls managed to corral him for the most part, but still there was the photographer-kicking incident and others.

Statistically, he was near the lowest-of-the-low in some stats (scoring, FT%), and highest-ever in others (rebounding). One in ten rating systems may rank his career highly enough; but it's doubtful that will sway the Hall.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
asimpkins



Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA

PostPosted: Tue Sep 11, 2007 10:23 am Post subject: Reply with quote
Flint wrote:
ASimpkins - re the WOW crediting players with team defensive accomplishments, I just don't really buy it. I know it is the big bone of contention, but I don't care to argue the point.


I don't really want to argue the point either. I'm not really trying to change your mind on who's-better-than-who as much as I was trying to explain why PER works the way it does. Your previous post expressed a lot of bewilderment on the matter (though perhaps it was just rhetorical). You may not ultimately agree with the assumptions that PER makes, but they are coherent and understandable:

Credit defensive rebounds at about 1/3 of a possession instead of the entire possession, and penalize players for hiding on the offensive end. That's how a fantastic one-dimensional rebounder like Rodman ends up ranked about the same as a good one-dimensional scorer like Curry.
Back to top
View user's profile Send private message
kjb



Joined: 03 Jan 2005
Posts: 864
Location: Washington, DC

PostPosted: Tue Sep 11, 2007 11:36 am Post subject: Reply with quote
When I read Berri's book, I thought I was starting to see some merit to his approach. And there may be some merit -- it's a good book. But...then I look a little closer and, I dunno.

Out of curiosity, I ran a search at b-r.com for the top rebound rates since 1977-78. Players needed at least 15,000 minutes to qualify. Then I ran Berri's simplified Win Score metric on those players -- pts + reb + stl + .5 x ast + .5 x blk - fga - tov - .5 x fta - .5 x pf. Then I divided by minutes to get a per minute Win Score.

According to this metric, Rodman rates as the 2nd best player among the top rebounders behind only Charles Barkley. He ranks ahead of players like David Robinson, Shaq, Moses Malone, Hakeem, Larry Bird, KG, Tim Duncan, Kareem, Karl Malone, Dirk, and Elton Brand.

I dunno, but to me there's something not right with a sytem that rates Rodman as "better" than players like Bird, Duncan, Shaq, and Kareem.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
Statman



Joined: 20 Feb 2005
Posts: 241
Location: Arlington, Texas

PostPosted: Tue Sep 11, 2007 1:00 pm Post subject: Reply with quote
kjb wrote:
When I read Berri's book, I thought I was starting to see some merit to his approach. And there may be some merit -- it's a good book. But...then I look a little closer and, I dunno.

Out of curiosity, I ran a search at b-r.com for the top rebound rates since 1977-78. Players needed at least 15,000 minutes to qualify. Then I ran Berri's simplified Win Score metric on those players -- pts + reb + stl + .5 x ast + .5 x blk - fga - tov - .5 x fta - .5 x pf. Then I divided by minutes to get a per minute Win Score.

According to this metric, Rodman rates as the 2nd best player among the top rebounders behind only Charles Barkley. He ranks ahead of players like David Robinson, Shaq, Moses Malone, Hakeem, Larry Bird, KG, Tim Duncan, Kareem, Karl Malone, Dirk, and Elton Brand.

I dunno, but to me there's something not right with a sytem that rates Rodman as "better" than players like Bird, Duncan, Shaq, and Kareem.


Any player that gets alot of rebounds, and doesn't do much of anything offensively (doesn't take, therefore miss, many shots and doesn't touch the ball enough to garner many turnovers) will rate very well in Berri's system.

As far as I can tell - Berri's system is flawed most in terms of overvaluing low usage/big rebound players, much the same way it can be argued that PER is flawed by overvaluing huge usage/low efficiency players.

Neither system (obviously by the nature of linear weights) sees any value of defensive stopper/no offense no rebounding (aka Bruce Bowen) type player.

If Dan R. ever fully hashed out and updated his system (he may have, I dunno) - that would almost certainly be the best player "ranking" system I know of.
_________________
Dan

My current national college player rankings (and other stuff):
http://www.pointguardu.com/f136/statman ... post355594
Back to top
View user's profile Send private message Send e-mail
bchaikin



Joined: 27 Jan 2005
Posts: 685
Location: cleveland, ohio

PostPosted: Tue Sep 11, 2007 1:28 pm Post subject: Reply with quote
...to me there's something not right with a sytem that rates Rodman as "better" than players like Bird, Duncan, Shaq, and Kareem.

agreed - but i can see rodman rated almost as good as some of the very best PFs. on a very poor offensive team, like for example the 02-03 denver nuggets, simulation shows - on a 40 min/g and 82 game basis - dennis rodman 91-92 generating 8-9 more wins than juwan howard 02-03. but it also shows PFs like karl malone 89-90, elton brand 05-06, tim duncan 04-05, larry bird 87-88 (ok SF not PF), and shawn marion 05-06 generating 14-15 more wins per average 82 game season replacing howard. so rodman here generates 6-7 less wins, quite a large differense....

but on a good offense, very poor defensive team (like the 05-06 seattle sonics, and whose PFs were not a key reason for the team's good offense), or on a team replacing a starting PF who was both a poor defender and a poor rebounder, like the 95-96 cleveland cavaliers (danny ferry, but who shot well and committed few turnovers), on a 40 min/g and 82 game basis simulation shows rodman generating close to as many wins (within 1-2 wins) as the above mentioned star PFs....


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 108 posts ]  Go to page Previous  1, 2, 3, 4, 5 ... 8  Next

All times are UTC


Who is online

Users browsing this forum: Yahoo [Bot] and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group