Player "position" and performance in the playoffs

[Crow]

Picking up on a question that arose to me from reading the refereed journal thread:

For "positions" at the level of PG, wings, bigs in recent playoffs (since 1991), which positions have the greatest rate of winshares per 48 minutes performances over .150 in those playoffs? (chosen as a quick cutoff for strong performance / "importance")

Using requirements of at least 16 playoff games that season and at least 25 minutes per game (capturing players who at least got to the conference finals and were probably starters or 6-7th men) I get a dataset of 127 players or barely 6 per season.

A bit short of 45% were big men. Almost 40% are wings. Only about 15% are primarily PGs for a modest under-representation.

Among those who actually won championships (a bit over 40) it was about 46% are big men, 46% wings and less than 8% were primarily PGs.

This may say stuff about the metric and maybe it looks different with another metric, but this is one possible answer to the question of deep playoff performance and perhaps "importance" by "position". For a starting answer, it doesn't look good for the importance of small guys who are PGs. Maybe they are getting hurt too much for turnovers and / or not given enough credit for distribution? Does RAPM cure that and do they perform better on this complete metric?

But go gaga for Westbrook, Paul, Wall, Jennings, Rubio, etc. if you want to. People did before for Stockton, Payton, Nash, Iverson, Kevin Johnson, Terell Brandon, etc. mostly with good rationale but maybe with too much reference back to Magic and Isiah and not enough to their faster pace era.

If Derrick Rose or someday Irving succeed it may be more because they were the #1 scoring perimeter than because of performance of the traditional distributional duites of PG.

[Crow]

In 2010-2011 4 PGs were +3 or better on prior informed RAPM in the regular season out of 26, or about 15% and a modest under-representation. None made the conference finals.

In 2009-2010 4 PGs were +3 or better on RAPM out of 25, or again about 15% and a modest under-representation. One made the conference finals but lost there.

In 2008-2009 5 PGs were +3 or better on RAPM out of about 33, or again about 15% and a modest under-representation. None made the conference finals.

In 2007-2008 (with the 3rd fastest pace in 18 years, pushing 93) had 6 PGs (the most in the period studied) were +3 or better on RAPM out of about 27, or about 22% and a slight over-representation. One made the conference finals but lost there.

In 2006-2007 4 PGs were +3 or better on RAPM out of about 27, or again about 15% and a modest under-representation. One made the conference finals but lost there.

In 5 years 23 PGs out of about 138 or about 17% and a slight under-representation. Only one in 8 of those made the conference finals and all lost there. Do you want to build around a PG as your #1 star these days?

[Mike G]

Using the same criteria of 16 playoff games and at least 25 mpg, but with a different sample -- currently active players -- I have 55 such player-postseasons, using (the equivalent) 1.50 eWins per 484 minutes cutoff.
I get 22 bigs, 21 wings and 12 PG.

Actually, your count of 15% (vs 20% expected from an unbiased sample) isn't far-fetched, since you started at 1991 -- right after Magic/Isiah era, and heading into the Bulls-then-Lakers titles with lineups that really didn't have (or need) PG's.

Then, too, it may be that Win Shares weigh turnovers too heavily, or assists too lightly.

[Crow]

21 years isn't a real long time but isn't real short either. The game is different than it was in the 80s or earlier. It wasn't picked specifically to exclude Magic and Isiah but maybe a 31 year comparison would have been different, perhaps more "neutral".

It doesn't appear to change the answer though. Primarily PGs are still underrepresented among the qualifiers for this longer period on playoff winshares per 48 even with Magic.

The result for primary offensive initiator would be stronger than it is for PGs by size / who they guard. Ball dominant wings and even some point forwards certainly have had an impact on the playoffs. So this may not be denigrating those who perform the role.


The last 5 years is somewhat different than what came before; I consider it a mini-era. For all the talk about PGs these days the superstar PGs have no recent titles. Whether that is mostly random or a sign of what wins now I can't say for sure of course. Billups was the last one in 2004, with under 6 assists per game (which migh t be how the next one does it, with some mix of shooting and / or defense).

Parker was nowhere near .150 WP /48 in 2005 or 2007 playoffs and neveer was over .15o in the playoffs ever. Though he was awarded playoffs MVP in 2007, he had a .100 WS/48, while Duncan and Ginobili were over .200.

Rondo was .121 in 2008 playoffs, good for 7th place on his team.

Fisher was over .150 in the playoffs for just 1 of the 5 titles.

None of the Rockets' PGs were over .150 in the playoffs for their championship seasons.

The Bulls usually had a PG or sometime PG over .150 in their title years but only 2 PGs were over 25 minutes per game, just barely and both were the 6-6 to 6-7 Ron Harper.

[Crow]

When Magic won his first title league average pace was 103 and the Lakers played at 104. At the very end of the 80s the league average pace had declined substantially but was still 98 when Isiah won his titles. The Pistons played much slower than average but were still at 94.

In at least the past 5 seasons the average pace has been below 90. A different era. Perhaps the pace is not as suited to very high performance / ultimate victory by a star PG.

[Mike G]

In this century, PG with eW/484 > 1.50, >16 G and >25 mpg in playoffs.
Standardized per36 rates (relative to 100 pts/tm and 44 reb/tm per game).

Code: Select all

PG      yr   tm   Eff%   Sco   Reb   Ast   Stl   TO   Blk   ws48    PER   e484
Kidd    02   NJ  .485   19.1   7.7   9.2   1.6   3.2   .3   .139   21.7   2.08
Kidd    03   NJ  .512   19.9   7.7   8.0   1.6   3.6   .2   .149   19.2   1.94
Billups 04  Det  .538   20.3   3.2   6.9   1.4   2.5   .1   .203   18.8   1.56
Billups 05  Det  .577   22.9   4.5   7.1   1.0   1.9   .2   .224   20.0   1.95

Billups 06  Det  .563   20.7   3.5   7.1   1.1   2.2   .1   .184   19.1   1.65
Nash    06  Phx  .605   20.4   3.3   8.9    .4   3.1   .2   .153   21.3   1.63
Parker  08   SA  .540   24.5   3.8   4.9    .9   2.8   .1   .122   19.9   1.61
Billups 09  Den  .646   22.6   4.0   5.6   1.2   1.9   .2   .249   22.9   1.61

Nash    10  Phx  .621   21.8   3.9   9.7    .3   4.1   .1   .175   22.4   1.85
Rondo   10  Bos  .495   15.4   5.5   7.9   1.7   2.7   .1   .131   17.8   1.65
Rose    11  Chi  .496   27.5   4.4   7.6   1.2   3.3   .6   .165   22.5   2.26
Westbrk 11  Okl  .490   23.3   5.5   7.5   1.4   4.5   .3   .084   19.6   1.86

WS loves a good shooting% and hates turnovers.
Recall that ws/48 of .150 is roughly PER = 20.
WS likes Billups more than eW does (X4), and eWins likes all the others more than WS does.

[Crow]

Just one title among these 12 leading PG seasons and it was the weakest one barely above 1.5 e484 and only a modest part of that success story. A high performing PG on this metric is a rarity for champions as with ws/48.

If you had 12 PGs over 1.5 e484 in the playoffs this century, how many did you have in the 90s and 80s?

[Mike G]

Since some years' playoffs have had a lot more games than other years, there's no very easy way to just look at conference finals participants. But when I look at all players with at least 1 eWin in a given postseason, I get these totals by nominal position.

Code: Select all

playoff eWins                    fraction of total
90s    1.50   1.25   all       90s    1.50   1.25    all
Bigs    151    192   217      Bigs    .486   .494   .467
Wings   100    126   161      Wings   .322   .324   .346
Points   60     71    87      Points  .193   .183   .187
Total   311    389   465      Total     1      1      1
                        
80s    1.50   1.25   all       80s    1.50   1.25    all
Bigs     95    109   139      Bigs    .403   .368   .365
Wings    88    124   158      Wings   .373   .419   .415
Points   53     63    84      Points  .225   .213   .220
Total   236    296   381      Total     1      1      1

The 1.50 column is eW by players with 1.50 or greater eW/484, etc.
Point guards are actually overrepresented in the '80s.

Of the top 25 postseason eWins in the '80s, just 8 are by Bigs (McHale 3, Moses 2, Kareem 2, Hakeem 1), 10 by Wings (Bird 6, Erving 2, Worthy, Jordan), 7 by Point Guards (Magic 6, Isiah 1)

Half of the top 30 in the '90s go to Bigs: Karl Malone 5, Olajuwon 4, Shaq 2, one each by Duncan, Barkley, Ewing, and Kemp.
Wings take 11 of 30: 7 by Jordan, 2 each by Drexler and Pippen
Just 4 point guards: Isiah, Magic, Porter, and Payton (the only one after '92). Stock is at #31 ('97)

[Crow]

Another metric that has PGs doing best when the pace was highest, the 80s. And people still want that despite the pace environment being so different and the record of star PGs being so different since then.

Team quickness up n down the floor might have been more impactful than personal quickness in the halfcourt.

At league level, PG shots at the rim represent only 5.5% of all shots taken. PGs are also slightly underrepresented in getting foul shots than the average for all players (they get more than wings but well less than bigs).

[mystic]

When I talk about "big" I really mean the height of the player, not so much the position assigned to him. Magic is listed above league average in height, we can hardly call him small. Magic's biggest strength in half court sets was not breaking down the defense with dribble penetrations, but posting up the smaller defender and creating scoring opportunities for his teammates from the post.

The game is horizontal, and teams with players above average height and the necessary talent level will have an advantage. The Lakers for example won the championships in 2009 and 2010 mainly, because of their height and size. They had above average height on each position while each of those players was talented. When you look at teams matching up well with them, you will find again that those have bigger players on each of the positions. Having a 6'6 SG instead of a 6'4 is an advantage. And when your 7ft player can play closer to the perimeter it will also help a lot. We have to take wingspan into that equation, some players can play bigger due to their wingspan others a bit due to their jumping abilities.

[EvanZ]

Here are some shot rates (per 100 possessions) that I calculated as part of doing my PSAMS metric:

Code: Select all

POS  INSR   INS%	  MIDR	 MID%	 3PTR	 3PT%	   FTR	 FT%	 AND1R
PG	6.21	59.64%	10.72	36.62%	6.67	50.24%	5.68	80.70%	0.35
SG	5.56	61.61%	12.08	37.50%	8.36	53.21%	6.47	81.78%	0.37
SF	6.48	63.89%	10.31	34.79%	6.55	48.39%	6.23	74.79%	0.37
PF	8.76	62.91%	11.99	36.99%	2.93	48.91%	6.30	69.33%	0.36
CN	10.8	61.73%	 8.72	36.87%	0.84	42.27%	7.10	63.41%	0.46
AV	7.49	62.02%	10.81	36.60%	5.16	50.39%	6.35	73.87%	0.38

Inside shots = dunks, layups, tips, and hooks per 100 poss played

mid-range = all jump shots except 3pt per 100 poss played

3pt = self-explanatory

free throw rate = FTA/100 possessions played

And1R = and1 per 100 possessions played

[Mike G]

Another metric that has PGs doing best when the pace was highest, the 80s. And people still want that despite the pace environment being so different and the record of star PGs being so different since then..

Crow, you think the '80s Lakers would not be annual contenders in this century?
Or that 20-year-old Magic Johnson inserted into, say, the Orlando Magic, wouldn't make them hands-down favorites?

It's unusual for a generational superstar like Magic to join a team already in contention, just as it's unusual for a rookie Duncan to team with a Robinson. Chris Paul didn't join his talents with other superstars in NO, but his record is pretty good, including postseasons. He could have gone to the Lakers this year.

No one, it seems, will be shocked if the Bulls proceed to the Finals or win it. They're clearly led by a point guard who is perhaps the PG just because he's so talented at 6-3.

[Mike G]

Evan, it appears you've multiplied 3fg% by 1.5 ?
Do your and1 rates refer to Made FT, or FTA?

[mystic]

No one, it seems, will be shocked if the Bulls proceed to the Finals or win it. They're clearly led by a point guard who is perhaps the PG just because he's so talented at 6-3.

The Bulls without Rose on the court are at +11. Rose can play like that, because the Bulls defense is exceptional. And Rose is not the main reason for this. Yes, the offense looks bad without Rose, but as I said, the Bulls still outscoring their opponents heavily despite having to play with John Lucas as replacement for Rose. The Bulls defense gets dictated by Luol Deng. He is likely the more important player on the Bulls in comparison to Rose, even though the boxscore metrics aren't showing it.
The Bulls also have incredible depth at the wing with Korver and Brewer coming from the bench, and in the frontcourt with Gibson and Asik. All of those named players are starting caliber players in the league today.

Nobody would be surprised that the Bulls are winning, because they are a great team. Take the supposed to be best player from each team and the Bulls are still a contender.


Magic is a 6-8 player with incredible talent. A less athletic James with much better post game, shooting and courtvision. He for sure would make a big impact, given the fact that he would be huge matchup problem for basically all teams.

[Mike G]

OK, and Magic was probably even less a part of the Lakers' defense, yet few have tried to say he wasn't their Leader, certainly after 1985 or so.

Thibodeau is a pretty good coach, and yesterday he had Rose back in quickly with 2 fouls in the 1st Q.
He sat just 3.5 minutes of the game.