Recovered old threads- miscellaneous topics

[Crow]

I have had almost no success using the site search tool for key words either since the site re-boot and others have reported problems as well. One exception was searching for an unusual word- regularization.

Will there be an administrative inquiry or fix to the problem?


I have had more success searching for member names.



If you really want / need to search the forum and the site search tool remains unwilling to do so by key words

one could:

1) just browse the pages of the forum and the several "bulk" recovered threads manually,

2) search by a member name, if known to be involved with the thread (doesn't have to be the thread's first author, I think it will find any poster's name if it is in a recovered thread)

or 3) use google. And I would strongly suggest this link and targeted, site-specific method as being the most efficient even though it is not completely straightforward and simple:

Add the key word you are searching for in the search bar here
http://www.google.com/search?q=site%3Ah ... tnG=Search
to the site's url that is already there to get a targeted search.



Start by looking for results that have a date of mid-April 2011 or more recent at the top left / very beginning of the text under the item's heading or have the (author, date) added to the title. Those results are going to be available by clicking the heading.

If there is not a recent date or a (author, date) addition in the title or mention of the word "recovered" in the item header then it is less likely the results will be available. But there maybe be exceptions to these rules of thumb, including items with long titles where the (author, date) addition might not be visible in the search results.

Many threads without any of these markings though will be from before the site was damaged and clicking on the item header will go to a statement that the topic does not exist (it doesn't directly at google now, but it still may as explained elsewhere in this explanation). However you could / probably should still check to be sure by clicking the item header.

If that doesn't work, it still may be possible to see some older materials by clicking on the "cache" button at the bottom right of the entry rather than the header. If it is still in google's cache records, it will pull up a saved copy of the material, but it will be just one page of a thread. Searching again for the thread title may find other pages but there are many duplicates and also missing pages so it is a tedious process to pull a full thread together if it is multiple pages.

I assembled 300 or so threads by finding and going thru the pages (mostly this way) and pasting them together. There are many other threads out there undone. The cache has shrunk by 1/3rd in the last 6 weeks and probably will shrink further, so if you want to recover a thread not yet recovered I'd try to do it soon.

If a thread or topic is very important to you and will be used for current research and / or new posting (here or elsewhere) and you have trouble recovering it yourself, mention the title or the subject and I might try to assist with the recovery if I have time.

[wilq]

1) just browse the pages of the forum and the several "bulk" recovered threads manually

Thanks to you there are a lot of them so that's not really an efficient way

If a thread or topic is very important to you and will be used for current research and / or new posting (here or elsewhere) and you have trouble recovering it yourself, mention the title or the subject and I might try to assist with the recovery if I have time.

I have some topics saved on my disc from old APBRmetrics forum and I wondered have you reposted them somewhere? I can't use searching tool so let's try this way... do you remember recovering at least some parts of those subjects [starting with the oldest ones which are from 2007]?
- Predicting games missed to injuries
- Individual Defensive Performance
- Foul Trouble
- Predictors of Adjusted Offensive and Defensive +/- Ratings
- Next in Basketball Analysis
- It's all about winning (championships)
- Quick Olympics Strength Progression Analysis since 1960
- PER Correlate To Offense/Winning?
- retrodiction
- Replacement player value
- Do younger rookies have longer careers?
- Learn R
- 2010-11 NBA Win Predictions
- 2011 deadline trade reviews
- All-Time rankings of current players.
- Value of Assists
- Scorecasting: The Hidden Influences Behind How Sports Are Pl

[Crow]

Already recovered and present on the site:

Predicting games missed to injuries (on page 7 of the forum)

Predictors of Adjusted Offensive and Defensive +/- Ratings (on page 8)

- Next in Basketball Analysis (page 5)

- It's all about winning (championships) (page 7)

- PER Correlate To Offense/Winning? (page 6)

- retrodiction (page 4)

Replacement player value (page 5)

- Do younger rookies have longer careers? (page 8)

- Learn R (page 8)

- 2010-11 NBA Win Predictions (page 2)

- 2011 deadline trade reviews (page 11)

- All-Time rankings of current players (page 10)

- Value of Assists (in the more recovered threads of assist thread on page, about the 5th post down)

Regressing to the Mean (page 11)


Not yet recovered (go ahead and post the material if you can)

Individual Defensive Performance

Foul Trouble (if that is the exact title, then it is needed)

- Quick Olympics Strength Progression Analysis since 1960

- Scorecasting: The Hidden Influences Behind How Sports Are Pl (only the first page is recovered and on page 11 of the forum. You can post all or all but the first page there or its own new thread and I'll adjust what is already up if needed.)

[wilq]

Already recovered and present on the site:

Thank you for the information. It worked out better than I've expected...

Not yet recovered (go ahead and post the material if you can)
Individual Defensive Performance
Foul Trouble (if that is the exact title, then it is needed)
- Quick Olympics Strength Progression Analysis since 1960
- Scorecasting: The Hidden Influences Behind How Sports Are Pl (only the first page is recovered and on page 11 of the forum. You can post all or all but the first page there or its own new thread and I'll adjust what is already up if needed.)

Can I send you those topics zipped into one file [with a size smaller than 1 MB]?
It's clearly your project at this point

[Crow]

Check for a private message on this.

[wilq]

Check for a private message on this.

Package sent. Thanks.

[Crow]

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Fri Nov 21, 2008 10:35 pm Post subject: Individual Defensive Performance Reply with quote
I haven't seen too much on this, besides +/- ratings, how many points the opposing player at a position scored. To me these methods are extremely flawed, defense is all about the team, one leak and every player on the other side of the ball can take advantage of it. I came up with a few categories and will examine them in tonight's Bulls/Warriors game, specifically analyzing the Bulls individual defense. I'm not even going to look at any traditional statistics, merely positive/negative plays. I'll post the results tomorrow probably.
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

Mountain



Joined: 13 Mar 2007
Posts: 1074


PostPosted: Fri Nov 21, 2008 11:40 pm Post subject: Reply with quote
Will be interested to see what you look for and find.
Of course one or a few games won't support strong statistical conclusions but you can notice stuff. Kevin Pelton has done several of these and I and a few others have too and written about it and I am sure many others have done it.

If trying to sort out defensive responsibility is considered too hard on many or most plays, at least for public sources, one fall-back position would be just to count unforgivable mental lapses and clear cases of physical limitations resulting in allowed baskets(for lateral quickness, speed, height, strength or hops or some combo). Defensive "giveaways" tracked to balance against defensive "takeaways".

To me a miss is not necessarily (or even truly) a "stop" and a field goal against is not always a failure.

Last edited by Mountain on Sat Nov 22, 2008 1:50 am; edited 1 time in total
Back to top
View user's profile Send private message

Ryan J. Parker



Joined: 23 Mar 2007
Posts: 285
Location: Charleston, SC

PostPosted: Fri Nov 21, 2008 11:42 pm Post subject: Reply with quote
Please post whatever you come up with.

Quantifying how a player impacts overall team defense is something we need to better nail down.
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sat Nov 22, 2008 1:17 am Post subject: Reply with quote
Thanks for the interest, basically what I did was set up a chart for every Bulls player with the following categories:

Pick and/or Roll Coverage
One-on-one
Help Defense
Turnover Attempts
Defensive Rebounds
Getting back/Leaving Open

If a player does something average for any category it does not get recorded, most plays are not average however, very, very few give no blame. Some plays have multiple positive or negative impacts, each +/- impact for a player gets recorded for those categories. It is pretty tough to keep up with, and my feed cut out for a few minutes, but I like the data I collected. It appears as if Rose and Sefolosha were attacked the most. Thabo and Gooden were pretty easily the worst defensive players on the court for the Bulls tonight. Sefolosha lost 11 1-on-1 battles, compared to only 3 positive 1-on-1 plays. Gooden had 5 negative help Ds, compared to 3 positve help Ds, and 6 negative pick and/or roll coverages to only 1 good. Surprisingly Gordon and Hughes were the best defenders, allowing +5 -2 and +5 -3 respectively for their 1-on-1 defense.

I can post more tomorrow, I would like to do this for every Bulls game this year, that is if it seems useful. I am excited with the data, but I am not really good with all of this analysis yet. I am currently reading Basketball on Paper, great so far. Let me know if anyone has questions, comments, critiques, anything, and I will try to get all of the data in tomorrow.
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

iamawesomer



Joined: 01 Sep 2008
Posts: 7


PostPosted: Sat Nov 22, 2008 2:57 am Post subject: Reply with quote
I watched this game too (as a Warriors fan) so I think I might have a few qualms with tonight's results. Sefalosha looked like he was doing OK, he was guarding Jackson most of the time I remember and Jackson was just draining everything, took some bad looking shots but still dropped in.

Hughes and Gordon were good mostly because Morrow reverted back to the mean a bit tonight and CJ Watson just sucks (offensively, defensively, I don't understand the love he gets a bit from Nelson/announcers. Hopefully Crawford cuts in to his time).

Of course this is just off the cuff thoughts, recording things like you did probably is better.
Back to top
View user's profile Send private message

Harold Almonte



Joined: 04 Aug 2006
Posts: 541


PostPosted: Sat Nov 22, 2008 10:55 am Post subject: Reply with quote
A miss is no a stop. But, why should we change the methodology when boxscoring defensive stats? When a scorer scores, metrics don't take account wether the shot was open, close defended, double teamed, etc.. It just mean +2 for the scorer's rating. Then, who allowed that scorer to score? the team? his assigned defender? his final defender after a shift? the coach because he called a zone? It means something wether he was defended and anyway he got to score?

Scorers are supposed to score and fail sometimes open, sometimes closed defended. Shot defenders are supposed to avoid scoring, sometimes close guarding, sometimes allowing the shot and relying on the law of average.

I believe in team defense, the same I believe in team offense plays called to be ran. Why don't we give a percent of every scorer's point to teammates like adj +/-? Don't deserve creators/1st options/high attemptors to be their scoring underrated? What about higher defensive attemptors? Might they be underrated because they end a lot of time defending players who are not their natural opponent? Are all defensive shifts the product of a man left behind, or a defensive strategy (the same at offense with passes) to maximize (or not to increase disadvantage to) the defensive position in the matchup during an action?

Are every FGA an exclusivelly shooter's decission making? How much of a team decission making they have? Is it fair to penalize shooters by a whole FGA, when it's supposed to be a team offensive play? That's extreme accounting, the same the less radical of not accounting teammates's not passing help/damage, not covering the lost man, etc. The quid is: How much is the value of "off the ball" help in the basketball game for a boxscore rating to start to fairly distribute team defenses and ofenses.
Back to top
View user's profile Send private message

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sat Nov 22, 2008 12:11 pm Post subject: Reply with quote
I got all of the data in an excel file, sorry I am kind of computer illiterate with anything beyond normal use, how do I post a link to the excel page?
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

Ryan J. Parker



Joined: 23 Mar 2007
Posts: 285
Location: Charleston, SC

PostPosted: Sat Nov 22, 2008 1:20 pm Post subject: Reply with quote
Try to share it with Google Docs
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sat Nov 22, 2008 2:48 pm Post subject: Reply with quote
http://spreadsheets.google.com/ccc?key= ... l=en&pli=1

There it is, the categories with just the title listed are positive, and the one next to it are the negative plays. Some categories need explaining: Help defense means either a good help or bad help play, bad as in not getting there, getting there too late, or unneccesary help. Turnover attempts refer to an attempted steal, most end up in steals, so a lot more positive than negative, but a failed attempt really hurts a team, I don't give the steal to who picked it up, just who created it. Rebounds only counts for a missed box out/ good box out or contested rebound, you see very little in this category since the Warriors played a ton of 1-on-1 and very rarely tried to get any off rebs. Leaving open/getting back refers to a missed assignment, or not hustling back, only positive given out for this would be an extreme effort to hustle back on defense.

Obviously everything is extremely subjective, and I must decifer what a player is being told to do on certain plays as well. If a player is just does something average then nothing gets recorded, because it is way too hard to assess average play for every player in a possession, and then try to watch Derrick Rose dunk on the other end. Oh, one more thing the 1-on-1 only applies for an opposing player attempting to break down, create, or score on a defender.

Let me know what you think, I will try to continue through the season, pretty pleased with day 1's results. Comments and criticism very appreciated.
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

Ryan J. Parker



Joined: 23 Mar 2007
Posts: 285
Location: Charleston, SC

PostPosted: Sat Nov 22, 2008 3:02 pm Post subject: Reply with quote
It says I don't have permission to access the document. Did you click the Share button?
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sat Nov 22, 2008 3:03 pm Post subject: Reply with quote
Ryan J. Parker wrote:
It says I don't have permission to access the document. Did you click the Share button?


Sorry, does it work now?
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

Ryan J. Parker



Joined: 23 Mar 2007
Posts: 285
Location: Charleston, SC

PostPosted: Sat Nov 22, 2008 5:13 pm Post subject: Reply with quote
Yup!
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sun Nov 23, 2008 12:19 pm Post subject: Reply with quote
I will be recording the same stats for tonight's game too. Anything to change or add?
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

deepak_e



Joined: 26 Apr 2006
Posts: 409


PostPosted: Sun Nov 23, 2008 12:42 pm Post subject: Reply with quote
Nice work on the defensive charting.

It occurs to me that certain positions are just more likely to be attacked on defense than others. Is this something that we should try to adjust for when crafting an overall defensive rating?

Ultimately, how a player impacts winning isn't just what he does and doesn't do on the court, but what he does compared to other players that could be taking those minutes. So some type of position adjustment seems appropriate.
Back to top
View user's profile Send private message

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sun Nov 23, 2008 1:29 pm Post subject: Reply with quote
deepak_e wrote:
Nice work on the defensive charting.

It occurs to me that certain positions are just more likely to be attacked on defense than others. Is this something that we should try to adjust for when crafting an overall defensive rating?

Ultimately, how a player impacts winning isn't just what he does and doesn't do on the court, but what he does compared to other players that could be taking those minutes. So some type of position adjustment seems appropriate.


Yeah, I would think the one-on-one ratings should have more importance for guards, and help D should have more importance for bigs. I guess I will think about that for now while I am just "collecting" data, but that is definitely something to keep my eye on, thanks.
_________________
Derrick Rose will be the ROTY


Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Fri Nov 21, 2008 10:35 pm Post subject: Individual Defensive Performance Reply with quote
I haven't seen too much on this, besides +/- ratings, how many points the opposing player at a position scored. To me these methods are extremely flawed, defense is all about the team, one leak and every player on the other side of the ball can take advantage of it. I came up with a few categories and will examine them in tonight's Bulls/Warriors game, specifically analyzing the Bulls individual defense. I'm not even going to look at any traditional statistics, merely positive/negative plays. I'll post the results tomorrow probably.
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

Mountain



Joined: 13 Mar 2007
Posts: 1074


PostPosted: Fri Nov 21, 2008 11:40 pm Post subject: Reply with quote
Will be interested to see what you look for and find.
Of course one or a few games won't support strong statistical conclusions but you can notice stuff. Kevin Pelton has done several of these and I and a few others have too and written about it and I am sure many others have done it.

If trying to sort out defensive responsibility is considered too hard on many or most plays, at least for public sources, one fall-back position would be just to count unforgivable mental lapses and clear cases of physical limitations resulting in allowed baskets(for lateral quickness, speed, height, strength or hops or some combo). Defensive "giveaways" tracked to balance against defensive "takeaways".

To me a miss is not necessarily (or even truly) a "stop" and a field goal against is not always a failure.

Last edited by Mountain on Sat Nov 22, 2008 1:50 am; edited 1 time in total
Back to top
View user's profile Send private message

Ryan J. Parker



Joined: 23 Mar 2007
Posts: 285
Location: Charleston, SC

PostPosted: Fri Nov 21, 2008 11:42 pm Post subject: Reply with quote
Please post whatever you come up with.

Quantifying how a player impacts overall team defense is something we need to better nail down.
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sat Nov 22, 2008 1:17 am Post subject: Reply with quote
Thanks for the interest, basically what I did was set up a chart for every Bulls player with the following categories:

Pick and/or Roll Coverage
One-on-one
Help Defense
Turnover Attempts
Defensive Rebounds
Getting back/Leaving Open

If a player does something average for any category it does not get recorded, most plays are not average however, very, very few give no blame. Some plays have multiple positive or negative impacts, each +/- impact for a player gets recorded for those categories. It is pretty tough to keep up with, and my feed cut out for a few minutes, but I like the data I collected. It appears as if Rose and Sefolosha were attacked the most. Thabo and Gooden were pretty easily the worst defensive players on the court for the Bulls tonight. Sefolosha lost 11 1-on-1 battles, compared to only 3 positive 1-on-1 plays. Gooden had 5 negative help Ds, compared to 3 positve help Ds, and 6 negative pick and/or roll coverages to only 1 good. Surprisingly Gordon and Hughes were the best defenders, allowing +5 -2 and +5 -3 respectively for their 1-on-1 defense.

I can post more tomorrow, I would like to do this for every Bulls game this year, that is if it seems useful. I am excited with the data, but I am not really good with all of this analysis yet. I am currently reading Basketball on Paper, great so far. Let me know if anyone has questions, comments, critiques, anything, and I will try to get all of the data in tomorrow.
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

iamawesomer



Joined: 01 Sep 2008
Posts: 7


PostPosted: Sat Nov 22, 2008 2:57 am Post subject: Reply with quote
I watched this game too (as a Warriors fan) so I think I might have a few qualms with tonight's results. Sefalosha looked like he was doing OK, he was guarding Jackson most of the time I remember and Jackson was just draining everything, took some bad looking shots but still dropped in.

Hughes and Gordon were good mostly because Morrow reverted back to the mean a bit tonight and CJ Watson just sucks (offensively, defensively, I don't understand the love he gets a bit from Nelson/announcers. Hopefully Crawford cuts in to his time).

Of course this is just off the cuff thoughts, recording things like you did probably is better.
Back to top
View user's profile Send private message

Harold Almonte



Joined: 04 Aug 2006
Posts: 541


PostPosted: Sat Nov 22, 2008 10:55 am Post subject: Reply with quote
A miss is no a stop. But, why should we change the methodology when boxscoring defensive stats? When a scorer scores, metrics don't take account wether the shot was open, close defended, double teamed, etc.. It just mean +2 for the scorer's rating. Then, who allowed that scorer to score? the team? his assigned defender? his final defender after a shift? the coach because he called a zone? It means something wether he was defended and anyway he got to score?

Scorers are supposed to score and fail sometimes open, sometimes closed defended. Shot defenders are supposed to avoid scoring, sometimes close guarding, sometimes allowing the shot and relying on the law of average.

I believe in team defense, the same I believe in team offense plays called to be ran. Why don't we give a percent of every scorer's point to teammates like adj +/-? Don't deserve creators/1st options/high attemptors to be their scoring underrated? What about higher defensive attemptors? Might they be underrated because they end a lot of time defending players who are not their natural opponent? Are all defensive shifts the product of a man left behind, or a defensive strategy (the same at offense with passes) to maximize (or not to increase disadvantage to) the defensive position in the matchup during an action?

Are every FGA an exclusivelly shooter's decission making? How much of a team decission making they have? Is it fair to penalize shooters by a whole FGA, when it's supposed to be a team offensive play? That's extreme accounting, the same the less radical of not accounting teammates's not passing help/damage, not covering the lost man, etc. The quid is: How much is the value of "off the ball" help in the basketball game for a boxscore rating to start to fairly distribute team defenses and ofenses.
Back to top
View user's profile Send private message

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sat Nov 22, 2008 12:11 pm Post subject: Reply with quote
I got all of the data in an excel file, sorry I am kind of computer illiterate with anything beyond normal use, how do I post a link to the excel page?
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

Ryan J. Parker



Joined: 23 Mar 2007
Posts: 285
Location: Charleston, SC

PostPosted: Sat Nov 22, 2008 1:20 pm Post subject: Reply with quote
Try to share it with Google Docs
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sat Nov 22, 2008 2:48 pm Post subject: Reply with quote
http://spreadsheets.google.com/ccc?key= ... l=en&pli=1

There it is, the categories with just the title listed are positive, and the one next to it are the negative plays. Some categories need explaining: Help defense means either a good help or bad help play, bad as in not getting there, getting there too late, or unneccesary help. Turnover attempts refer to an attempted steal, most end up in steals, so a lot more positive than negative, but a failed attempt really hurts a team, I don't give the steal to who picked it up, just who created it. Rebounds only counts for a missed box out/ good box out or contested rebound, you see very little in this category since the Warriors played a ton of 1-on-1 and very rarely tried to get any off rebs. Leaving open/getting back refers to a missed assignment, or not hustling back, only positive given out for this would be an extreme effort to hustle back on defense.

Obviously everything is extremely subjective, and I must decifer what a player is being told to do on certain plays as well. If a player is just does something average then nothing gets recorded, because it is way too hard to assess average play for every player in a possession, and then try to watch Derrick Rose dunk on the other end. Oh, one more thing the 1-on-1 only applies for an opposing player attempting to break down, create, or score on a defender.

Let me know what you think, I will try to continue through the season, pretty pleased with day 1's results. Comments and criticism very appreciated.
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

Ryan J. Parker



Joined: 23 Mar 2007
Posts: 285
Location: Charleston, SC

PostPosted: Sat Nov 22, 2008 3:02 pm Post subject: Reply with quote
It says I don't have permission to access the document. Did you click the Share button?
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sat Nov 22, 2008 3:03 pm Post subject: Reply with quote
Ryan J. Parker wrote:
It says I don't have permission to access the document. Did you click the Share button?


Sorry, does it work now?
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

Ryan J. Parker



Joined: 23 Mar 2007
Posts: 285
Location: Charleston, SC

PostPosted: Sat Nov 22, 2008 5:13 pm Post subject: Reply with quote
Yup!
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sun Nov 23, 2008 12:19 pm Post subject: Reply with quote
I will be recording the same stats for tonight's game too. Anything to change or add?
_________________
Derrick Rose will be the ROTY
Back to top
View user's profile Send private message AIM Address

deepak_e



Joined: 26 Apr 2006
Posts: 409


PostPosted: Sun Nov 23, 2008 12:42 pm Post subject: Reply with quote
Nice work on the defensive charting.

It occurs to me that certain positions are just more likely to be attacked on defense than others. Is this something that we should try to adjust for when crafting an overall defensive rating?

Ultimately, how a player impacts winning isn't just what he does and doesn't do on the court, but what he does compared to other players that could be taking those minutes. So some type of position adjustment seems appropriate.
Back to top
View user's profile Send private message

Jimmysmithhof



Joined: 29 Oct 2008
Posts: 24
Location: Chicago, Illinois

PostPosted: Sun Nov 23, 2008 1:29 pm Post subject: Reply with quote
deepak_e wrote:
Nice work on the defensive charting.

It occurs to me that certain positions are just more likely to be attacked on defense than others. Is this something that we should try to adjust for when crafting an overall defensive rating?

Ultimately, how a player impacts winning isn't just what he does and doesn't do on the court, but what he does compared to other players that could be taking those minutes. So some type of position adjustment seems appropriate.


Yeah, I would think the one-on-one ratings should have more importance for guards, and help D should have more importance for bigs. I guess I will think about that for now while I am just "collecting" data, but that is definitely something to keep my eye on, thanks.
_________________
Derrick Rose will be the ROTY

[Crow]

Eli W



Joined: 01 Feb 2005
Posts: 327


PostPosted: Sat Dec 09, 2006 11:11 am Post subject: DeanO Interview Reply with quote
Some good stuff in here about what an NBA stats analyst actually does:

http://www.sports-central.org/sports/20 ... he_nba.php
Back to top
View user's profile Send private message Send e-mail Visit poster's website
HoopStudies



Joined: 30 Dec 2004
Posts: 578
Location: Near Philadelphia, PA

PostPosted: Sat Dec 09, 2006 12:11 pm Post subject: Reply with quote
There are some typos in there that I'm trying to get cleaned up...
_________________
Dean Oliver
Author, Basketball on Paper
http://www.basketballonpaper.com
Back to top
View user's profile Send private message Visit poster's website
JonathanG



Joined: 20 Apr 2005
Posts: 32


PostPosted: Sun Dec 10, 2006 10:50 pm Post subject: Reply with quote
Really interesting stuff Dean. Thanks to John for sharing.
_________________
DraftExpress.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
deepak_e



Joined: 26 Apr 2006
Posts: 361


PostPosted: Mon Dec 11, 2006 10:27 am Post subject: Reply with quote
So players aren't really aware of the advanced number crunching that's happening in the background.

What about the coaching staff? Are they really involved in it, or do you mostly communicate with the front office?
Back to top
View user's profile Send private message
gabefarkas



Joined: 31 Dec 2004
Posts: 951
Location: Durham, NC

PostPosted: Mon Dec 11, 2006 11:11 am Post subject: Reply with quote
Dean,
One of the things I took away from the article is that you basically say that most people who are working on quant/stat stuff for teams don't know what they are talking about:

Quote:
There are about five teams with people who do what I do. It involves pretty serious statistical analysis, and very few are qualified. Houston is the easiest example, in that they have a few quantitative analysts on staff...


But then you go on to say:

Quote:
...Everyone in the NBA in this field is pretty successful. It is a tough position to fill. It's not just about producing spreadsheets and information. A person must be both knowledgeable about the game, and technically competent, because the statistics must tell a clear story. This it's about translation — one can't simply have a computer science background, or a basketball background, but both.


So now I'm a little confused. The analysts don't know what they are doing, but they're really successful? Is this one of the typos you're trying to get fixed?

I'm hoping you could elaborate a little. I'm in the middle of doing graduate studies in statistics right now precisely because I would like to ensure that I'm technically qualified, while also trying to ensure that the knowledge of the game is there as well. Is this not worth it? Should I drop out right now?
Back to top
View user's profile Send private message Send e-mail AIM Address
kjb



Joined: 03 Jan 2005
Posts: 715
Location: Washington, DC

PostPosted: Mon Dec 11, 2006 11:14 am Post subject: Reply with quote
gabe: That's not how I interpreted Dean's remarks at all. I think his point was that there are few people in the general population who are qualified to be stat analysts for NBA teams -- NOT that people currently in those positions are unqualified.
_________________
My blog
Back to top
View user's profile Send private message Visit poster's website AIM Address Yahoo Messenger
HoopStudies



Joined: 30 Dec 2004
Posts: 578
Location: Near Philadelphia, PA

PostPosted: Mon Dec 11, 2006 1:16 pm Post subject: Reply with quote
WizardsKev wrote:
gabe: That's not how I interpreted Dean's remarks at all. I think his point was that there are few people in the general population who are qualified to be stat analysts for NBA teams -- NOT that people currently in those positions are unqualified.


Yup. Thanks, Kev.
_________________
Dean Oliver
Author, Basketball on Paper
http://www.basketballonpaper.com
Back to top
View user's profile Send private message Visit poster's website
tpryan



Joined: 11 Feb 2005
Posts: 68


PostPosted: Tue Dec 12, 2006 12:39 am Post subject: Reply with quote
HoopStudies wrote:
There are some typos in there that I'm trying to get cleaned up...

I noticed three: lower case "i" instead of capital, a word missing near the end, and "This" should be "Thus" (I assume), close to the beginning.
Back to top
View user's profile Send private message
tpryan



Joined: 11 Feb 2005
Posts: 68


PostPosted: Tue Dec 12, 2006 1:41 am Post subject: Reply with quote
I believe the primary limiting factors regarding "very few are qualified" are probably (1) knowledge of the game of basketball, and (2) knowledge of the statistical approaches that have been used to date and how successful they have been.

Gabe, you are apparently pursuing an M. A. in statistics at Columbia. That should provide the necessary statistical background, but courses in statistical theory, probability theory, measure theory, etc. will be of little, if any, help.

As far as I know, there is no one who (1) has a Ph.D. in statistics and is well known in the field, (2) is an active participant here, and (3) has a very solid knowledge of contemporary NBA basketball. Are all three of these necessary for success in the NBA? Probably not, but anyone who is not close to meeting all three of these ideal characteristics might struggle at least slightly if employed by an NBA team.
Back to top
View user's profile Send private message
holymoly



Joined: 30 Jul 2005
Posts: 43


PostPosted: Tue Dec 12, 2006 6:14 am Post subject: Reply with quote
Nice comments Dean. One point I wanted to raise was the way you kinda dismissed the use of your tools for scouting Euro/International players. I happen to think that yours and the communitys tools give some great insight into these players and their roles despite their lower minutes. It is also worth noting that many teams participate in up to 4 different competitions (Balkan teams especcially), so accumulating these minutes can give you some very serious playing time. Bearing in mind how difficult it is for NBA teams to traditionally scout a player in Brazil, China, Argentina in comparison with an NCAA player, these stats tools give GM's and scouts a much needed extra indicator dont they?
Back to top
View user's profile Send private message Send e-mail MSN Messenger
HoopStudies



Joined: 30 Dec 2004
Posts: 578
Location: Near Philadelphia, PA

PostPosted: Tue Dec 12, 2006 9:47 am Post subject: Reply with quote
holymoly wrote:
Nice comments Dean. One point I wanted to raise was the way you kinda dismissed the use of your tools for scouting Euro/International players. I happen to think that yours and the communitys tools give some great insight into these players and their roles despite their lower minutes. It is also worth noting that many teams participate in up to 4 different competitions (Balkan teams especcially), so accumulating these minutes can give you some very serious playing time. Bearing in mind how difficult it is for NBA teams to traditionally scout a player in Brazil, China, Argentina in comparison with an NCAA player, these stats tools give GM's and scouts a much needed extra indicator dont they?


I'm still learning how interviews go, but I will let people know that it is common for your actual words to not show up at all. I'm not sure that any sentence that is printed of that interview is exactly what I said. Most of it captures the essence.

If it comes across dismissing analysis in other leagues, it shouldn't. I know a few people (coaches and other analysts) who use it extensively in Europe. I have applied the methods with several international leagues for one reason or another as well as junior high and high school kids. However, there are a lot of "prospects" over in Europe who do get very limited minutes and, in those cases, it is difficult to do much in the way of analysis. Yes, when minutes accumulate, you can do better.
_________________
Dean Oliver
Author, Basketball on Paper
http://www.basketballonpaper.com
Back to top
View user's profile Send private message Visit poster's website
mikez



Joined: 14 Mar 2005
Posts: 36


PostPosted: Tue Dec 12, 2006 12:19 pm Post subject: Reply with quote
Quote:
So players aren't really aware of the advanced number crunching that's happening in the background.


There are actually some players who have a decent idea of some of the stuff that's going on. But not most of the players, and certainly not all of the stuff.

As for peoples' qualifications, the other people I know who doing this sort of work are uniformly very smart. But IMHO it takes more than just smarts and some basketball/stats knowledge to do well at this sort of thing, as I mentioned at the bottom of a post to Dan R.'s blog about a year ago:

http://danrosenbaum.blogspot.com/2005/0 ... ar-is.html

A stats PhD definitely won't hurt though. Smile

-MZ
Back to top
View user's profile Send private message
tpryan



Joined: 11 Feb 2005
Posts: 68


PostPosted: Tue Dec 12, 2006 4:14 pm Post subject: Reply with quote
Dean,

I was interviewed by the Cleveland newspaper about 10 years ago for an article that was not sports related. Some of the "quotes" from me contained words that I never even use. The interviewer simply made up his own quotes. Quite frankly, I consider that to be poor journalism and I was annoyed by it.
Back to top
View user's profile Send private message
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 537
Location: Greensboro, North Carolina

PostPosted: Tue Dec 12, 2006 4:23 pm Post subject: Reply with quote
mikez wrote:
As for peoples' qualifications, the other people I know who doing this sort of work are uniformly very smart. But IMHO it takes more than just smarts and some basketball/stats knowledge to do well at this sort of thing, as I mentioned at the bottom of a post to Dan R.'s blog about a year ago:

http://danrosenbaum.blogspot.com/2005/0 ... ar-is.html

A stats PhD definitely won't hurt though. Smile

-MZ

Your comment is right on the mark. Communcation skills are just as important as basketball smarts or stats smarts. It is obviously not a sufficient condition, but it arguably more important than either basketball or stats smarts.

And I would be amiss in not mentioning that an MA or PhD in economics may be a better launching point into an NBA analytical position than even a degree in statistics. Economists have a lot of practical experience dealing with the omitted variable and/or endogeneity issues that are at the heart of basketball statistical analysis. The econometrics problems are very similar to those faced in empirical fields like labor and public economics. In addition, these analytical positions generally are not assistant coaching jobs; they are management jobs and so a good understanding of the economics involved in the free agent market and the salary cap/luxury tax system is very useful.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
RocketsFan



Joined: 23 Oct 2005
Posts: 398


PostPosted: Tue Dec 12, 2006 6:09 pm Post subject: Reply with quote
Dean, you got your degree in Environmental Engineering, correct? How do you apply what you learned in that field to the analysis you do in the NBA?

Oh, and John, thanks for posting the article.

page 2 of 3 missing

SGreenwell



Joined: 12 Feb 2005
Posts: 76
Location: Rhode Island

PostPosted: Tue Dec 19, 2006 1:16 pm Post subject: Reply with quote
I've done many stories without a recorder - Normally, it means using shorter quotes, and as Dean said, using paraphrasing to capture at least the intent if you can't get the exact wording.

However, I've been amazed at the advancements in recording technology the past few years, so I finally broke down and bought a digital recorder for $40. They pick up talking crystal clear from about 10 feet away, you can dump the audio to computers easily via USB, much easier to sort quotes out, etc. It surprises me that more reporters don't use them, since the last time I covered a college basketball game maybe a quarter had them.

I think that what I was trying to get at is that many people don't realize that they are not great public speakers. If you ever want to test the theory, pull a Tobias from Arrested Development and try recording yourself for a day. I've done it, and it was somewhat painful to listen to Wink
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
John Hollinger



Joined: 14 Feb 2005
Posts: 175


PostPosted: Wed Dec 20, 2006 9:45 am Post subject: Reply with quote
Wiz Kev is right that everybody at NBA games has a recorder, and I can vouch for the accuracy of quotes I've seen in the paper where I was also present. All I can say is that I've scratched my head a few times when I've looked at how I've been quoted. Maybe I need to record myself as I go.
Back to top
View user's profile Send private message AIM Address
KnickerBlogger



Joined: 30 Dec 2004
Posts: 180


PostPosted: Thu Dec 21, 2006 1:31 pm Post subject: Reply with quote
That's why I interview by email. Well that and I have no access to anyone in person or by phone. Ahhh the life of a lowly blogger.

Other than grammatical/spelling changes, or occasionally changing the order of questions in order to make the article read better, there's not much to do other than cut & paste. Although, you have to think things out in advance & send all the questions at once. I know occasionaly I've sent a 2nd email, but you don't want to send 20 emails to do an interview.
_________________
KnickerBlogger.Net - now indispensable!
Back to top
View user's profile Send private message Visit poster's website AIM Address Yahoo Messenger
gabefarkas



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

PostPosted: Wed Jan 03, 2007 10:44 am Post subject: Reply with quote
HoopStudies wrote:
mikez wrote:

As for getting "breaking in" or getting the job, the vast majority of people I know who work for sports teams worked for the teams for free (mostly via internship) before working there. To get hired off of one's reputation is a much more difficult task, whether you're talking about a basketball analytics position or any other non-coaching position with a team. The best advice I have for anyone looking to get one of these analytics jobs is to meet with someone at a team & leave them with something you think could help them -- if you manage to convince them to look at it & take it seriously (not a trivial task) then you might get the chance to do some more work, either for free or as a consultant, & get your foot in the door that way. Much easier (though still not easy) to do this than it is to get a job by becoming well-known in the field. Just like for any other job, it also doesn't hurt to network, contact everyone you know who works for a team to ask if they know of anyone looking for someone, etc...


There are thousands of people trying to get into NBA organizations that have almost no jobs to offer. This includes lots of smart people who say they "have a passion" for the game. Many say it, but then don't show it, either because they really don't have it or because they can't afford to.

MikeZ is right. I would only add that what you leave them with should be pretty convincing and unique, hard to get other ways. You have to create your own position or plan on taking a pretty low pay, low responsibility position to just start with.


I think that last point is true of a lot of high-profile, high-desire jobs. But how do you evaluate whether or not someone is conveying a real passion?
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: Wed Jan 03, 2007 12:34 pm Post subject: Reply with quote
gabefarkas wrote:

I think that last point is true of a lot of high-profile, high-desire jobs. But how do you evaluate whether or not someone is conveying a real passion?


There are a lot of people who say they want to get in really badly, but if I ask them to do something, it is usually pretty clear who spends time thinking about it and putting in the real work. There needs to be more insight in what they give me than I can get in a short period of just thinking about it.

Many people email me with their own pre-packaged player rating formula also -- the problem here can be justifying that it is better than any of the other things out there. I have never gotten good evidence from the random emailers and most of the time, I have never gotten anything, just words that if I trust them, I will make much better decisions.

If you wanted to try to become a nuclear engineer, how much time would you spend trying to learn the lingo, reading books, taking classes, understanding the business? You have to love what you want to do a lot to spend your free time learning the things you need to learn. If you are already a nuclear engineer, imagine someone coming to you and asking how to get in? If they showed some credentials, you might give them an informational interview or even an internship to do pretty menial, but necessary jobs. That way you can see whether they have the interest and ability.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
mikez



Joined: 14 Mar 2005
Posts: 75


PostPosted: Wed Jan 03, 2007 5:27 pm Post subject: Reply with quote
Quote:
I think that last point is true of a lot of high-profile, high-desire jobs. But how do you evaluate whether or not someone is conveying a real passion?


I'd second everything DeanO wrote. I don't have any sort of public profile (certainly not in relation to Dean's "fame" Smile ) and yet I still get multiple resumes/cover letters every week, most of which are from people who say they have a passion for the game and claim (believably) to have followed it closely for years. And most of the resumes evince at least OK credentials. But a resume and cover letter are unlikely to convey real passion alone -- you'll most likely need to do something else to distinguish yourself, either by having a relatively unique skillset or analysis, or by working hard for free for a bit and doing an excellent job, or more likely both. I think you're right that this is the case for most high-profile, high-desire jobs, but I'd argue that even in those jobs the people with real passion (and real skill) do tend to stand out, one way or another.

-MZ
Back to top
View user's profile Send private message
adandar



Joined: 27 Dec 2005
Posts: 13
Location: Philadelphia

PostPosted: Fri Jan 26, 2007 3:26 pm Post subject: Reply with quote
I have to agree with the sentiments of Dean and Mikez because I know at some point I was one of those random emailers. I may still even be one...

I used to think that spending hours pouring over numbers or reading books or watching games was enough to justify that I had an intense interest (a.k.a. passion) and I would be destined for great things if given the "chance."

It's only now that after speaking with many people who actually have made their way to some degree or another that it takes so much more than any of that stuff and I only hope at 25 that I didn't learn that lesson a little too late.

I think many of "us" on the outside think that since some pioneers paved the way with great sacrifice and that if we know a thing or two that we can simply stand on the shoulders of geniuses to become the next big thing. And with all of that, suddenly we are "qualified" and deserve an opportunity.

To be clear I'm not passing judgment. Rather, think of me like a recovering addict; an addict of intellectual arrogance and this is my confession. The best thing to ever happen to me was a wicked ego check laid on me not too long ago. If my story sounds familiar, I'd recommend you receive a nice ego stomping as quickly as possible because it may sting a bit at first, but will be worth it in the end if a life in sports is what you truly desire.

[Crow]

The Specialist



Joined: 23 Apr 2008
Posts: 16
Location: Houston, TX

PostPosted: Mon Sep 06, 2010 4:53 pm Post subject: DeanO interpretation Reply with quote
Can someone help me understand how DeanO started out with the number 20 for the variance (which sounds like it's an estimate) and what that number represents if I were to try to calculate it. Also the 150, not sure if that is a constant to use the whole time or varies based on another calculation that he does not detail.

Any insight is appreciated.

http://www.rawbw.com/~deano/articles/kalman.html
Back to top
View user's profile Send private message
HoopStudies



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

PostPosted: Tue Sep 07, 2010 4:10 pm Post subject: Reply with quote
Wow. Digging deep into archives.

It's just a Bayesian prior variance, essentially framing the estimate as plus or minus 9 points. It reflects how well you know the value of a team's rating. You could start with a bigger variance and it doesn't matter too much. I probably wouldn't shrink it much, though.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.

[Crow]

Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 978
Location: Seattle

PostPosted: Thu Jan 27, 2005 10:59 am Post subject: More standardization Reply with quote
Trying to explain what we're doing to other people constantly reminds me how confusing it is to have multiple definitions. So would it be possible for us to agree on a free-throw multiplier?
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jkubatko



Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH

PostPosted: Thu Jan 27, 2005 11:14 am Post subject: Re: More standardization Reply with quote
admin wrote:
Trying to explain what we're doing to other people constantly reminds me how confusing it is to have multiple definitions. So would it be possible for us to agree on a free-throw multiplier?


Could we have people present cases for each of the above? Perhaps John Hollinger could make the case for 0.44, Dean Oliver for 0.4, and ??? for 0.45.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Ed Küpfer



Joined: 30 Dec 2004
Posts: 785
Location: Toronto

PostPosted: Thu Jan 27, 2005 11:46 am Post subject: Re: More standardization Reply with quote
jkubatko wrote:
Could we have people present cases for each of the above? Perhaps John Hollinger could make the case for 0.44, Dean Oliver for 0.4, and ??? for 0.45.

It's important to remember that there's an essential tension in everything we do between simplicity and accuracy. Almost always, a move towards one brings us away from the other.

On the simplicity side, there's 0.4. That's a single significant digit to remember.

On the other side, there's 0.44. I think different methods have all supported the accuracy of this number.

I did a little test. Using the same player's numbers, I changed only the multiplier. Here's how his stats changed:
Code:
POSS ORTG FLR% DPOSS DRTG STP%
0.4 761 118.5 53.3% 869 103.6 51.1%
0.44 774 115.6 53.9% 878 102.6 50.6%
0.5 794 111.5 54.8% 891 101.2 49.9%

Those changes are huge. We need to decide on a number. I'm going to go with the most accurate -- there is no need for simplicity.
_________________
ed
Back to top
View user's profile Send private message Send e-mail
Dan Rosenbaum



Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina

PostPosted: Thu Jan 27, 2005 11:46 am Post subject: Reply with quote
Ed Küpfer wrote:
Here's how the types of FTs break down:
Code:
FTAs
TYPE 1 2 3 TOTAL

Away from Play 35 4 0 39
Clear Path 28 0 0 28
Delay Technical 4 0 0 4
Double Technical 1 0 0 1
Elbow 0 4 0 4
Flagrant Type 1 0 77 0 77
Flagrant Type 2 0 12 0 12
Hanging Technical 12 0 0 12
Illegal Defense 341 0 0 341
Inbound 0 24 0 24
Loose Ball 3 1002 0 1005
Non Supported Technical 2 0 0 2
Offensive 0 8 0 8
Personal 8 6628 3 6639
Punching 2 0 0 2
Shooting 2473 18578 303 21354
Taunting Technical 6 0 0 6
Technical 446 0 0 446

TOTAL 3361 26337 306 30004

The FTAs that represent the end of a possession -- Shooting, Loose Ball, Away from Play, and Personal -- sum to 13259, or 44.2% of the total. Nice to have that FTA coefficent confirmed.

I think this data represents all of the free throws this season up through January 21st. This is just fabulous to have these data.

I think the number of possessions is equal to:

two-shot away from play divided by two = 4/2 = 2
two-shot elbow divided by two = 4/2 = 2
two-shot inbound divided by two = 24/2 = 12
two-shot loose ball divided by two = 1002/2 = 501
two-shot offensive divided by two = 8/2 = 4
two-shot personal divided by two = 6628/2 = 3314
three-shot personal divided by three = 3/3 = 1
two-shot shooting divided by two = 18,578/2 = 9289
three-shot shooting divided by three = 303/3 = 101

Add this all up and you get 13,226, which is 44.0808% of the 30,004 total free throws.

Note that a couple of these categorizations are a little fishy, but ...
40 percent of 30,004 is 12,002
43 percent of 30,004 is 12,902
44 percent of 30,004 is 13,202
45 percent of 30,004 is 13,502
50 percent of 30,004 is 15,002

So even if we haggle a bit over a few of these foul categories, we are going to end up pretty close to 0.44. Like others, I have also found 0.44 to work best in terms of balancing offensive and defensive team possessions.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Yahoo Messenger
HoopStudies



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

PostPosted: Thu Jan 27, 2005 1:07 pm Post subject: Re: More standardization Reply with quote
Ed Küpfer wrote:

It's important to remember that there's an essential tension in everything we do between simplicity and accuracy. Almost always, a move towards one brings us away from the other.

On the simplicity side, there's 0.4. That's a single significant digit to remember.

On the other side, there's 0.44. I think different methods have all supported the accuracy of this number.

... I'm going to go with the most accurate -- there is no need for simplicity.


In all cases, there are 3 considerations in making this kind of decision:

1. Reality or accuracy
2. Simplicity
3. Consistency

These are listed in the importance I personally place on them. The 0.44 is more accurate. A few people have shown that. And it's not really much more complex or inconsistent since adding a digit ain't hard and Hollinger has had .44 out for a while. I'm certainly planning on using it and do use it in some things I've done, even though I've been using .4 for nearly 20 years.

The only realm in which it has a problem is in the possession calculation. I've been tracking real possessions this year pretty carefully, even knocking off the fake ones that come out of play by plays at the end of quarters. And if you just use 0.44*FTA you seem to overestimate those real possessions because of slight errors in the other part of the equation. This is a minor point. I've been dealing with various formulas for possessions for years. Renormalizing them to some mean is not a huge deal.

Anyway, I do think the .44 is the best factor to use, but it has some cascading effects in terms of fixing other things.

I'll try to post some of the study on possessions so someone else can polish the job. Just not this week.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
gabefarkas



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

PostPosted: Fri Jan 28, 2005 5:01 pm Post subject: Re: More standardization Reply with quote
Ed Küpfer wrote:

Those changes are huge. We need to decide on a number. I'm going to go with the most accurate -- there is no need for simplicity.


they are indeed huge, but they are huge for every player's stats for which the multiplier is applied (ie, somewhat evenly across the board).
Back to top
View user's profile Send private message Send e-mail AIM Address
Ed Küpfer



Joined: 30 Dec 2004
Posts: 785
Location: Toronto

PostPosted: Fri Jan 28, 2005 5:07 pm Post subject: Re: More standardization Reply with quote
gabefarkas wrote:
they are huge for every player's stats for which the multiplier is applied (ie, somewhat evenly across the board).

They are internally consistent, but if someone tells me that some player is hitting 1.02 points per possession, I don't want to be asking about which FT coefficient was used so I can compare it to my numbers. I mean, if it's going to cause some confusion among us knowledgable types, imagine the confusion among the less statistically savvy.
_________________
ed

[Crow]

BorisD



Joined: 29 Jul 2007
Posts: 14


PostPosted: Sat Aug 07, 2010 2:29 am Post subject: Need Help with DeanO's Net Points & Project Def. Score S Reply with quote
I was reviewing Basketball on Paper this week (as I'm often wont to do) and I mulled over the chapter on Individual Won-Lost records a little more thoroughly than I have, especially the segment on the Net Points ideas. Dean had two ideas for how to calculate net points, one formula based on efficiency, the other on production and the total would be the average of the two. His second one was largely derived from the estimated Defensive Rating.

However, I was inspired by Dean's Project Defensive Score Sheet idea and kept score in every single Raptors' game this year (not friendly on the eyes or the blood pressure this past year, let me assure you). Now, I have that data, but here's my problem for the net points formula. I know exactly how many points each player directly allowed, but I also have 1017 points that were allowed by nobody in particular - thin air, if you will. I figured I would calculate how many points that works out to per player-minute (which was .0428), multiply that by the number of minutes each player played and just add it to the points allowed to get my answer, but then I thought more deeply. It isn't really fair or justifiable to throw those points on equally to players who bear more than an average amount of the team's possessions defensively (which, by definition, would be .200) than it would be to those who don't.

My question is, would anybody whose mathematical skills are beyond my first-year undergrad calculus for business which I took 14 years ago be able to advise on a formula I could use to adjust the amount of those "thin air" points allowed to the player on a sliding scale based on their DPoss%? 'twould be handy to just dump the formula into Excel and let it do the work for me...

Thanking you all in advance.

[Crow]

mikez



Joined: 14 Mar 2005
Posts: 75


PostPosted: Mon Oct 12, 2009 8:48 am Post subject: DeanO Article from Denver Post Reply with quote
Nice Denver Post article on DeanO today.

-MZ
Back to top
View user's profile Send private message
HoopStudies



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

PostPosted: Mon Oct 12, 2009 8:55 am Post subject: Reply with quote
I told Ben Hochman that I couldn't give away much. So don't go through this looking for clues to how I practice the craft. It has some Basketball on Paper methodologies and that's already public.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Mike G



Joined: 14 Jan 2005
Posts: 3603
Location: Hendersonville, NC

PostPosted: Mon Oct 12, 2009 9:34 am Post subject: Reply with quote
Quote:
As much as Oliver helps executives decide whom they should try to acquire, he helps even more pointing out which players the Nuggets should avoid. "There is not a huge predictive value," Nuggets executive Mark Warkentien said. "There is a huge — bigger than downtown Denver — eliminating value."

Comments
1) Warkentien possesses uncommon quantification-to-verbalization skill
2) Invert a list ranking "Players to Avoid", and wouldn't you have a list of "Players to Covet"?
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
mtamada



Joined: 28 Jan 2005
Posts: 377


PostPosted: Mon Oct 12, 2009 7:24 pm Post subject: Reply with quote
Mike G wrote:

1) Warkentien possesses uncommon quantification-to-verbalization skill


Yeah, that was my big take-away from the article. DeanO we already know about, but Warkentien's quotes were impressive. His background seems to be traditional (college coach, NBA scout), but he seems to have a solid grasp of the value of quantitative analysis and how to use it.
Back to top
View user's profile Send private message
HoopStudies



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

PostPosted: Mon Oct 12, 2009 7:49 pm Post subject: Reply with quote
mtamada wrote:
Mike G wrote:

1) Warkentien possesses uncommon quantification-to-verbalization skill


Yeah, that was my big take-away from the article. DeanO we already know about, but Warkentien's quotes were impressive. His background seems to be traditional (college coach, NBA scout), but he seems to have a solid grasp of the value of quantitative analysis and how to use it.


Mark is tremendous. A fine boss, a fine GM, a fine person. He challenges me with his requests and I challenge him in some of his beliefs. It's great working with him.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Serhat Ugur (hoopseng)



Joined: 13 Oct 2006
Posts: 209
Location: Basketball Research

PostPosted: Tue Oct 13, 2009 2:02 am Post subject: Reply with quote
Quote:
In his eyes, games are a series of possessions, and the simple way to win this game is to maximize your possessions and minimize your opponent's possessions.

This is lost in translation part of the author, I think...
Possession should be replaced with the word efficiency?
_________________
http://www.nbastuffer.com
Back to top
View user's profile Send private message Visit poster's website
gabefarkas



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

PostPosted: Tue Oct 13, 2009 7:22 am Post subject: Reply with quote
HoopStudies wrote:
Mark is tremendous. A fine boss, a fine GM, a fine person. ... It's great working with him.

HoopStudies wrote:
The postings are my own & don't necess represent positions, strategies or opinions of employers.

I'm sorry but those two in the same post just made me giggle. I realize one is the signature, and that I'm making a completely fictitious stretch, but I found it humorous nonetheless.
Back to top
View user's profile Send private message Send e-mail AIM Address
Qscience



Joined: 22 Jun 2009
Posts: 70
Location: Phoenix, Arizona

PostPosted: Tue Oct 13, 2009 9:50 am Post subject: Reply with quote
Quote:
Asked if Lawson, the 18th overall pick, should have gone higher in the draft, Oliver could only smile and say, "I don't care — we got him."


This is the beauty of Quantitative Analysis - numbers do not listen to whom the media picks as its best player. As a matter of fact I bet Dean also had Ty above Jonny Flynn.
The floor analysis I did on Ty however showed that he is much more suited for an uptempo style. Denver is trending towards defensive personnel and that usually equates to fewer possessions and a slower pace to the game.

The Quantitative Analysis field needs more articles like this ...gj Dean.
Back to top
View user's profile Send private message Visit poster's website
John Hollinger



Joined: 14 Feb 2005
Posts: 175


PostPosted: Mon Oct 19, 2009 1:46 pm Post subject: Reply with quote
Echoing what Dean said, I've found that among those with the old school "traditional" background, Warkentien grasps analytics concepts as well or better than anyone I've talked to.
Back to top
View user's profile Send private message AIM Address
bchaikin



Joined: 27 Jan 2005
Posts: 689
Location: cleveland, ohio

PostPosted: Sun Oct 17, 2010 10:46 pm Post subject: Reply with quote
more from deano:

http://www.slamonline.com/online/nba/20 ... -analysis/
Back to top
View user's profile Send private message Send e-mail Visit poster's website
schtevie



Joined: 18 Apr 2005
Posts: 412


PostPosted: Mon Oct 18, 2010 8:58 am Post subject: Reply with quote
There have been enough of these that would be interesting to do an interpretive retrospective on all of Dean's print interviews, from the perspective of seeing how things have changed over time in this little world we comment on.

Having read several over the years, I am always struck by the question that the interviewer never asks - not that Dean would answer, but still. Given that the purpose of "quantitative analysis" is to discover, quantify, and evaluate competitive opportunities, why does the interviewer never ask (and not just Dean, but any of the several who dabble in the trade) for an estimate of the value of his output to the team: that of advice taken, both positive and negative, and advice ignored.

That this general question is never asked (and those that could answer have zero incentive to volunteer such estimates) contributes to maintaining the basic state of the world we observe, where those who do quantitative analysis for teams - despite the growth in their number over the past decade or so - appear to be distinctly undervalued within the NBA economy.
Back to top
View user's profile Send private message
greyberger



Joined: 27 Sep 2010
Posts: 50


PostPosted: Mon Oct 18, 2010 5:25 pm Post subject: Reply with quote
Undervalued? Underheralded? Comparisons with other sports are difficult. NBA teams like to travel light and the economy might be slowing the infiltration process a bit. Hiring a quant consultant is a hard sell when teams are firing scouts and skimping on coach and FO staff. We are a ways away from where teams feel they're missing out on wins or revenue if they're not paying a nerd.
Back to top
View user's profile Send private message Send e-mail
jim



Joined: 01 Aug 2009
Posts: 13


PostPosted: Mon Oct 18, 2010 6:13 pm Post subject: Reply with quote
Quote:
NBA teams like to travel light and the economy might be slowing the infiltration process a bit. Hiring a quant consultant is a hard sell when teams are firing scouts and skimping on coach and FO staff.

All this talk about frugality seems funny when teams are still signing a 29 year old Joe Johnson to a 6 year $119 million dollar contract. There are pennies, and then there are benjamins.

Quote:
We are a ways away from where teams feel they're missing out on wins or revenue if they're not paying a nerd.

Well, if teams are missing out on wins or revenue by not hiring some pocket protectors, then they're undervaluing those geekazoids.
Back to top
View user's profile Send private message
gabefarkas



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

PostPosted: Mon Oct 18, 2010 8:44 pm Post subject: Reply with quote
schtevie wrote:
That this general question is never asked (and those that could answer have zero incentive to volunteer such estimates) contributes to maintaining the basic state of the world we observe, where those who do quantitative analysis for teams - despite the growth in their number over the past decade or so - appear to be distinctly undervalued within the NBA economy.
Typically, an interviewer will come to the table with many more questions prepared than those ending up in the final printed version of the interview. The questions (and subsequent responses) not appearing achieve their status due to one or more of several reasons. Chiefly among these, viz. the lack of incentive to provide an interesting answer, or prior knowledge regarding this.
Back to top
View user's profile Send private message Send e-mail AIM Address
BobboFitos



Joined: 21 Feb 2009
Posts: 200
Location: Cambridge, MA

PostPosted: Mon Oct 18, 2010 10:34 pm Post subject: Reply with quote
greyberger wrote:
Undervalued? Underheralded? Comparisons with other sports are difficult. NBA teams like to travel light and the economy might be slowing the infiltration process a bit. Hiring a quant consultant is a hard sell when teams are firing scouts and skimping on coach and FO staff. We are a ways away from where teams feel they're missing out on wins or revenue if they're not paying a nerd.


Eh, teams aren't really saving money by not hiring the 70k/yr quant analyst. The teams that don't have a quant position simply don't realize they cover their costs+more by having one. Unfortunately, (up to this point at least, subject to change) the field hasn't really advanced to the point that all teams feel a quant analyst is necessary to compete. They may be right.
_________________
-Rob

acollard



Joined: 22 Sep 2010
Posts: 56
Location: MA

PostPosted: Tue Oct 19, 2010 9:55 am Post subject: Reply with quote
jim wrote:
Well, if teams are missing out on wins or revenue by not hiring some pocket protectors, then they're undervaluing those geekazoids.


I think in order to determine the value of the geeks, we first need to figure out the marginal profit created by a win in the NBA. The cost of a win share on the free agent market I believe has been determined, but in terms of profits for each additional win, I'm under the opinion its not a very strong correlation. I couldn't quickly find any studies that have determined this, though I couldn't access "Profitability in professional sports and benchmarking: the case of NBA franchises", which seemed to be relevant.

So I did a little bit of research on my own. Based on Forbes' operating income for 2008-2009 season, I got each marginal win to account for .27 million dollars, and obviously with only one year of data it was not significant (P=.29) and it only represents the market for 2008-2009, with the recession and everything else.

The expense of a quant person isn't just the income, its benefits, time he takes up, and the marginal cost that they could use on something else. Could a single guy even be worth enough (.5 wins to 1 win) in the average year? Who knows? And could a quant guy also lower costs as well as increase wins? Sure, probably. But I don't think its a slam dunk that a team hires one guy to look at statistics, and they suddenly make millions more dollars per year.
Back to top
View user's profile Send private message
kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC

PostPosted: Tue Oct 19, 2010 11:02 am Post subject: Reply with quote
Here's one of my encounters with an NBA front office. The team expressed interest in my defensive tracking. I proposed to track the 2nd half of the season and provide written reports every 5th game for less than $10k. They told me that they thought the info was valuable and that it would help them, but that it wasn't in their budget (at least not in their front office/scouting budget area).

So, I offered to do it for less. They poor-mouthed again, and then questioned my commitment to breaking into the league when I refused to do it for free. In the following weeks, they went out and signed a guy to two 10-day contracts, and then for the remainder of the season. Including playoffs, the player was on the floor for 145 minutes and totaled 0.1 win shares and a PER below 10.0.

For those contributions, they paid him more than $100k. I think my data would have helped the team win more games, and it would have cost them very little. Oh well.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
schtevie



Joined: 18 Apr 2005
Posts: 414


PostPosted: Tue Oct 19, 2010 12:05 pm Post subject: Reply with quote
gabefarkas wrote:
schtevie wrote:
That this general question is never asked (and those that could answer have zero incentive to volunteer such estimates) contributes to maintaining the basic state of the world we observe, where those who do quantitative analysis for teams - despite the growth in their number over the past decade or so - appear to be distinctly undervalued within the NBA economy.
Typically, an interviewer will come to the table with many more questions prepared than those ending up in the final printed version of the interview. The questions (and subsequent responses) not appearing achieve their status due to one or more of several reasons. Chiefly among these, viz. the lack of incentive to provide an interesting answer, or prior knowledge regarding this.


I infer your conjecture (or perhaps actual knowledge) to be that the line of questioning I suggest is not untypically followed, but the results always end up on the cutting room floor. Hmm. Not sure that makes sense to me, but perhaps that is the existing norm in sports journalism.

If so, here's some solid, free advice to the sporting fourth estate: ask the damn question. Not just of our analyst friends, but also of GMs, coaches, and players. Wouldn't the result of such inquiries be really, really interesting?

"Carmelo, you are reported as being interested in playing for the Knicks. Should you leave the fair confines of the Rockies, what do you think would be the effect on both Denver's and New York's win totals?" And the follow-ups, whether the question is answered or obfuscated, write themselves....
Back to top
View user's profile Send private message
mtamada



Joined: 28 Jan 2005
Posts: 377


PostPosted: Tue Oct 19, 2010 1:58 pm Post subject: Reply with quote
schtevie wrote:
If so, here's some solid, free advice to the sporting fourth estate: ask the damn question. Not just of our analyst friends, but also of GMs, coaches, and players. Wouldn't the result of such inquiries be really, really interesting?

"Carmelo, you are reported as being interested in playing for the Knicks. Should you leave the fair confines of the Rockies, what do you think would be the effect on both Denver's and New York's win totals?" And the follow-ups, whether the question is answered or obfuscated, write themselves....


The results would be TOO interesting. For a couple of reasons. First, there are anti-tampering rules, so that teams can't come along and e.g. induce a player to demand a trade from his current team. Second, almost any answer that Carmelo gave to that Denver/NY question would result in him appearing to insult both teams, resulting in burned bridges in both cities. About the only answer the Carmelo could give that would not result in incendiary blowback would be to give the standard cliche self-effacing answer. That's why we see so many cliche answers in the first place, because otherwise there'd be too much bulletin board material for foes and teammates alike.
Back to top
View user's profile Send private message
schtevie



Joined: 18 Apr 2005
Posts: 414


PostPosted: Tue Oct 19, 2010 2:27 pm Post subject: Reply with quote
Asking any relevant party for an estimate of his own value in terms of his effect on wins (or points scored, allowed, differential, whatever) can easily be formulated in a way (and asked at a time) that it is not disallowed by league rules or directly provides bulletin board fodder. Indeed, what about posing the question to agents who in representing their clients are advocating precisely on these grounds?

I understand why individuals might be uninterested in providing direct answers, but not why the questions aren't publicly asked (or maybe they are?)

Anyway, getting back to the point at hand, I would be quite surprised if the informed belief is that the value of the marginal, competent analyst (from a starting point of none) would be less than one in the win column. And I would be really shocked if that value in terms of revenue - never mind the owner's consumer surplus - were only $270,000. What is the expected profit from just the gate receipts of one playoff home game?
Back to top
View user's profile Send private message
acollard



Joined: 22 Sep 2010
Posts: 56
Location: MA

PostPosted: Tue Oct 19, 2010 3:10 pm Post subject: Reply with quote
I agree my methods probably weren't best but based on these numbers and the wins that each team had in that year, that's the coefficient I come up with. I apologize I'm not familiar with linking to graphs, but I'd show you the distribution, its all over the place.

Here are the numbers I was using for operating income: http://www.forbes.com/lists/2009/32/bas ... ncome.html

I agree that it probably underestimates marginal wins, and obviously that one win that makes a difference from playoffs to staying home matters a lot. The point I was trying to make is that I think we are biased and over-valuing our own craft a bit.

If profitability is the goal, market, marketing, media rights and arena cost/financing seem to be a bigger deal than winning. If winning is the goal, then sure, get some analysts. But if you really think one guy being paid 70k a year will somehow come up with data that is so much more valuable to a team than the data that already is in the public domain or instinctively known by the coach, I disagree. To create a valuable metric, convince a front office of its value, then convince the coach to implement it, and then get the coach to get the players to do it seems much more than trivial, and that entire process seems like it would cost so much more in terms of time and money than 70k.
Back to top
View user's profile Send private message
bchaikin



Joined: 27 Jan 2005
Posts: 690
Location: cleveland, ohio

PostPosted: Tue Oct 19, 2010 4:13 pm Post subject: Reply with quote
But if you really think one guy being paid 70k a year will somehow come up with data that is so much more valuable to a team than the data that already is in the public domain or instinctively known by the coach, I disagree. To create a valuable metric, convince a front office of its value, then convince the coach to implement it...

this was during the 02-03 season:

http://www.nytimes.com/2003/04/27/sport ... -edge.html

and apparently based on what mark cuban has mentioned at the sloan conference the mavs made some lineup decisions based on their data. that was $100,000 for two guys - eight seasons ago...
Back to top
View user's profile Send private message Send e-mail Visit poster's website
gabefarkas



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

PostPosted: Tue Oct 19, 2010 8:12 pm Post subject: Reply with quote
schtevie wrote:

I infer your conjecture (or perhaps actual knowledge) to be that the line of questioning I suggest is not untypically followed, but the results always end up on the cutting room floor. Hmm. Not sure that makes sense to me, but perhaps that is the existing norm in sports journalism.

If so, here's some solid, free advice to the sporting fourth estate: ask the damn question. Not just of our analyst friends, but also of GMs, coaches, and players. Wouldn't the result of such inquiries be really, really interesting?
Perhaps you misunderstood. They may ask the question, and the response may be "well, you know I'm not allowed to answer that."
Back to top
View user's profile Send private message Send e-mail AIM Address
schtevie



Joined: 18 Apr 2005
Posts: 414


PostPosted: Tue Oct 19, 2010 8:50 pm Post subject: Reply with quote
Bob, excellent get! I had forgotten about that article. And it, in combination with the recent remarks at the MIT Sloan Conference the essential points can be inferred.

The most interesting and relevant paragraphs for the topic at hand are found bridging pages 3 and 4:

"Sagarin and Winston have tried to sell their system to other teams besides the Mavericks. But only the Seattle SuperSonics have bought it, and they did so for just one month last season.

''It's a real nice system, it's just real expensive,'' Rich Cho, the Sonics' assistant general manager, said. ''And it's only as good as what you make of it. In general, coaches don't have a lot of time to look at the analysis. They're not going to sit there during a game and look at a spreadsheet. They coach by feel a lot.''

Though I am guessing that Rich Cho doesn't think $100k is too much money right now, consider the context of his remarks, being made at the end of the 2002-03 season. The Sonics finished oh-so-close to the last playoff spot in the West. An improvement of a touch over 1 point per game, over the season, and they would have been expected to match Phoenix in strength. And the minimum of two home play-off games and all related income could have been theirs. Though they certainly had a statistical analyst on staff, could WinVal have put them over the hump?

I am not sure of its capacities back then, but cue the tape from last year's Basketball Analytics Panel to the 19:30 mark. Mark Cuban implicitly asks himself the question "what is the value of an analyst?" (or at least a lower bound in terms of a subset of potential services) in discussing the fact that he can tell when teams are using advanced techniques (in the instance, implicitly referring to what WinVal produces) by the basic line-up errors he observes. And just thereafter, at 19:42, Dean says "he is right about that". Cuban's follow-up quote, regarding the consequences of these errors, is: "...nasty, when it comes to results".

Now, maybe I am reading these emphatic remarks incorrectly, but they suggest to me that the opportunity cost of analytics is greater than the aformentioned 1+ point per game, which relative to a 41 win baseline, is about 4 additional expected wins.

I wonder what Rich Cho's opinions are these days about not having cut a check to Winston and Sagarin.
Back to top
View user's profile Send private message
acollard



Joined: 22 Sep 2010
Posts: 56
Location: MA

PostPosted: Wed Oct 20, 2010 9:33 am Post subject: Reply with quote
Thanks for sharing that article Bob. I didn't mean to get into this big debate, I just think that this board has a lot of like-minded people who obviously value analytics a lot, and it might be worthwhile to think of it from another prospective. Two things I noticed about the article.

By my understanding, Winston and Sagarin were not full time analysts and they did not have exclusivity rights. They shared a database with the Mavs, but could have shared it with other teams, too. If their data was exclusive or they worked full time for the Mavs, I'm guessing it would be more expensive. In a few years their winval calculations (~2005 when adjusted +/- was on places like 82games) were essentially available in different forms for free.

Secondly, Dallas had the second worse operating income in 2009, and has had a negative operating income for 10 of the last 11 seasons. So they don't necessarily make decisions at that franchise based on profits. I don't think a profit is the end all be all of running a sports franchise, but I also don't think that just because Dallas/Cuban paid someone a certain amount of money, that amount is the market value. Front-office people are paid a lot of money to do their job well, and if it was as easy as hiring an analytic guy and making millions of dollars, I think every team would have a department. My guess is a lot of it comes down to institutional philosophy, if your coach is receptive, and if you want to win vs. make money.
Back to top
View user's profile Send private message
kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC

PostPosted: Wed Oct 20, 2010 4:45 pm Post subject: Reply with quote
I ran into the same "the coaches won't use it" comments when I briefly consulted for a team. I sorta understood, because the coaches are incredibly busy. And, there are a lot of voices trying to get in the coach's ear and tell him how to do his job.

But, I found that when I could actually get a coach to engage in the conversation, there was learning on both sides. My stat work often challenged the coach's thinking; and his superior experience and hoops knowledge challenged me to dig deeper into the numbers and work harder in the analysis.

Even fairly basic stat work would contradict what a coach believed about a player. Probably because the coach is human and is as subject to confirmation bias as anyone else.

My experience working with coaches is limited, but I've thought that just having a guy on the staff and in the meetings who can test what the coaches are saying and thinking would have value. Perhaps that's part of Roland's roll with Dallas.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
Marver



Joined: 21 Jun 2009
Posts: 5


PostPosted: Thu Oct 21, 2010 2:21 am Post subject: Reply with quote
If the goal as an analyst is to increase wins (assuming a static player payroll), and solely to increase wins, the sources of that improvement, via analytics, would be one of the following:
- Player acquisitions/roster composition (draft, free agency, trade, etc.)
- Player usage (minutes, lineups)
- Playing emphasis (facets of the game the players should actively stress while on the floor)

With respect to the 'coaches won't use it' comment:
It is very unlikely that a team would hire an analyst only to not value their input regarding the first bullet, roster composition, so I think we can make the fair assumption that the analyst's worth in that area would be fully manifested upon the team. Moving down the list to player usage you reach analyst/coach interaction, though with the proper voices in his ear and sound logic, I would assume the analyst's worth in this area would be at least partially manifested. Finally, you reach an area that I would imagine would be nearly impossible to convince a coach -- any coach -- to use, since it is intrinsically changing how players would move and interact on the floor, ie. how they are coached.

So while I see some limitation to the analyst's worth reaching full manifestation, due to coaching stubbornness, I am somewhat surprised that it would be a factor in hiring an analyst; I believe the overwhelming majority of the analyst's worth is in roster composition.

However, I strongly agree with one of acollard's earlier points about weighing the difficulty of landing (and keeping) an analyst job in the NBA against the subsequent payoff. If you're like me, and I imagine many of you are, jumping through a myriad of hoops to reach an insecure job as another sort of analytic scientist, with the same pay, seems like a strange exercise.
Back to top
View user's profile Send private message
acollard



Joined: 22 Sep 2010
Posts: 56
Location: MA

PostPosted: Thu Oct 21, 2010 9:36 am Post subject: Reply with quote
Marver, this comes back to the main problem I have with all this. If hiring an analyst were such a surefire way to increase wins, and coaches/front offices were receptive to the input of an analyst, why don't all teams have analysts working for them? Its a very competitive environment, with a lot of smart people. I just figure there has to be a logical reason about half the teams don't have one (16/30 based on the teams with an analytical staff thread, if you're including consultants).
Back to top
View user's profile Send private message
Ryan J. Parker



Joined: 23 Mar 2007
Posts: 711
Location: Raleigh, NC

PostPosted: Thu Oct 21, 2010 10:36 am Post subject: Reply with quote
Here's another way to pose the problem: are there actually 30+ quality analysts hanging around that can do quality work for an NBA team? Like basketball players, analysts have varying skill sets with their own strengths and weaknesses.
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website
kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC

PostPosted: Thu Oct 21, 2010 10:56 am Post subject: Reply with quote
acollard wrote:
Marver, this comes back to the main problem I have with all this. If hiring an analyst were such a surefire way to increase wins, and coaches/front offices were receptive to the input of an analyst, why don't all teams have analysts working for them? Its a very competitive environment, with a lot of smart people. I just figure there has to be a logical reason about half the teams don't have one (16/30 based on the teams with an analytical staff thread, if you're including consultants).


Earlier this week, I watched the latest Frontline (PBS news show) about a man who'd been executed for setting a fire that killed his three small children. According to the original arson investigator, there were 20 indicators of arson.

Except, since the man's conviction (but before he was executed) researchers delved into indicators for started fires vs. accidental fires. The research and analysis of the evidence gathered by arson investigators in this case showed that all 20 of those indicators were wrong.

A comment from one of the researchers jumped out at me. He said that most arson investigators are firemen, and that the mantra among investigators is that you have "get in there and feel the beast." An attitude he said embarrassed him because of its lack of rigor. Feeling the beast, having hunches, and using experience has led to arson convictions, imprisonments and executions when the accused is actually innocent.

Why aren't arson investigations conducted based on the solid science that exists? Because of organizational inertia. Because the investigators "know" what to look for. In some cases, probably because some of that science contradicts what these investigators have been trained to look for.

I know in my case, one coach rejected the defensive tracking data I'd collected because a guy they'd been publicly praising as a defender showed up as a poor defender in my stuff. Maybe my stuff was wrong. But, maybe my stuff was picking up something the coaches were missing. (As it turns out, when I got one of the assistants to sit down and look at film with me (after the season), he ended up agreeing with my analysis.)

I don't think anyone is claiming that having an analyst on staff guarantees that team will improve. But, having an analyst might help guard against confirmation bias. In my case, the data contradicted what the coaches were thinking. Why? Well, the guy in question was a veteran and a vocal leader. He talked a good game and he played with effort. He just wasn't very effective in certain situations, and it was costing the team significantly when he was on the floor. That information might have led the coaches to adjust playing rotations or strategies and may have helped them win more games.

acollard



Joined: 22 Sep 2010
Posts: 56
Location: MA

PostPosted: Thu Oct 21, 2010 1:34 pm Post subject: Reply with quote
Thanks to everyone! I really enjoy statistics and obviously think they are of interest and valuable. I didn't mean to be relentlessly argumentitive or to "troll". I enjoyed hearing the experiences of people who have actually worked for a team and the economics of the analytics is also really interesting to me.
Back to top
View user's profile Send private message
schtevie



Joined: 18 Apr 2005
Posts: 408


PostPosted: Thu Oct 21, 2010 2:18 pm Post subject: Reply with quote
A particular point, followed by a bemused rant...

acollard, I really don't think that looking at operating income is in any way useful for determining the value of a marginal win. There is a basic, first cut at reality, and it is quite simple: you win, you get a big playoff payout.

NBA franchises are in many businesses (selling tickets, TV shows, weak beer, T-shirts, etc.), are separately and distinctly endowed with revenue-generating potential (different sized population bases with differing historic interest in professional basketball), and have different abilities to disguise the profitability of their basketball operations (via co-ownership of businesses that provide goods and services to the franchise) which they all have a strong incentive to do. And on top of that, many owners want to win "irrationally" and pour in extra money on productive inputs, beyond their marginal value. But apparently not, for some reason, with analtyics. And the question there is: why?

I think the answer is twofold. First, the competitive forces that one might think would lead to a strong demand for such services is, in fact, quite weak - certainly as a historical matter. And second, well, the movie title says it quite well: He's Just Not That Into You.

You see the NBA product on the court - players sweating, straining, cursing (well, not anymore), coaches yelling - and you get the strong and correct impression that teams really want to win. But such conspicuous, on-court desire for victory has never translated into a strongly competitive, off-court environment for improving team coaching and personnel management.

As a historical matter, can anyone identify a competitive innovation that a coach or GM has discovered or embraced that has led to any discernible professional reward? Were there relegation, like in foreign football leagues, there might be some institutional inducement, but in the NBA such forces are extremely weak.

As evidence for their absence, cue the MIT videotape and listen to the various recommendations for breaking into the business as an analyst: work for free and learn how to speak non-geekily to the coaches. Know your place! But if the competitive environment were such that victories were highly prized, the advice would be the reverse and wouldn't even need to be explicitly offered to the opposite parties. Coaches would find it in their personal interest to learn to speak Geek and GMs would be actively seeking out and opening up their checkbook for the most promising analytic talent.

And another piece of evidence as to the historic, relative indifference of members of the NBA coaching guild toward acquiring a competitive, on-court advantage, this posed in the form of a question: Why is Louis Dampier not a household name?

Apologies to those who have read my argument more than once before, but what else can explain the excruciatingly slow adoption of the 3 point shot in professional basketball? From its first days more than 40 years ago, the competitive benefits, actually realized and potentially imagined, were there for anyone and everyone who could multiply by 1.5 and was savvy enough to realize that mid-range shots were the worst. And Pat Riley was his college teammate!

If perfecting average practice has always been solid career advice for aspirants to front office and coaching jobs, and that was all that was surpressing innovation, that would be one thing. But there is a more obvious impediment to innovation than indifference, and that is hostility. If coaches and front office personnel are working in an environment where, if push came to shove, they would find it very difficult to justify their own employment and high salaries, why would they countenance or encourage the hiring of others who might help to identify their shortcomings to their employers?

Wouldn't it be amusing, were Donald Sterling to offer analyst's terms, the next time he is in the market for a coach? I can see it now. He's in LA. He surely knows the right people. Perhaps a reality show: America's - scratch that - the World's Got Coaching Talent. The promise would be zero salary, but the opportunity of a lifetime. I bet he could get a pretty darn good hire, and almost certainly he would increase his operating income.
Back to top
View user's profile Send private message
Marver



Joined: 21 Jun 2009
Posts: 5


PostPosted: Sat Oct 23, 2010 6:32 pm Post subject: Reply with quote
acollard wrote:
Marver, this comes back to the main problem I have with all this. If hiring an analyst were such a surefire way to increase wins, and coaches/front offices were receptive to the input of an analyst, why don't all teams have analysts working for them? Its a very competitive environment, with a lot of smart people. I just figure there has to be a logical reason about half the teams don't have one (16/30 based on the teams with an analytical staff thread, if you're including consultants).

Your argument is based on the assumption that all NBA front offices make near-perfect logical decisions a near-perfect amount of the time. That's a pretty far-fetched assumption.

schtevie wrote:
If perfecting average practice has always been solid career advice for aspirants to front office and coaching jobs, and that was all that was surpressing innovation, that would be one thing. But there is a more obvious impediment to innovation than indifference, and that is hostility. If coaches and front office personnel are working in an environment where, if push came to shove, they would find it very difficult to justify their own employment and high salaries, why would they countenance or encourage the hiring of others who might help to identify their shortcomings to their employers?

To which I reply:
Thomas Paine wrote:
...a long habit of not thinking a thing wrong, gives it a superficial appearance of being right, and raises at first a formidable outcry in defense of custom. But the tumult soon subsides. Time makes more converts than reason.

[Crow]

Doc319
Guest






PostPosted: Sat Feb 26, 2005 5:39 am Post subject: Kobe Bryant--Superstar or Selfish Reply with quote
The debate about whether Kobe Bryant is a legit superstar or a
talented but selfish player inspires very strong reactions in
support of both contentions. I have just published an article on
this issue at Hoopshype.com that includes quotes from NBA TV analyst
Fred Carter; I also cite Bob Chaikin's scoring field goal percentage
statistic, which Chaikin himself mentioned recently as an indicator that Bryant is more efficient as a shooter than
his 41% field goal percentage suggests.

The article can be found here:
http://hoopshype.com/columns/kobe_friedman.htm
Back to top

Mike G



Joined: 14 Jan 2005
Posts: 3547
Location: Hendersonville, NC

PostPosted: Sat Feb 26, 2005 7:43 am Post subject: Reply with quote
While Kobe was out (for 14 G), it seemed the other Lakers' effective shooting went down. Before, Kobe was the lowest of the top 6 guns; now he's about in the middle. This might be interpreted as : Kobe sets up the other guys for high-% opportunities.

Checking 82games.com, it appears everyone on the team does better with Kobe on the floor (excepting Walton):

http://82games.com/0405LALP.HTM

But 2 other guys have similarly positive net effects: Brian Cook and Jumaine Jones. These guys play 18 and 23 MPG, respectively; Cook has played about 55% of his minutes with Kobe, while Jones has only played 431 of his 1023+ with Kobe. When Kobe was out, Jones' minutes soared and he became a starter.

Jones is the highest eff% and lowest-volume shooter (more than half his pts are from 3) of the top 6 Lakers; yet he seems to be Kobe's de facto Replacement. And, he seems to have the same effect on his teammates.

Jumaine is no Kobe, but he'll do in a pinch?
Back to top
View user's profile Send private message Send e-mail
bchaikin



Joined: 27 Jan 2005
Posts: 681
Location: cleveland, ohio

PostPosted: Tue Apr 19, 2005 12:33 pm Post subject: Reply with quote
now that the season is just about over, just wanted to see what any kobe-bashers had to say. yes the lakers aren't going to make the playoffs, but they lost the best C in the league in shaq and replaced him with chris mihm...

as for kobe? many will cite his FG% of .431 as being his worst in seven seasons, but the reality is that his Scoring FG% (combining 2pters, 3pters, and FTs), despite the absence of shaq, is at a career high for him at .549, due to him taking more 3's than ever before - almost 6 per game - and hitting 34/35% of them, and getting to the line for a career best 10 FTA/g, and hitting 80+% of those. that Scoring FG% is also higher than the Scoring FG%s this season of ray allen, vince carter, michael redd, and jason richardson - as a matter of fact, its the highest Scoring FG% of any SG scoring at least 20 pts/g this season...

others will cite his poorer defense, but my numbers show me he is still a good defender (just not the best defending SG). his rebounding of 7reb/48min is similar to what it's been for most of the past 5 seasons, his shot blocking is as good as it's been the past 5 seasons, and although he's committing 4.0 TO/g, which is high, his touches/min are high (1.6 to 1.7 touches/min) and his rate of 6% turnovers (6 TO per 100 touches) is similar to that of other SGs such as manu ginobili, mike miller, and jerry stackhouse...

he also has passed the ball - per touch - at an identical rate to each of the past three seasons, with 50/51% of his touches...

the only kink i see in his numbers is a career low in steal rate...
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Mike G



Joined: 14 Jan 2005
Posts: 3547
Location: Hendersonville, NC

PostPosted: Tue Apr 19, 2005 1:07 pm Post subject: Reply with quote
bchaikin wrote:
now that the season is just about over, just wanted to see what any kobe-bashers had to say. yes the lakers aren't going to make the playoffs, but they lost the best C in the league in shaq and replaced him with chris mihm...



Actually, Shaq was traded for 3 guys who together have averaged 44.4 pts, 18.5 reb, and 10 ast. Granted, they needed 107 minutes to do these things. But 85% of the Lakers' non-Kobe minutes are from guys they didn't have last year.

Kobe's TS% (the term most of us here have agreed to use) is the most remarkably consistent stat I've seen from any player: between .536 and .549 his whole career; and .559 this year.

I wouldn't say a 34% rate from the arc is responsible for an improvement in TS% -- that's only 51% TS in itself. Rather, it's due to more FTA this year.

Kobe's career seems to be in a plateau phase over the last 5 years, with the middle of that span (02-03) being a distinct peak. Perhaps if the Lakers' situation stabilizes, he may continue what had been an upward trajectory.

But he'll be 3 years removed from that peak. And in the last 6 years, he's only played 70 games twice.
Back to top
View user's profile Send private message Send e-mail
Kurt



Joined: 10 Jan 2005
Posts: 30
Location: Los Angeles

PostPosted: Tue Apr 19, 2005 7:15 pm Post subject: Reply with quote
Mike G wrote:
Kobe's career seems to be in a plateau phase over the last 5 years, with the middle of that span (02-03) being a distinct peak. Perhaps if the Lakers' situation stabilizes, he may continue what had been an upward trajectory.

But he'll be 3 years removed from that peak. And in the last 6 years, he's only played 70 games twice.


We're still talking about a 27-year-old here, even if this is his ninth season. I find it hard to think he could not reach that peak again or go higher with the right system (as opposed to the multiple and questionable offensive systems used by the Lakers this year) and with better support around him.

I know several people on this site (I'm thinking Ed was one) who had done work about the general age players at specific positions started to see rapid declines in performancy (they went over the top of the curve and started to fall fast). I'm curious if those are more tied to years in the league and the physical wear that happens because of that or just age of the player.
Back to top
View user's profile Send private message Visit poster's website AIM Address
Greg D



Joined: 30 Dec 2004
Posts: 8


PostPosted: Wed Apr 20, 2005 12:39 am Post subject: Kobe: Superstar or Selfish Reply with quote
Statistically, I admit Kobe is an effective player. But he appears to be a poor leader. I've seen him double and tripled teamed and refuse to give up the ball, take crazy turnaround jumpers, decide not to shoot for a half, to shoot left-handed (ok, it went in) and do other crazy stuff that pissed of PJ, Shaq and his other teammates. This leads to dissension and resentment, hallmarks of the Lakers last year and this year. You don't get your team to make the max effort on D and to do other things that don't necessarily show up in the glory stats when your best player does not play team ball consistently, pouts when others don't pass him the ball or when they miss shots. Like AI, I cannot imagine there are other good or great players who want to play second banana to Kobe. At least for more than one season. Showing up your HOF coach (whether it be PJ or Larry Brown) is no way to lead either.

So like AI, I think Kobe's teams in the future will be doomed unless he matures or management finds just the right kind of player (Kirilenko? Battier? Dampier?) to put up with Kobe's ego driven play. It's not Lamar Odom, that's for sure. But if Antoine Walker can reform himself (a debate for another day) I guess there's hope.

Yes, Kobe was a major factor in three championships. He has great skills, and as Bob C points out, good to great stats. But my limited observation of his play leads me to believe that he is selfish offensively all too often and that this will impair management's ability to surroind him with better players.

I know this is a stat-based site but I do not think every argument of this type can be captured in stats.
Back to top
View user's profile Send private message
HoopStudies



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

PostPosted: Wed Apr 20, 2005 4:35 pm Post subject: Re: Kobe: Superstar or Selfish Reply with quote
Greg D wrote:
Statistically, I admit Kobe is an effective player. But he appears to be a poor leader. I've seen him double and tripled teamed and refuse to give up the ball, take crazy turnaround jumpers, decide not to shoot for a half, to shoot left-handed (ok, it went in) and do other crazy stuff that pissed of PJ, Shaq and his other teammates. This leads to dissension and resentment, hallmarks of the Lakers last year and this year. You don't get your team to make the max effort on D and to do other things that don't necessarily show up in the glory stats when your best player does not play team ball consistently, pouts when others don't pass him the ball or when they miss shots. Like AI, I cannot imagine there are other good or great players who want to play second banana to Kobe. At least for more than one season. Showing up your HOF coach (whether it be PJ or Larry Brown) is no way to lead either.

So like AI, I think Kobe's teams in the future will be doomed unless he matures or management finds just the right kind of player (Kirilenko? Battier? Dampier?) to put up with Kobe's ego driven play. It's not Lamar Odom, that's for sure. But if Antoine Walker can reform himself (a debate for another day) I guess there's hope.

Yes, Kobe was a major factor in three championships. He has great skills, and as Bob C points out, good to great stats. But my limited observation of his play leads me to believe that he is selfish offensively all too often and that this will impair management's ability to surroind him with better players.

I know this is a stat-based site but I do not think every argument of this type can be captured in stats.


I should mention that Basketball on Paper has a chapter that discusses measuring leadership. Or at least measuring the ability of the de facto leaders of teams. With Kobe this year as leader of the Lakers, that measure says that he is only about average. The measure often shows that the leader of a team defines his team's record. Given that the Lakers are mediocre, it makes sense that Kobe's measure is only mediocre.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Doc319
Guest






PostPosted: Wed Apr 20, 2005 5:56 pm Post subject: Leadership, limitations of statistical analysis Reply with quote
Dean, I take your post to mean that you believe that the two questions that I asked of Bob Chaikin in the Antoine Walker topic--"Do you buy the idea that Walker's leadership/presence may have some value that is not measurable statistically? Or is it your contention that each NBA player is no more and no less than the sum of his simulated statistics?"--are valid issues to be discussed in a statistics forum. Perhaps you would reword my first question to say "not easily measurable statistically." Please indicate if this is a fair appraisal of your views or not.

My point is that Bob stated that Antoine Walker's future dropoff in production is a "fact" but without stating how many games it will take for Walker's numbers to do so and without indicating how (or if) his sim accounts for leadership (and other "intangibles") that many observers consider to be important or at least relevant to evaluating basketball players. This makes his prediction impossible to prove or disprove, which places it in the realm of speculation, as opposed to scientific theory. I'm not trying to discredit his sim or statistical analysis in general; I'm saying that if his sim is scientific then it must conform to the scientific method, which means proposing a specific, testable hypothesis and having the willingness to modify the hypothesis if new evidence is found--in the specific instance of Walker, at some point the sample size of games is not small and a scientist would be willing to adjust his hypothesis if Walker maintains a higher than predicted level of performance. DanR and Dean do this all the time, referring to "noise" and indicating that there is a higher degree of certainty with some stats than with others. Acknowledging the possibility of error and a certain degree of uncertainty would not be a weakness in the sim or its creator--it would be a strength. The Titanic's engineers were 100% certain that they had crafted an unsinkable shift; I prefer my science and engineering with a degree of skepticism and uncertainty.

I have no personal stake in how Walker performs or whether or not Bob's sim is accurate or scientific. I'm not sure what Walker's career suggests about statistical analysis or "leadership." All I am doing--with the Kobe thread and with the Walker thread--is raising questions that I think are worthy of statistical evaluation in order to learn what statistical analysts have observed about these issues. The answer that Walker is "playing for a contract" (one of Bob's early responses) strikes me as practicing psychiatry without a license and the statement that his future production is a "fact" seems to be much bolder than the accepted norm for scientists and researchers. Maybe his future production is 75% certain (or some other percentage, if this can be supported by numerical evidence) but a "fact"? I guess we no longer need to play the games--we can just put the numbers in a sim and post the results in the newspaper.

For Bob to say that I should not question his sim's ability to measure leadership because I cannot do so myself frankly misses the entire point. I am not asserting that I can predict Walker's statistical future so precisely that it is an absolute fact (if that were possible one should do this with the stock market...), so the onus is not on me to demonstrate anything. I am asking questions to determine what is and is not being measured. If leadership (and other "intangibles" like setting screens, "help" defense, etc.) is considered irrelevent, then simply say so and others can decide for themselves if that makes the sim faulty or not (I never said that it does). If leadership is relevant but not measurable, then some kind of margin of error should be included in the analysis--Einstein kept including and removing the so-called "cosmological constant" in one of his formulas; maybe there is a way, as Dean suggests, to account for leadership in some way.

In any case, as DanR mentioned in the Walker thread and Dean points out here, these are indeed relevant questions to consider in a statistical analysis group.

I guess I could have posted this in the Walker thread, but since the "leadership" issue was raised here, with Bob referring to me indirectly, I felt that this was the correct thread in which to place this reply.

For what it's worth, Bob and I agree about Kobe and I cited his shooting percentage numbers in my article about Bryant. Despite the occasional (or not so occasional) venturing into vitriol and hyperbole, I find both this thread and the Walker thread very interesting and I am glad that I started both of these threads in this forum. Bryant and Walker are two players who we will be able to follow for the next several seasons and it will be interesting to see what happens in both of their careers.

--David Friedman
Contributing Editor, Basketball Spotlight
www.suite101.com/welcome.cfm/basketball_spotlight
Back to top

HoopStudies



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

PostPosted: Wed Apr 20, 2005 6:38 pm Post subject: Re: Leadership Reply with quote
Doc319 wrote:
Dean, I take your post to mean that you believe that the two questions that I asked of Bob Chaikin in the Antoine Walker topic--"Do you buy the idea that Walker's leadership/presence may have some value that is not measurable statistically? Or is it your contention that each NBA player is no more and no less than the sum of his simulated statistics?"--are valid issues to be discussed in a statistics forum. Perhaps you would reword my first question to say "not easily measurable statistically." Please indicate if this is a fair appraisal of your views or not.


In Basketball on Paper, I stated that measurement of the effect of leadership was pretty soft science. I still believe that. I think it is fair to say that without evidence that it matters, you don't have to consider it most applications. However, that doesn't mean that we shouldn't talk about measuring it in a scientific way.

With regard to Walker, I would say that he was more a designated leader in Atlanta than he is in Boston, where Paul Pierce is alpha male. My concept of a leader is the person who does what it takes on his own and with his teammates to achieve the goal: winning. If we're saying leadership is something else, what is it? I don't think Walker has the role of doing what it takes in Boston to win. That is Pierce's role. Walker may have some wisdom of taking the right shot, uhh, maybe not. Maybe wisdom on defense that he can impart. Walker has been on teams with success and even surprising success with those Celtic teams. So maybe there is something there. But it has to be measurable somehow. What wins is performance that is fully measurable. If we can narrow down 95% of that performance to other things and 5% is left unexplained, maybe that is Walker's indirect (not unmeasurable) influence. That is the most it can be, at least.

Doc319 wrote:
My point is that Bob stated that Antoine Walker's future dropoff in production is a "fact" but without stating how many games it will take for his numbers to do so...


I do think that this is a good subject in general, well beyond Walker. How long does it take for numbers to stabilize, regress to the mean, whatever you want to call it.

Doc319 wrote:
Maybe his future prodution is 75% certain (or some other percentage, if this can be documented by numerical evidence)


I think you can say that a player is going to produce below or above a certain level with 75% or 95% certainty right now. Over the course of a season. But there is a lot of sampling variation between games.

Doc319 wrote:
maybe there is a way, as Dean suggests, to account for leadership in some way.


My old company had this "leadership training" for all the managers. It was this whole day of management and teamwork exercises. It was interesting, kinda fun, and I didn't drink the buttermilk (for those that know about this). But it did seem a bit silly to have so many people in the training. I will back up from that in a second, but, sheesh, 80 people all training to be leaders of the company? Some of those people clearly had no real leadership ability in the way that we think about it. Of course, the class had a way of sorting people into 4 different types of leaders, one or two of which had people that we wouldn't typically think of as leaders. Here is where I back up -- yes, those people were in a position of leadership, managing people, so they better have some idea of how to do it. But only a few people in that room had the stuff to, say, run a company.

I'll stop being a bastard now.

My point with that diatribe is that some teams don't need leadership and really shouldn't consider it at all. I don't think Boston needed leadership. They might have considered it with Walker. I really don't know. As much as anything, Boston got rid of dead weight (guys getting no minutes), got Gary back for nothing, and could move forward with one extra option they didn't have before.

Anyway, read Chapter 20 in Basketball on Paper. It talks about measuring leadership. I felt it was the most controversial stuff I put out (in that my support for it was less than elsewhere)...
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Doc319
Guest






PostPosted: Wed Apr 20, 2005 7:07 pm Post subject: Measuring leadership, Roboscout Reply with quote
Dean,

I just went back to Chapter 20 to refresh my memory about what you did regarding leadership. Do you have this year's numbers handy with regard to Kobe and Antoine? Specifically, based on the composition of their teams (the actual games played, discounting players like Divac and George who missed almost all of the season) how do Kobe's and Antoine's W-L records compare to their teams and what does this suggest about their "leadership"? Realizing that you do consider this "soft science," I am still interested if your numbers suggest that Kobe did not "lead" well or if he simply did not have enough talent around him to win consistently (particularly when factoring in the games that he and Odom missed).

What do you think of the contention that the Shaq-Kobe Lakers were built around Shaq and the half court game (lots of three point shooters), so that when Shaq was out the Lakers were not equipped to play the up-tempo game that is best suited to Kobe's skills. Therefore, Kobe does not showup as the "leader" and he indeed was not the leader of those teams, but that with the right personnel around him (a fast breaking team) he would show up as the "leader" by the metric that we are discussing. As a corollary to this idea, maybe Kobe does not show up as a "leader" this year because this Laker roster is still not constructed to properly take advantage of what he does best. Who shows up by your measurement as the Suns' leader--Nash, Amare or somebody else? Is Nash more of a "leader" this year than he was with the Mavs? I interviewed Del Harris recently and asked him why Nash and Nowitzki, who seemed to be such a perfect duo, are each having career years playing apart. His basic reply was that the Phoenix roster is a running squad that best suits Nash's skills, while this year Dallas is a deeper, less predictable team so that teams cannot load up on one or two plays (pick and roll, etc.) to stop Dirk. Here is a link to the interview for those who are interested: http://www.suite101.com/article.cfm/bas ... ght/115277

This season which players show up as the "best" leaders by the leadership metric you used in Basketball on Paper? Again, I understand that you consider this to be "soft" and inexact; I'm just interested what the numbers show to this point. Your discussion of D-Rob in Basketball on Paper is very interesting.

Also, on a different subject, last year you used Roboscout to predict a Pistons' triumph in the Finals. I know that you are working for the Sonics now and are probably limited in what you can say publicly, but can you share any Roboscout insights/predictions about this year's playoffs?

Thank you for your contributions to this field and to this forum. Specifically, I appreciate your thoughtful replies to this thread.


--David Friedman
Contributing Editor, Basketball Spotlight
www.suite101.com/welcome.cfm/basketball_spotlight

Last edited by Doc319 on Wed Apr 20, 2005 7:52 pm; edited 2 times in total
Back to top

HoopStudies



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

PostPosted: Wed Apr 20, 2005 7:18 pm Post subject: Re: Measuring leadership, Roboscout Reply with quote
Doc319 wrote:
Dean,

I just went back to Chapter 20 to refresh my memory about what you did regarding leadership. Do you have this year's numbers handy with regard to Kobe and Antoine? Specifically, based on the composition of their teams (the actual games played, discounting players like Divac and George who missed almost all of the season) how do Kobe's and Antoine's W-L records compare to their teams and what does this sugges about their "leadership"? Realizing that you do consider this "soft science," I am still interested if your numbers suggest that Kobe did not "lead" well or if he simply did not have enough talent around him to win consistently (particularly when factoring in the games that he and Odom missed). This season which players show up as the "best" leaders by this metric? Again, I understand that you considered this to be "soft" and inexact; I'm just interested what the numbers show. Your discussion of D-Rob in Basketball on Paper is very interesting.


Kobe 31-34
Walker 14-39 with Atlanta in a leadership role
Walker 7-15 with Boston in a nonleadership role

I won't run the full leadership thing until later (it does more than individual game-by-game win-loss records). I've got a toy to do some analysis with this but I've not seen a lot of value to it yet, so I haven't refined it to do things easily.

Doc319 wrote:

Also, last year you used Roboscout to predict a Pistons' triumph in the Finals. I know that you are working for the Sonics now and are probably limited in what you can say publicly, but can you share any Roboscout insights/predictions about this year's playoffs?


Nope, can't say anything. It's getting used pretty heavily with a lot of new stuff. We've been using it all year to some degree in Seattle. I've got some experimental stuff that I'm still debating whether to introduce to the staff for the playoffs...
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Doc319
Guest






PostPosted: Wed Apr 20, 2005 7:35 pm Post subject: Reply with quote
Dean,

Maybe there is an obvious answer to this, but how does Walker have a losing record this year with Boston?

Also, can you shed any light on the Nash/Nowitzki issue or are those among the numbers that you have not crunched yet?

Finally, regarding your "alpha male" concept of leadership, there is a certain logic to this but two questions occur to me: (1) In the ESPN the Magazine article about Walker it is asserted that Pierce was not as successful in this "alpha male" role without Walker's presence on the team. Is there a way to quantify how Walker's ability to set screens, feed the post, relieve Pierce of ball-handling responsibilities, etc. helps Pierce or are all of these effects, if they are significant, going to appear in the "measurables" that you, Bob Chaikin and others are already tracking? (2) There have been "role players" who are considered to be important leaders on their teams even though they were clearly not the "alpha male." I am thinking here specifically of the Bulls' Ron Harper and the Sonics' Nate McMillan. Is the type of "leadership" that these players provided statistically significant in your opinion and is there any way to track it? Does it show up in the way that you chart teams' w-l records vs. individual w-l records?

--David Friedman
Contributing Editor, Basketball Spotlight
www.suite101.com/welcome.cfm/basketball_spotlight
Back to top

HoopStudies



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

PostPosted: Wed Apr 20, 2005 8:20 pm Post subject: Reply with quote
Doc319 wrote:
Dean,

Maybe there is an obvious answer to this, but how does Walker have a losing record this year with Boston?


Frankly, by the numbers in BoP, his just aren't very good even in Boston.

Doc319 wrote:

Also, can you shed any light on the Nash/Nowitzki issue or are those among the numbers that you have not crunched yet?


Not even sure what the issue is.

Doc319 wrote:

Finally, regarding your "alpha male" concept of leadership, there is a certain logic to this but two questions occur to me: (1) In the ESPN the Magazine article about Walker it is asserted that Pierce was not as successful in this "alpha male" role without Walker's presence on the team. Is there a way to quantify how Walker's ability to set screens, feed the post, relieve Pierce of ball-handling responsibilities, etc. helps Pierce or are all of these effects, if they are significant, going to appear in the "measurables" that you, Bob Chaikin and others are already tracking? (2) There have been "role players" who are considered to be important leaders on their teams even though they were clearly not the "alpha male." I am thinking here specifically of the Bulls' Ron Harper and the Sonics' Nate McMillan. Is the type of "leadership" that these players provided statistically significant in your opinion and is there any way to track it? Does it show up in the way that you chart teams' w-l records vs. individual w-l records?


Leadership from someone who is not your best player just doesn't seem to help much. It usually means that they are good role players who can't carry a team. Were people talking about Harper's leadership of the horriBulls in 1999?

Obviously, it is soft science. And, frankly, I've always been more of the Nate McMillan player because I can't shoot but I definitely know how to complement good players. Maybe I can help a horrible team get better, but I'd also do it by trying to identify our truly best player (who should be better than I) and have them show the actions of a leader.

I'd love to do more on the subject. It's very interesting. Just not high enough on the priority list to get my time now.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Doc319
Guest






PostPosted: Thu Apr 21, 2005 4:53 am Post subject: Reply with quote
Dean, the Nash/Nowitzki issue is not a big deal--I was just wondering, according to your "alpha male" determination of who the team leader is, how you would categorize these two players last year and this year. Which one was Dallas' leader when they were teammates? Is Nash the "alpha male" on this year's Phoenix team or is it Amare or somebody else? Again, we're talking about your stat of comparing individual wins with team wins and your idea that the individual on the team whose wins most correlate with the team wins is the team leader (I hope that I did not mangle that in my attempt to restate it...)

The way this all relates to the Kobe thread is that I was wondering if maybe this "alpha male" trait is at least somewhat dependent on who one's teammates are. That gets back to Del Harris' quote to the effect that Nash is thriving in Phoenix because he has players who can run the floor very well. I'm wondering if your leadership numbers could shed some light on this--my theory is that if this is true, that maybe Nowitzki shows up as the "alpha male" when he and Nash played together but that Nash shows up as Phoenix's "alpha male" because the style of game suits him even better than Dallas' did. As for Kobe, to get back to your BoP example, maybe he did not show up as the Lakers' "alpha male" when he played with Shaq in part because the roster was filled with players who thrived in the half court set, making threes after Shaq is doubleteamed; I think that Kobe's skills are better suited to an open court game in which he can attack before the defense is set. That seemed to be the source of at least some of the friction with Shaq--Kobe wanting to run and Shaq taking his time to come down court and secure post position. I realize that some of this science is a little "soft," as you say, and that there are some oversimplifications here--Shaq sometimes runs the floor and Kobe can also score (and distribute) in a half court situation.

As for Harper's leadership on what you aptly termed the Horribulls, is their record a reflection on his leadership? To take an extreme example to make the point, if Harper is the greatest leader in the world he still could not lead four guys from the Y to victory against even the worst NBA team. At some point talent--or lack thereof--supercedes leadership. However, when the requisite talent to win is present it still may require the right leadership to get over the top. Portland seemed to have enough talent to win a few years back, but couldn't quite get it done. Harper could not rescue the Horribulls and Walker could not save the Hawks, but Harper seemed to have an importance on the Bulls that was greater than his measurable stats (the games the Bulls lost in the '96 Finals were when he did not play); the anecdotal evidence about Walker this year is that he is having that kind of effect in Boston. Watching Kobe this year it seemed that a lot of his dribble penetrations ended in missed jump shots or dropped passes; he didn't have a dead-eye shooter or a great finisher to convert his passes into field goals.

--David Friedman
Contributing Editor, Basketball Spotlight
www.suite101.com/welcome.cfm/basketball_spotlight
Back to top

mtamada



Joined: 28 Jan 2005
Posts: 376


PostPosted: Fri Apr 22, 2005 1:19 am Post subject: Reply with quote
Doc319 wrote:
As for Harper's leadership on what you aptly termed the Horribulls, is their record a reflection on his leadership?

[...]

Harper could not rescue the Horribulls and Walker could not save the Hawks, but Harper seemed to have an importance on the Bulls that was greater than his measurable stats (the games the Bulls lost in the '96 Finals were when he did not play); the anecdotal evidence about Walker this year is that he is having that kind of effect in Boston.


I think this thread is trying to use the word "leadership" in two different ways, leading to some confusion and ambiguity.

I don't recall people talking about Harper's "leadership" on the Bulls, but they certainly did talk about his (non-statistically measured) importance to them, mainly defensively, with the key example being exactly the one that Doc319 mentions, the Bulls' poorer performance in the 1996 Finals when Harper was injured. It was parallel to Nate McMillan's role with the Sonics in 1996; the Sonics beat the Bulls in the games in which McMillan was able to play significant minutes (the key element however being McMillan's playmaking rather than his defense, important though that was). And I don't recall people calling McMillan a "leader" on that team -- but George Karl did call him the "glue" which held the Sonics together.

IOW, I don't think role players such as Harper and McMillan can be called "leaders", at least not in the sense of a Jordan or Magic. What they are providing should be called something else (I'm not sure what term we should use; maybe "glue" is as good as any. Or "role-playing leadership".)

An example of the difference between "leadership" and "glue": one would never expect a Harper or a McMillan to be able to transform a roster such as the HorriBulls of the late 1990s. Role players can't do that.

But franchise players (especially those who are "leaders") can. Jordan, Kemp+Payton, maybe Pippen. They might not be able to do so singlehandedly (even Kobe clearly needs help to take the Lakers into the playoffs), but they are the kinds of players who you DO expect to have a big impact on such rosters. Especially the franchise players who are expected to provide "leadership". Jordan arrives in 1984 and the Bulls win 11 more games and make the playoffs for the first time in x years. Bird arrives and the Celtics win 32 more games and advance to the Eastern Conf finals. Etc.

Whereas the role-playing leaders such as Harper and McMillan cannot have such a transformative effect on a lousy roster. But they can make a good roster into a great one; with Harper on the 1996 Bulls were a 72-win team; with Harper injured, the Bulls couldn't even beat the Sonics. Conversely, with McMillan on, the Sonics could beat the 1996 Bulls; without him the Bulls could smother the Sonics. (Obviously, those wins and losses can't be ascribed just to one player -- but maybe they can be ascribed to two players, namely Harper and McMillan. Plus Payton being able to contribute a lot more in the last 3 games of the series.)

Walker's leadership, if indeed he's providing it to the Celtics, is in between the "franchise" level and the "role player" level, because he's clearly not a franchise player but more than a role player for Boston.

Doc319 mentioned an espn.com column; it may've been the same one that I read, by Bill Simmons, who said that when Walker's on the team, the Celtics, Paul Pierce in particular, play harder. If true, that'd certainly be a form of leadership -- not the franchise player kind of leadership, where Walker transforms the Hawks into a decent team (obviously he didn't, and can't, do that). But it could be something like the role-player, glue kind of leadership (except that he's more than role-player).

As I mentioned in another thread, since Walker rejoined the Celtics, if anything Pierce's rebounding seems down a bit (but Walker could be grabbing some of the rebounds which previously went to Pierce) and his scoring hasn't changed much. But Pierce's TS% is way up, through February (roughly corresponding to the non-Walker months), Pierce was at .630; in March and April he's been at .683.

Is that due to Walker? Who knows, this is where someone who's closer to the team or who sees their games firsthand is needed.


Does Kobe provide "leadership"? He's a very talented player, but his poor shot selection (of course this part IS measureable; Kobe's TS% is around league average, which in my book is darned poor for a player who should be, and by all indications in previous years when teamed with Shaq could have been, shooting a much better percentage) and antagonism of teammates IMO say "no". I believe it was the 2001 season in which Kobe during the first two months was neck and neck with Iverson for the league scoring lead -- and practically tearing the team apart with his lack of passing and shot selection. Gradually, from January to the All-star break, pressure from Phil Jackson, Shaq, (and I'll bet Jerry West too) got Kobe to play in a more team-oriented mode, so by spring the Lakers were firing on all cylinders. He settled for being 4th in the league in scoring (behind, in particular, Shaq), and the Lakers were the better for all this, winning the second of their three championships.

Kobe was fine in 2002 and the Lakers won another championship, but in 2004 the antagonisms surfaced again, with the contrast between the Pistons' dedication to team play compared to the Lakers being quite stark.

Doc319
Guest






PostPosted: Fri Apr 22, 2005 4:17 am Post subject: Reply with quote
MikeT, I agree that in this thread (and also in the Walker thread) we have been speaking at various times of two different types of "leadership" or, more precisely, two different kinds of--as you put it--"non-statistically measured" elements. This all started over in the Walker thread when BobC steadfastly maintained how awful a player Walker is, that any statistical improvement he is showing is only temporary and that the first hand statements of coaches, teammates and writers about his intangible contributions are irrelevant. I cited "leadership" as an example of an intangible that many observers would consider to be important and asked whether it is accounted for in the sim; the sim can either measure it or include some margin of error to account for not quantifying it directly.

I was using "leadership" as a stand in for all of the things that you call "non-statistically measured" elements but eventually the discussion focused on "leadership" to the exclusion of the larger issue of importance--what are the limits/margins of error for what can and cannot be measured by a sim or any other form of stats analysis. DanR and DeanO seem to agree that this is a question of some importance. If a sim or other form of stats analysis cannot produce measurable, testable hypotheses that can be proven or disproven, it is not science. DeanO's Roboscout predicted that Detroit would beat L.A. in the Finals--that is an example of producing a measurable, testable hypothesis; the next question would be if Roboscout can repeatedly do this with a high degree of success--scientists don't run an experiment once and then consider the results definitive for all time (I know that DeanO understands this, I'm just emphasizing it to make my point clear). The issue is really not about "leadership" at all--although it has led to an interesting, separate discussion of leadership here, which included DeanO pointing out that in his book he actually included a measurement of leadership, albeit one that he considers "soft" science.

Your characterization of Walker as a hybrid between DeanO's "alpha male" leader and a Harper/McMillan "glue" role player seems to be as apt a description of Walker as any--it fits the fact that the Celtics improved their W-L record when he arrived (and that they made the Eastern Finals during his first tenure, but lost in the first round when he departed) but that what he is providing is elusive to quantify. Other than Bob, who is certain that he has Walker's entire career trajectory plotted out, most people who posted replies about Walker seemed to find him to be a vexing and contradictory case from a stats standpoint. Perhaps this is because he is the type of hybrid that you are describing. Maybe sims/stats analyses have a higher degree of accuracy about "regular" players than the rare hybrid case.

There are two "ESPN" articles about Walker--one on ESPN.com by Bill Simmons and the other in ESPN the Magazine by Tom Friend. I was the first to mention Friend's article in this forum; someone else brought up the Simmons article, which actually appeared in print first. Friend's article mentions that Ainge basically says that he was wrong about Walker and did not realize what kind of positive impact he had on the team and in the locker room until Walker was gone. I've made reference to that article on a few occasions, wondering if there is a way to directly or indirectly quantify the positive traits of Walker's that are mentioned there. If not, we must either conclude that the NBA coaches, GMs and writers who refer to these things are idiots (DanR states in a different thread that BobC seems to believe this) or that there is some margin of error/degree of uncertainty about player ratings/sims because they cannot quantify effects that close observers of the game perceive. The last point is what I'm really getting at--what exactly can and cannot be measured and how much of a margin of error is there when a sim/rating system, etc. produces a numerical evaluation of a player. I certainly do not believe that it is producing 100% factual predictions of future performance. DeanO mentioned earlier that the accuracy is at least 75% at this point in time (not sure if he meant this in general or specifically in reference to his methods); this may or may not be true--I wonder how exactly that would be documented.

--David Friedman
Contributing Editor, Basketball Spotlight
www.suite101.com/welcome.cfm/basketball_spotlight
Back to top

Mike G



Joined: 14 Jan 2005
Posts: 3628
Location: Hendersonville, NC

PostPosted: Fri Apr 22, 2005 8:48 am Post subject: Reply with quote
Quote:
Kobe's TS% is around league average,...


Actually, league average is about .529, and Kobe winds up at .563. Earlier in the year, I made a couple of posts about how Kobe's TS% was less than all his primary teammates. But when he was injured, their efficiencies faltered; by year's end, they come out like this:

TS% . . player . . . . . . PPG
.563 Bryant,Kobe . . 27.6
.558 Atkins,Chucky . 13.6
.556 Jones,Jumaine .. 7.6
.552 Mihm,Chris . . . . 9.8
.539 Grant,Brian . . . . 3.8
.539 Odom,Lamar . . 15.2
.528 Butler,Caron . .. 15.5
.522 Cook,Brian . . . . 6.4
.437 Brown,Tierre . . . 4.4


This is almost a perfect distribution of who-should-be-taking-the-shots. It may even be that Kobe was being Too deferential toward the end of the season.

Also, this is Kobe's highest TS% ever. Due largely to the huge increase in his FTA.


Quote:
Walker 14-39 with Atlanta in a leadership role
Walker 7-15 with Boston in a nonleadership role


Here's what I have for Walker's 2005 sub-seasons:

tm. TS% Sco Reb Ast Stl TO.. Blk -- eW/48
Atl. .474 16.7 9.4 3.6 1.1 (3.3) 0.6 - .119
Bos .490 15.9 9.3 3.1 1.1 (3.1) 1.2 - .124

eWins per 48 minutes: I'm still tinkering with weights on this one. But for a given player, you see basically the same Antoine in Boston. They didn't need him to shoot as much; he "improved" by monopolizing the ball less and by blocking a few shots.

But this is still an above-average player. He helps most teams, and he may be the type who doesn't help teams with great depth.


Quote:
...last year you used Roboscout to predict a Pistons' triumph in the Finals...


Did Robo do this ?
Back to top
View user's profile Send private message Send e-mail
HoopStudies



Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA

PostPosted: Fri Apr 22, 2005 8:57 am Post subject: Reply with quote
Mike G wrote:

Quote:
...last year you used Roboscout to predict a Pistons' triumph in the Finals...


Did Robo do this ?


What it did is highlight a few strategies that could be used to win in 5 games. I had no way of knowing whether Detroit would use those strategies. So it wasn't a prediction and Roboscout is not about predictions. But in the sense that it showed how close Detroit was to winning in 5 and not getting killed as most people seemed to think, it makes a prediction. Still, I'd say it's more a strategy tool than a predictor.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Doc319
Guest






PostPosted: Fri Apr 22, 2005 6:22 pm Post subject: Reply with quote
DeanO, I did not mean to overstate or misrepresent what you did with Roboscout. I do remember you mentioning that IF the Pistons followed certain strategies then they had a good chance of winning, which is not the same as predicting that Detroit would win. Still, what Roboscout did fits in with what I am talking about in terms of making a verifiable hypothesis and then observing what happens. Apparently Detroit followed enough of those strategies to be successful--I know that after the Finals you posted info with details about how extensively Det did or did not use Roboscout's recommended strategies, but I don't have that info in front of me at the moment.

MikeG, your comment about Kobe being TOO deferential is very intriguing. Of course, such a sentiment flies in the face of conventional wisdom, but if one actually watched the Lakers play this year, it became clear that Kobe deferred to his teammates quite often--driving to the hoop and then kicking the ball to three point shooters when he could easily have decided to shoot the ball himself. If he were truly determined to win the scoring title, he would have shot every one of those point blank opportunities, particularly since Iverson had a good lead over him in the scoring race. Kobe also dumped the ball off a lot to Mihm, who sometimes caught the ball and finished well and who sometimes fumbled the ball out of bounds. This is not to say that Kobe did not shoot a lot, but shooting a lot does not mean that he was not trying to involve his teammates also. Kobe, Iverson, T-Mac, LeBron, Wade, etc. all shoot a lot and they all also involve their teammates--it's just that LeBron and Wade get a lot more credit for doing so.

I wonder if anyone measures what I like to think of as "hand-grenade" FGAs--the shots that someone gets stuck with when the shot clock is winding down and his teammate passes him the ball so he doesn't get caught holding the "grenade" when the shot clock "explodes." To be specific, a "hand-grenade" FGA could be defined as a shot that occurs after a player receives a pass outside the paint with less than 4 seconds on the shot clock. We could make a sub-category for plays that begin with less than 4 seconds left in the quarter or game, since the team has no choice but to launch such a shot (as opposed to starting a possession with enough time and mismanaging the situation to the point that someone gets stuck holding the "grenade"). This discounts feeds to post players for easy dunks as the shot clock expires and discounts guys like Marbury and Francis who dribble out the whole clock and then launch these kinds of shots. I suspect that Kobe shot a high number of such attempts this year and not solely because he was monopolizing the ball. Often the Lakers would run a play or set, it wouldn't result in an open shot and someone would pass the ball back to Kobe as the clock wound down, so Kobe had no choice but to shoot.

--David Friedman
Contributing Editor, Basketball Spotlight
www.suite101.com/welcome.cfm/basketball_spotlight
Back to top

kjb



Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC

PostPosted: Fri Apr 22, 2005 7:56 pm Post subject: Reply with quote
David: The "hand-grenade" possession information is available at 82games. Check the shooting details on the right hand side for that player's shot clock usage. Here's Kobe's page: http://www.82games.com/04LAL6A.HTM

17% of Kobe's FGA came in the final 3 seconds of the shot clock, according to this data. Contrast with Paul Pierce -- 10%. Or Iverson -- 6%. McGrady 10%, Garnett 16%, Kidd 13%, Marbury 12%, Nash 6%, Duncan 16%, Ginobili 13%, Arenas 10%, Dirk 13%, Wade 12% and Shaq 18%.

The fact that Kobe took so many shots late in the shot clock does suggest he took a significant portion of the tougher/lower percentage shot attempts. Roland's shot clock data shows that shots late in the shot clock are lower percentage than shots taken earlier.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
jambalaya



Joined: 30 Jan 2005
Posts: 282


PostPosted: Sun Apr 24, 2005 2:54 am Post subject: kobe next year - 21 areas for possible improvement Reply with quote
we dont know the coach or the full lineup of course but among these which do you think kobe is likely to achieve next year and which should he and the coach most emphasize?

1. score 30 points a game
2. dish out 9 assists
3. grab 7.5 rebounds
4. improve 3 pt % to greater than 36%
5. increase 3ptrs made to 3-4 a game
5. raise eFG% for all shots from 48% to over 50%
6. raise eFG% on jump shots from 43% to over 45%
7. increase inside shots

(by the way in addition to bob's comments about kobe's career high scoring FG%, i noticed two other things- he shot a resptable 47% from just 2pt land overall. but % inside shots fell from 34% last year to a still high 29%. the presence of malone and payton probably got kobe more shots on cuts and now that is gone. odom and atkins arent malone and payton but they are fairly capable passers. kobe's inside shots are still up compared to 2002-03 when it was only 22%. this was probably a conscious effort on kobe's part + some ability/finds by the new teammates.)

8. improve clutch eFG% from 35% to 40% (like year) or above
9. cut his turnovers to under 3.
10. raise his steals to over 2.
11. play point full-time
12. let somebody else initiate more of the offense
13. help improve team D by at least 3 pts
14. raise his +/- from around +3 to over +8 (it was only +3 this year, +6 last year, +10 the year before
15. raise his teammates scoring by at least 3-5 points
16. recruit effective complementary free agents
17. do or stay with what the coach asks more often on offensive plays
18. be a better leader off the court and in the locker room
19. develop more chemisty with butler and odom
20. play fewer minutes and make the minutes he does play more efficient and impactful
21. stop trying to be michael jordan or even better than that
Back to top
View user's profile Send private message
gabefarkas



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

PostPosted: Mon Apr 25, 2005 8:27 am Post subject: Reply with quote
#1 and #5 are the only ones i see really happening. but since i have no statistics to back that up, i don't have anything else to say about it.
Back to top
View user's profile Send private message Send e-mail AIM Address
bchaikin



Joined: 27 Jan 2005
Posts: 690
Location: cleveland, ohio

PostPosted: Sat Dec 31, 2005 5:13 pm Post subject: Reply with quote
any thoughts on the latest kobe bryant fracas (with mike miller of the grizzlies)? more selfishness or a show of leadership?...

[Crow]

hoyasaxa



Joined: 23 Jun 2008
Posts: 5
Location: NY

PostPosted: Tue May 19, 2009 8:19 am Post subject: Denver Nuggets "Assembled Largely by Instinct" Reply with quote
Don't know how many on the board read the Wall St Journal. Thought you guys might find this interesting.

http://online.wsj.com/article/SB1242699 ... al_journal

Rex Chapman:

Finally as nearly all sports executives rely increasingly on statistical analysis, Mr. Chapman says the Nuggets act on instinct rather than spreadsheets. "When you are talking about chemistry and camaraderie, you have to," he says.
Back to top
View user's profile Send private message
KD



Joined: 30 Jan 2005
Posts: 163


PostPosted: Tue May 19, 2009 9:42 am Post subject: Reply with quote
And sometimes instinct falls in line with what has been proven to work.

And sometimes your coach decides to get his act together.

And sometimes some other team wants to give you an all-world player for the return of some cap relief.

And I appreciate the instinct that did away with Chris Andersen's drug problem, Kenyon Martin's knee issues, and Nene Hilario's illness.

Forget Ming. ALL HAIL INSTINCT.

(That said, I should actually read the article, now.)
Back to top
View user's profile Send private message
NewCollegeHokage



Joined: 14 Apr 2009
Posts: 9


PostPosted: Tue May 19, 2009 10:28 am Post subject: Reply with quote
Doesn't Dean Oliver work for the Nuggets? The guy wrote the book on basketball stats so far? New York Times > Wall Street Journal
Back to top
View user's profile Send private message AIM Address
Ryan J. Parker



Joined: 23 Mar 2007
Posts: 711
Location: Raleigh, NC

PostPosted: Tue May 19, 2009 10:39 am Post subject: Reply with quote
The terminology sucks here.

I figure "instinct" means "basketball knowledge" or maybe even perhaps "subjective probability". Cool
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website
Mike G



Joined: 14 Jan 2005
Posts: 3608
Location: Hendersonville, NC

PostPosted: Tue May 19, 2009 10:58 am Post subject: Reply with quote
Where in statistical analysis can one find 'camaraderie'?

Don't forget that they dumped Camby, getting nothing, and some wrote this off as idiotic or worse. Still they've done pretty well, to put it mildly.

Maybe DeanO has just been a good-luck token (Den, Sea'05); but I doubt it.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
SGreenwell



Joined: 12 Feb 2005
Posts: 76
Location: Rhode Island

PostPosted: Tue May 19, 2009 9:43 pm Post subject: Reply with quote
NewCollegeHokage wrote:
Doesn't Dean Oliver work for the Nuggets? The guy wrote the book on basketball stats so far? New York Times > Wall Street Journal

Rex Chapman thinks he hired this guy. Don't tell him, lest you want the number of employed statisticians to drop by one.
_________________
Now blogging, although nobody asked!

http://stephenonstuff.blogspot.com/
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address
HoopStudies



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

PostPosted: Wed May 20, 2009 1:03 pm Post subject: Reply with quote
I just got a chance to see that article.

Of course we use stats! (and spreadsheets) Good business uses both instinct and analysis. Our owner has done business for a long time and been quite successful using both hard analysis and instinct. I certainly spend a lot of time translating numbers to words -- so maybe it sometimes looks like mostly instinct. Did we know "it was going to turn out this way"? No, but we were doing things that analytically and instinctively were toward the goal of having it "turn out this way." The point is to have an organization that, through its resources -- analysis and instinct -- goes the right direction on a consistent basis. Is there some six-shooter in that? I think we'd all agree to that. (And we'd all be mystified as to what Six Sigma is.)
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
gabefarkas



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

PostPosted: Wed May 20, 2009 5:23 pm Post subject: Reply with quote
KD wrote:
And sometimes instinct falls in line with what has been proven to work.

And sometimes your coach decides to get his act together.

And sometimes some other team wants to give you an all-world player for the return of some cap relief.

And I appreciate the instinct that did away with Chris Andersen's drug problem, Kenyon Martin's knee issues, and Nene Hilario's illness.

Forget Ming. ALL HAIL INSTINCT.

(That said, I should actually read the article, now.)

Sounds like someone bet against the Nuggets. Let's just call this someone "Kelly D." No, wait, let's call him "K. Dwyer".

Oh, I've said too much...
Back to top
View user's profile Send private message Send e-mail AIM Address
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 979
Location: Seattle

PostPosted: Wed May 20, 2009 5:29 pm Post subject: Reply with quote
I think it is an interesting sign of progress that the notion of a successful team not using stats (even if we know that's not the case) is considered revolutionary.
Back to top
View user's profile Send private message Send e-mail Visit poster's website
mtamada



Joined: 28 Jan 2005
Posts: 377


PostPosted: Thu May 21, 2009 3:53 pm Post subject: Reply with quote
Mike G wrote:
Don't forget that they dumped Camby, getting nothing, and some wrote this off as idiotic or worse. Still they've done pretty well, to put it mildly.

Maybe DeanO has just been a good-luck token (Den, Sea'05); but I doubt it.


Yeah, I've been wondering about the same thing as the playoffs progress. The Lakers seem not to hesitate to hire top-notch talent, but have they beefed up their statistical analysis capabilities? The Ming-and-McGrady-less Rockets gave them all they could handle, and so did the Nuggets in Game 1. I can't tell from watching how much of the Rockets' and Nuggets' success was do to scouting and analysis, but surely the NY Times article alerted the Lakers to the fact, if they weren't aware of it already, that some teams, notably the Rockets and Nuggets, have backrooms buzzing with activity as analysts seek out ways to neutralize the Lakers.

I like Camby as a player, but a recipe which includes declining to spend big bucks on Camby while choosing to pay big bucks to Billups looks like a good recipe.
Back to top
View user's profile Send private message
Italian Stallion



Joined: 04 Mar 2009
Posts: 112


PostPosted: Fri Jun 05, 2009 6:57 pm Post subject: Reply with quote
I think stats are generally superior to intuition. However, I think some people are more naturally gifted at "intuiting" things that are very complex and difficult to measure statistically.

I hate to say this about myself because it's in bad taste, but I believe I have that skill in some areas of life (though not necessarily at judging basketball talent, skill etc...). I sometimes sort of know things are true without being able to demonstrate it or prove why. But I'd be more than willing to bet on it and wait out the results.

I think the human brain absorbs a lot of information and on some level can process that knowledge and experience etc... and use it to solve very complex things and give you answers without you even knowing why you know.

In fact, I think some people that have that skill at high levels and also have high level mathematical skills can actually work backwards with great sucesss. They can start with what they know and then try to explain it mathematically.
Back to top
View user's profile Send private message
Neil Paine



Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA

PostPosted: Fri Jun 05, 2009 11:09 pm Post subject: Reply with quote
Italian Stallion wrote:

I think the human brain absorbs a lot of information and on some level can process that knowledge and experience etc... and use it to solve very complex things and give you answers without you even knowing why you know.

Somebody's been reading Gladwell...
Back to top
View user's profile Send private message Visit poster's website
HoopStudies



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

PostPosted: Sat Jun 06, 2009 10:51 am Post subject: Reply with quote
davis21wylie2121 wrote:
Italian Stallion wrote:

I think the human brain absorbs a lot of information and on some level can process that knowledge and experience etc... and use it to solve very complex things and give you answers without you even knowing why you know.

Somebody's been reading Gladwell...


I like reading Gladwell, but Blink did leave me wondering what exactly was the message. I think the messages were these:

- Your instinct can be very wrong unless you really study things well, at which point there will be times you know things without having to go through the analysis, without knowing why you know.

- We do all clamor for "more data!" but sometimes it isn't data that is necessary, rather the knowledge of how to use the data.

I get these mainly from the examples he had of the person studying couples and the statue experts. The latter example is important because it shows how being an expert who doesn't know why they know isn't as useful. The museum still bought the statue because the experts couldn't explain their intuition whereas all the scientific tests could explain why they were valid.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Ryan J. Parker



Joined: 23 Mar 2007
Posts: 711
Location: Raleigh, NC

PostPosted: Sat Jun 06, 2009 11:56 am Post subject: Reply with quote
Quote:
- We do all clamor for "more data!" but sometimes it isn't data that is necessary, rather the knowledge of how to use the data.


Along these lines, my feeling is that the data available hasn't been used to the fullest extent yet (on the outside, at least). Sure we'd like more individual defensive data, but have we really used the play-by-play/shot charts/etc to the fullest extent yet? I'm pretty sure the answer is no...
_________________
I am a basketball geek.
Back to top
View user's profile Send private message Visit poster's website
gabefarkas



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

PostPosted: Sat Jun 06, 2009 1:13 pm Post subject: Reply with quote
Italian Stallion wrote:
They can start with what they know and then try to explain it mathematically.
I think that's called "confirmation bias".

page 2 missing

[Crow]

holymoly



Joined: 30 Jul 2005
Posts: 63


PostPosted: Thu Jun 21, 2007 10:55 am Post subject: Similarity Scores Reply with quote
I have been trying to run my own Similarity Scores and for a first view I looked at Basketball Reference's calculations.

www.basketball-reference.com/about/similar.html

Justin uses the following metrics;

Height (Ht)
Minutes per game (Min/G)
Percentage of team field goals attempted while on the court (%Shots)
Percentage of team offensive possessions used while on the court (%Poss)
Effective field goal percentage (eFG%)
Free throw percentage (FT%)
Free throw attempts per field goal attempt (FTA/FGA)
Percentage of teammate field goals assisted (%Ast)
Turnovers per 100 possessions (%TO)
Steals per 100 opponent possessions (%Stl)
Blocked shots per 100 opponent two-point field goal attempts (%Blk)
Percentage of possible offensive rebounds (OR%)
Percentage of possible defensive rebounds (DR%)

Seems good to me. But after running the normalised scores myself and adding the weightings etc I found one major problem, namely it didn't allow for variations in their 3PT Distribution. I was getting back PF's who are similar in all the above metrics but completely different when it comes to the number of 3's they are shooting. For me this compromises how 'similar' they are.

I just wanted to gauge peoples opinion on this and if they saw any other metrics which could be useful additions to the model.
_________________
.....
60% of the time, it works every time
Back to top
View user's profile Send private message Send e-mail MSN Messenger
Mike G



Joined: 14 Jan 2005
Posts: 3566
Location: Hendersonville, NC

PostPosted: Thu Jun 21, 2007 12:01 pm Post subject: Reply with quote
This is surely an upgrade over earlier versions which only compared season totals. For low-minute players, 'similar' players were often not at all similar.

If you are looking for similar production rates in the various categories, then it shouldn't matter whether points are gotten by 3's, 2's or 1's. Since a given player, if he's a versatile scorer, may take what the defense allows, he may shoot a lot of 3's one night, get to the line a lot the next game, and score mostly on 2's in a third game. He could have equal scoring production and efficiency thru these different avenues; and obviously he should be considered similar to himself, whether in different games or different seasons.

Height is not a productivity, and if a 6-4 player (Barkley) can produce figures similar to a 6-10 player (Malone), then I'd say they have similar game, could be traded for one another without much change to either team, etc. You may pick and choose which elements are essential to your analysis, from among the available 'productivity' and 'non-productivity' (age, vertical, weight, etc) variables.

The thing about adding more variables is that the more of them you have, the less each one counts. If you had just one variable, you can compare 'offensive rebounders', say. At the other end of the spectrum, if you have a zillion comparisons, none of them carry much weight; and you might conclude that 2 players are 'similar' because they are the same race, weight, wingspan, birth state, etc, etc; even though one averages 17-5-2 and the other averages 10-9-4.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Thu Jun 21, 2007 1:09 pm Post subject: Reply with quote
I agree with you holymoly on the importance of the 3pt shot in getting on the court similars that match the way they play (not essential to all and every circumstance but usually of interest to me) and would add inside scoring consideration (% shots taken inside from 82games and FG% there).

ideally I'd also want to add shot defense but perhaps in a limited way: players are clumped in wide middle "normal" (for inability to get agreement using present stats for precise calls about average , just above average and just below average) unless their 82games FG% and pts allowed for both are a certain distance above or below average in terms of standard deviations.
Back to top
View user's profile Send private message
gabefarkas



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

PostPosted: Thu Jun 21, 2007 2:16 pm Post subject: Reply with quote
Mike G wrote:

If you are looking for similar production rates in the various categories, then it shouldn't matter whether points are gotten by 3's, 2's or 1's. Since a given player, if he's a versatile scorer, may take what the defense allows, he may shoot a lot of 3's one night, get to the line a lot the next game, and score mostly on 2's in a third game. He could have equal scoring production and efficiency thru these different avenues; and obviously he should be considered similar to himself, whether in different games or different seasons.


Aha, I see. So Player A is 6'9", 220 lbs, and his career Pt/Reb/Ast per game are 18.4, 5.1 and 2.0. Player B is 6'9", 225 lbs, and his career rates are 17.6, 5.1 and 3.0.

They sound pretty similar, right?

At least, until you note that Player A's 3PM/3PA for his career are 2.1/5.2, while Player B's are 0.1/0.5.

Although their similarity scores might be pretty high, I would be hard-pressed to say these two were/are similar at all.


I'll even throw in Player C, at 6'9" and 215 lbs, with career averages of 16.0/4.3/1.9, but whose career 3-point shooting is/was 0.0/0.1.
Back to top
View user's profile Send private message Send e-mail AIM Address
kjb



Joined: 03 Jan 2005
Posts: 864
Location: Washington, DC

PostPosted: Thu Jun 21, 2007 2:21 pm Post subject: Reply with quote
Does Justin (or anyone else doing similarity scores) adjust for pace?
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
Kevin Pelton
Site Admin


Joined: 30 Dec 2004
Posts: 978
Location: Seattle

PostPosted: Thu Jun 21, 2007 6:18 pm Post subject: Reply with quote
kjb wrote:
Does Justin (or anyone else doing similarity scores) adjust for pace?

I'm confused. Looking through the metrics holymoly used, all of them appear to me to be independent of pace.
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Thu Jun 21, 2007 7:23 pm Post subject: Reply with quote
I agree with Kevin P that Justin's criteria are pace adjusted but there is no explicit added label proclaiming it.

Kevin B, I went to your blog and read the article about consistency and it might be another thing to add to the max version of a similarity list.

I am in the process of reading some of your other pieces I might have missed. This MVP article of yours might interest some if they didnt see it before. http://www.nbcsports.com/nba/1129882/detail.html

Last edited by Mark on Wed Jul 04, 2007 4:13 pm; edited 2 times in total
Back to top
View user's profile Send private message
jkubatko



Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH

PostPosted: Thu Jun 21, 2007 7:59 pm Post subject: Reply with quote
Mark wrote:
I agree with Kevin P that Justin's criteria are pace adjusted but there is no explicit added label proclaiming it.


They're not pace-adjusted, they're independent of pace. It wouldn't make any sense to say that these variables are adjusted for pace.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Mark



Joined: 20 Aug 2005
Posts: 807


PostPosted: Thu Jun 21, 2007 8:22 pm Post subject: Reply with quote
"Independent of pace" is what Kevin P. said and may be the preferred way to say it to some (and I could have gone with it without problem but instead just happened to used Kevin's B.'s choice of word) but I believe that it is the same basic understanding underneath as when I said "adjusted for pace" (i.e. taking the step of dividing by 100 possessions or calculating and using % of various team stat totals in recognition that the team total and player's stat level are affected by pace) and I don't think it is crucial. Adjusting the raw stats to assess them "independent" of pace (or affects). But the independence seems to me an assumption, simplification or looser use of that term and not a clearcut proven fact. I'll probably stick with "adjusted for pace" (and understand that to myself as adjusted for the obvious main affect of pace on raw stats to try to somewhat normalize comparisons), but I can accept how others might use "independent" and prefer it.
Back to top
View user's profile Send private message
kjb



Joined: 03 Jan 2005
Posts: 864
Location: Washington, DC

PostPosted: Fri Jun 22, 2007 8:28 am Post subject: Reply with quote
I'm guilty of sloppy reading. Sad Justin standardizes his stats (says so right in his frigging write-up), which is really what I was asking.

Mark: Thanks for the comments about the blog. I just started it, but with my work schedule, I update infrequently. I finally hired someone to fill the last slot on my staff, so I'm hoping that will ease the time crunch a bit and I'll be able to write more. The next few entries will probably be about screenwriting, though.
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
Mike G



Joined: 14 Jan 2005
Posts: 3566
Location: Hendersonville, NC

PostPosted: Fri Jun 22, 2007 9:00 am Post subject: Reply with quote
gabefarkas wrote:
Player A's 3PM/3PA for his career are 2.1/5.2, while Player B's are 0.1/0.5.

Although their similarity scores might be pretty high, I would be hard-pressed to say these two were/are similar at all.
.


A player in 1991 averaged 0.4/1.3 3's/40min. In 1997, a player has almost identical Pts/40, Reb/40, etc. But that player shoots 1.4/3.8 3's/40. Both play 37-38 mpg. Are they dissimilar players just because one shoots 3 times as many threes?

Both added 15.8 PW and 0.8 PL to their teams. Justin gives them 56 and 57 WS. In fact, they are both Michael Jordan. They had different teams and different rules, and both took what they got.

Is some other player more similar to Jordan -- based on 3-point frequency -- than he is to himself?
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
jeffpotts77



Joined: 18 Feb 2005
Posts: 150
Location: Cambridge, MA

PostPosted: Fri Jun 22, 2007 9:30 am Post subject: Reply with quote
Mike G wrote:
A player in 1991 averaged 0.4/1.3 3's/40min. In 1997, a player has almost identical Pts/40, Reb/40, etc. But that player shoots 1.4/3.8 3's/40. Both play 37-38 mpg. Are they dissimilar players just because one shoots 3 times as many threes?

Both added 15.8 PW and 0.8 PL to their teams. Justin gives them 56 and 57 WS. In fact, they are both Michael Jordan. They had different teams and different rules, and both took what they got.

Is some other player more similar to Jordan -- based on 3-point frequency -- than he is to himself?


I'm sure that, as long as you don't weight the 3-point frequency too heavily, 91' Jordan would still rate fairly high on the similarity score to 97' Jordan. I do think 3-point frequency should be a factor. 97' Jordan, though similar, was not exactly the same player as 91' Jordan.
Back to top
View user's profile Send private message Yahoo Messenger
holymoly



Joined: 30 Jul 2005
Posts: 63


PostPosted: Fri Jun 22, 2007 9:52 am Post subject: Reply with quote
Mike G wrote:
gabefarkas wrote:
Player A's 3PM/3PA for his career are 2.1/5.2, while Player B's are 0.1/0.5.

Although their similarity scores might be pretty high, I would be hard-pressed to say these two were/are similar at all.
.


A player in 1991 averaged 0.4/1.3 3's/40min. In 1997, a player has almost identical Pts/40, Reb/40, etc. But that player shoots 1.4/3.8 3's/40. Both play 37-38 mpg. Are they dissimilar players just because one shoots 3 times as many threes?

Both added 15.8 PW and 0.8 PL to their teams. Justin gives them 56 and 57 WS. In fact, they are both Michael Jordan. They had different teams and different rules, and both took what they got.

Is some other player more similar to Jordan -- based on 3-point frequency -- than he is to himself?



I use similarity scores to not simply match production but also to attempt to match style, i.e. how they achieved their production. FTA/FGA is already in the model so how often they drive to the lane is loosely covered (1pts) but there is not distinction between 2's and 3's which for me relating to 'style', undermines the results.

I guess it also comes down to how your using this info. If I was a GM looking to see which players could replace the production of a player i've just lost in free agency, then 3PT distro isnt a factor. If I am looking to see which players are most similar to a player in style and production then I would include 3PT Distro. Unfortunatley I'm not a GM so the latter part is more relevant for me.

With regards to the Jordan example above, Jeff Potts is right you would find that the similarity scores would be very similar even with 3PT distro included. But don't forget when looking at similarity scores only players of the same age are compared. Maybe there is evidence that as players (guards) get older they rely on shooting the 3 a little more as they lose their quickness/ability to get to the foul line. So comparing players at different ages isnt that relevant for me.

I'm not suggesting that 3PT distro should become the prominent metric in the model but I do think it would benefit from being included.
_________________
.....
60% of the time, it works every time
Back to top
View user's profile Send private message Send e-mail MSN Messenger
Mike G



Joined: 14 Jan 2005
Posts: 3566
Location: Hendersonville, NC

PostPosted: Fri Jun 22, 2007 9:57 am Post subject: Reply with quote
Of course players aren't exactly the same from year to year or game to game. I was suggesting that overall impact, in every area of production, can be the same whether you score on dunks, 3's, FT, whatever.

Another whose interest in the 3 changed: Larry Bird. From a low of 0.7 att/G to a high of 3.1 , he became a better scorer and passer, and did less rebounding.

But here,

http://www.basketball-reference.com/pla ... dla01.html

...for that 1988 season, Bird's similars list the Malones and Webbers and Hills we'd expect; also McGrady, Antoine, and: George Gervin 1979.

Gervin didn't shoot 3's in '79, of course. While their scoring rate and TS% were very similar, Bird was just about twice the rebounder and assister. About half the fouler, and less turnover-prone. How are these, overall, very similar players? The 909 simscore falls squarely between Webber '01 and Webber '05.

My own similarity spreadsheet yields 366 more-similar player-seasons, since 1978, including a few by Gervin. Can we really suppose a 30-10-6 player can be replaced by a 30-5-3 player? Are variables like Minutes and Height weighed too heavily, relative to AstRt and RebRt?

Quote:
If I am looking to see which players are most similar to a player in style and production then I would include 3PT Distro

And as per my initial response, the more 'style' elements you include, the further your similarity strays from 'production'. You get players who look alike, rather than who play alike.
Quote:
when looking at similarity scores only players of the same age are compared.

Actually, at b-r.com you get it with or without age-restriction. Since age is arbitrarily relative to some date in the year, you may be looking at players who are 11-months different and excluding some who are 1-month off in age.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
penbeast0



Joined: 11 Aug 2005
Posts: 36


PostPosted: Fri Jun 22, 2007 10:46 am Post subject: Reply with quote
I do tend to agree that you want to include 3 pointers in there, probably 3pt made since you have both FG attempts and eFG% already. It's a rough measure of outside shooting which helps open up the floor for other player plus, similarity scores should show style to some degree as well as effectiveness. . .

Similarly, I would not only remove height, I think that having two rebounding numbers (off and def) overstates rebounding's importance particularly since you are using this for perimeter players too; just use the total rebounding number. Thus you add one factor and remove two.

page 2 of 2 missing