Best Playoff Players, 2003-2011 (*according to statistical +/-)
Posted by Neil Paine on June 12, 2011
As a follow-up to Thursday's post about the best Finals performances according to Statistical Plus/Minus, here's a playoff ranking since 2003 with a few tweaks:
- I finally re-ran the Offensive SPM formula without steals and blocks. Steals in particular were causing certain players to be extremely overvalued offensively, and there's little reason to include those defensive stats in an offensive regression. (DSPM is the same as before -- and yes, it still includes several offensive stats, but DSPM wouldn't explain more than 25% of defense without them, while OSPM's explanatory power was barely affected by dropping steals & blocks out.)
- At the request of readers, instead of per-minute SPM players are ranked by per-game "Impact", which is SPM times the % of team minutes played.
- All of a player's games are weighted by Championship Leverage, which takes into account how much the game will potentially swing the odds of a team winning the NBA title. Leverage is relative to the average playoff game in a given season (which always has a leverage index of 1.00). For instance, Game 1 of the Magic-Hawks 1st-round series had a leverage of 0.44, while Game 5 of the Finals had a leverage of 5.28. This means that, in terms of influence on championship probability, Thursday's game was 12 times as important as Game 1 of a 1st-round series, and the rankings will reflect this.
Finally, why 2003? Because that was the year the NBA adopted best-of-7 first-round series, allowing me to use the series win probabilities found here.
Anyway, here were the top playoff performers since 2003 according to per-game SPM impact, weighted by the importance of the game (minimum 10 games):
| Rank | Player | Games | Impact | Avg Lev |
|---|---|---|---|---|
| 1 | LeBron James | 91 | 7.14 | 1.26 |
| 2 | Dwyane Wade | 86 | 6.38 | 1.24 |
| 3 | Kobe Bryant | 123 | 5.71 | 1.41 |
| 4 | Tim Duncan | 128 | 5.63 | 1.28 |
| 5 | David Robinson | 23 | 5.08 | 1.96 |
| 6 | Manu Ginobili | 122 | 4.98 | 1.32 |
| 7 | Dirk Nowitzki | 105 | 4.88 | 1.13 |
| 8 | Dwight Howard | 57 | 4.50 | 1.07 |
| 9 | Chris Paul | 23 | 4.47 | 0.68 |
| Rank | Player | Games | Impact | Avg Lev |
| 10 | Kevin Garnett | 82 | 4.14 | 1.45 |
| 11 | Kevin Durant | 23 | 4.11 | 0.97 |
| 12 | Chauncey Billups | 134 | 3.90 | 1.37 |
| 13 | James Harden | 23 | 3.36 | 0.97 |
| 14 | Baron Davis | 23 | 3.35 | 0.60 |
| 15 | Steve Kerr | 10 | 3.30 | 2.71 |
| 16 | Pau Gasol | 89 | 3.07 | 1.42 |
| 17 | Deron Williams | 44 | 2.89 | 0.77 |
| 18 | Jason Kidd | 99 | 2.85 | 1.10 |
| Rank | Player | Games | Impact | Avg Lev |
| 19 | Ben Wallace | 120 | 2.81 | 1.33 |
| 20 | Paul Pierce | 94 | 2.80 | 1.27 |
| 21 | Elton Brand | 17 | 2.35 | 0.66 |
| 22 | Kerry Kittles | 31 | 2.27 | 1.50 |
| 23 | Rasheed Wallace | 134 | 2.27 | 1.51 |
| 24 | Tracy McGrady | 27 | 2.10 | 0.58 |
| 25 | Shaquille O'Neal | 92 | 2.03 | 1.23 |
| 26 | Rajon Rondo | 73 | 1.89 | 1.49 |
| 27 | James Posey | 65 | 1.83 | 1.36 |
| Rank | Player | Games | Impact | Avg Lev |
| 28 | Steve Nash | 92 | 1.81 | 0.93 |
| 29 | Trevor Ariza | 41 | 1.63 | 1.55 |
| 30 | Derrick Coleman | 12 | 1.61 | 0.69 |
| 31 | Amare Stoudemire | 56 | 1.53 | 0.74 |
| 32 | Speedy Claxton | 25 | 1.52 | 1.83 |
| 33 | Robert Horry | 91 | 1.47 | 1.24 |
| 34 | Marc Gasol | 13 | 1.40 | 0.95 |
| 35 | Alonzo Mourning | 40 | 1.38 | 1.44 |
| 36 | Allen Iverson | 26 | 1.38 | 0.53 |
LeBron James? That choker is #1?
Well, take a look at LeBron's performances in his highest-leverage playoff games (leverage >= 1.5):
| Date | Tm | Opp | MP | OSPM | DSPM | SPM | Lev | %Min | Impact |
|---|---|---|---|---|---|---|---|---|---|
| 6/9/2011 | MIA | DAL | 45.6 | -0.46 | -4.55 | -5.01 | 5.282 | 95.1% | -4.76 |
| 6/5/2011 | MIA | DAL | 45.3 | 0.80 | 0.46 | 1.27 | 4.246 | 94.3% | 1.19 |
| 6/10/2007 | CLE | SAS | 38.0 | 1.50 | -4.04 | -2.54 | 4.010 | 79.2% | -2.01 |
| 6/7/2011 | MIA | DAL | 45.7 | -5.93 | 2.15 | -3.79 | 3.874 | 95.2% | -3.61 |
| 6/7/2007 | CLE | SAS | 44.0 | -7.32 | -0.43 | -7.75 | 3.864 | 91.7% | -7.10 |
| 6/2/2011 | MIA | DAL | 39.6 | 0.03 | 5.03 | 5.06 | 3.654 | 82.5% | 4.17 |
| 6/2/2007 | CLE | DET | 46.2 | 3.85 | 6.81 | 10.66 | 3.622 | 96.3% | 10.26 |
| 5/31/2011 | MIA | DAL | 45.3 | 8.49 | 4.26 | 12.75 | 3.521 | 94.3% | 12.02 |
| 5/30/2009 | CLE | ORL | 44.9 | 3.73 | -3.30 | 0.44 | 3.179 | 93.5% | 0.41 |
| 5/31/2007 | CLE | DET | 50.3 | 21.89 | -3.39 | 18.49 | 3.139 | 86.7% | 16.04 |
| 6/12/2007 | CLE | SAS | 42.0 | 0.44 | 5.19 | 5.62 | 3.009 | 87.5% | 4.92 |
| 5/18/2008 | CLE | BOS | 46.8 | 20.63 | -2.05 | 18.58 | 2.667 | 97.5% | 18.12 |
| Date | Tm | Opp | MP | OSPM | DSPM | SPM | Lev | %Min | Impact |
| 5/21/2006 | CLE | DET | 46.8 | 2.93 | -1.86 | 1.08 | 2.335 | 97.4% | 1.05 |
| 5/24/2011 | MIA | CHI | 49.4 | 5.01 | 0.59 | 5.61 | 2.244 | 93.2% | 5.23 |
| 5/26/2009 | CLE | ORL | 49.4 | 7.73 | -1.27 | 6.47 | 2.161 | 93.2% | 6.03 |
| 5/29/2007 | CLE | DET | 44.3 | 7.49 | 3.18 | 10.67 | 2.126 | 92.3% | 9.84 |
| 5/22/2011 | MIA | CHI | 43.9 | 12.36 | 3.46 | 15.82 | 2.123 | 91.5% | 14.47 |
| 5/24/2009 | CLE | ORL | 42.0 | 15.82 | -0.61 | 15.21 | 2.044 | 87.4% | 13.29 |
| 5/24/2007 | CLE | DET | 44.7 | -5.40 | 3.75 | -1.64 | 2.005 | 93.1% | -1.53 |
| 5/21/2007 | CLE | DET | 45.4 | -3.06 | 8.50 | 5.44 | 1.932 | 94.5% | 5.14 |
| 5/18/2011 | MIA | CHI | 45.7 | 6.05 | 4.27 | 10.32 | 1.827 | 95.1% | 9.82 |
| 5/15/2011 | MIA | CHI | 41.2 | -1.71 | -0.51 | -2.22 | 1.761 | 85.8% | -1.90 |
| 5/20/2009 | CLE | ORL | 41.2 | 26.49 | -4.56 | 21.92 | 1.696 | 85.9% | 18.83 |
| 5/22/2009 | CLE | ORL | 42.5 | 7.33 | -4.16 | 3.18 | 1.609 | 88.6% | 2.81 |
| 5/27/2007 | CLE | DET | 46.1 | 12.35 | 2.33 | 14.68 | 1.504 | 96.1% | 14.11 |
I will concede that he has had negative performances in 4 of his 5 most important games, but right below those we also see double-digit impact ratings in extremely high-leverage games.
Besides, if you look at someone like Kobe Bryant, you'll also see a Top 10 littered with negative impact ratings in huge games:
| Date | Tm | Opp | MP | OSPM | DSPM | SPM | Lev | %Min | Impact |
|---|---|---|---|---|---|---|---|---|---|
| 6/17/2010 | LAL | BOS | 44.9 | -3.29 | 3.58 | 0.28 | 9.238 | 93.4% | 0.27 |
| 6/13/2010 | LAL | BOS | 43.9 | 13.71 | -2.07 | 11.64 | 4.434 | 91.4% | 10.64 |
| 6/17/2008 | LAL | BOS | 42.8 | -3.11 | -8.72 | -11.83 | 4.267 | 89.1% | -10.54 |
| 6/13/2004 | LAL | DET | 45.0 | -5.29 | -3.76 | -9.06 | 3.924 | 93.8% | -8.49 |
| 6/12/2008 | LAL | BOS | 43.4 | -0.81 | 4.86 | 4.05 | 3.756 | 90.3% | 3.66 |
| 6/11/2009 | LAL | ORL | 48.9 | 4.60 | 0.50 | 5.10 | 3.731 | 92.2% | 4.70 |
| 6/10/2004 | LAL | DET | 45.0 | -5.94 | -1.81 | -7.76 | 3.712 | 93.8% | -7.27 |
| Date | Tm | Opp | MP | OSPM | DSPM | SPM | Lev | %Min | Impact |
| 6/8/2010 | LAL | BOS | 43.9 | 6.43 | 2.98 | 9.40 | 3.564 | 91.4% | 8.59 |
| 6/8/2008 | LAL | BOS | 40.5 | 8.42 | -1.75 | 6.67 | 3.543 | 84.3% | 5.62 |
| 6/7/2009 | LAL | ORL | 48.5 | -0.16 | -0.92 | -1.08 | 3.519 | 91.5% | -0.99 |
| 6/15/2004 | LAL | DET | 45.0 | 1.62 | -7.79 | -6.18 | 3.463 | 93.8% | -5.79 |
| 6/5/2008 | LAL | BOS | 41.9 | -3.25 | -3.25 | -6.50 | 3.414 | 87.2% | -5.67 |
| 6/4/2009 | LAL | ORL | 38.0 | 20.29 | 4.64 | 24.94 | 3.391 | 79.1% | 19.73 |
| 6/10/2010 | LAL | BOS | 43.1 | 4.55 | 1.11 | 5.66 | 3.253 | 89.7% | 5.08 |
| 6/15/2010 | LAL | BOS | 39.6 | 7.27 | 9.17 | 16.45 | 3.233 | 82.5% | 13.57 |
Not even the great Dwyane Wade, perhaps the best Finals performer of his era, is immune to poor games in big spots. While his overall record (like James' and Bryant's) is outstanding, in the 3rd-most-important game of his playoff career to date, he threw up a 7-for-20, 5 turnover, 113 DRtg performance vs. Detroit.
The point? Everyone has bad games in big spots, some far more odious than James' worst offerings (Kobe's 2008 Game 6 vs. Boston was perhaps the worst big-game performance by a great player in the last decade*). But on average, the best SPM playoff performer since 2003 was LeBron -- something to keep in mind as you watch the 6.1-leverage index Game 6 tonight.
(* - Btw, the best big-game performance -- meaning the highest-leverage double-digit impact game since 2003 -- was this closeout by Tim Duncan: 21 points, 20 rebounds, 10 assists, 8 blocks, and played all but 2 minutes for a team that allowed 0.835 pts/possession.)
June 12th, 2011 at 3:15 am
Very interesting. It's nice to see Baron Davis up there too. I've always felt that he's been one of those players who has legitimately lifted their performance in the big games.
June 12th, 2011 at 5:04 am
Nice post as usual Neil, and I agree with you on the steal issue.
BTW, what's the avg. SPM and avg. impact for the playoff players who played over 50% of team minutes played?
June 12th, 2011 at 11:42 am
#2 - Thanks. The minute-weighted average SPM for players who played at least 50% of team playoff minutes was 0.81, and the average impact was 0.80.
June 12th, 2011 at 1:31 pm
Great stuff as always,
Am I missing something or are defensive rebounds more valuable for overall SPM than offensive rebounds now?
June 12th, 2011 at 1:46 pm
Indulge my ignorance a second, Neil... Is Kerr's leverage so high because he made the majority of his contributions in the closeout game against NJ in 2003 (believe that was when Scott figured out that the whole Kenyon Martin straight up D thing wasn't hacking it and started doubling down on Timmy, so Pop countered by inserting Steve)?
No Ray Allen surprises me.
June 12th, 2011 at 2:01 pm
If I was a Kobe defender, I might think that multiplying all 3 columns would be the best measure.
June 12th, 2011 at 2:25 pm
#4 - Yes, the coefficient for OReb decreased when I removed steals and blocks from the regression. Offensively, steals were seen as hugely positive while blocks were somewhat negative, so it makes sense that a new regression w/o them would compensate by decreasing the value of OReb (players who get a lot of those tend to get more blocks than steals).
#5 - Basically. By far, his best impact games of the 2003 playoffs were 5/29/2003 vs Dallas (Game 6, 4-4 3fg, 2 reb, 3 ast in 13 minutes) and 6/13/2003 vs New Jersey (Game 5, 6 pts, a rebound and a steal in 9 minutes). Those games happened to carry leverages of 2.13 and 5.11 respectively, and then in his other high-leverage games his average impact was a little better than 0, so he comes out looking good here. If your bad games come with almost no leverage (on 5/7/2003 vs. LA he had an impact of -3.31, but that game only had a leverage of 0.88), and your good games come with a lot of leverage, you'll do well, which is as it should be.
#6 - That would double-count leverage, since impact already takes into account the leverage of every game the player played. But yeah, an adjustment for total games could be made, since impact is a per-game metric.
June 12th, 2011 at 7:12 pm
Thanks for all the analyses, Neil. Cool stuff.
I'm wondering if you could calculate correlation coefficients for SPM vs. leverage for each player? A statistically significant positive coefficient would suggest better play when the leverage is high; a negative would indicate a choker.
June 12th, 2011 at 8:44 pm
I prefer this list to the broader antecedent, as it takes place over a timeline that more accurately accounts for rule changes that have softened the league.
Dwyane Wade's infamous Finals of 2006 would be an example -- there's no way I believe he puts up those numbers, has that impact, under 90s rules.
The same could be said and, I suppose, *is*, in relation to Dirk Nowitzki's Finals and overall playoffs this year. Defenders are helpless, more because of the way the game is now officiated than Dirk's skillsets; at the very least, the rule changes are greatly playing up and complementing those skills.
Individual performance is often higher, made easier, thanks to the machinations to "open the game up" that have been pushed over the last decade.
June 13th, 2011 at 1:00 am
Neil, I'm curious how sensitive this list would be if you used WinShares instead of SPM?
June 13th, 2011 at 1:16 pm
Nice to see the MONSTER impact James has had in the playoffs though, I knew his 2008 series against Boston was pretty dominant. A shame Erik wasted him but I bet James wins a few rings playing dominantly, and even when he's not in his prime too.