37 Responses to “Career Playoff Quality of Defenses Faced (1991-2010)”

1. Neil Paine Says:
   May 28th, 2010 at 10:57 am

   Manu Ginobili has been a beast in crucial games (although the opposing defenses have been relatively weak); Tony Parker has played surprisingly poorly, especially when you consider the opposing Ds.

2. Mike G Says:
   May 28th, 2010 at 1:25 pm

   This is great! Maybe.
   What the heck is OppDPAA_poss?

   (I could guess, but most likely I'm not the only one wondering.)

3. Jason J Says:
   May 28th, 2010 at 1:26 pm

   Neil - I'm having a little trouble interpreting this last table. Is this saying that Shaq put up a 113ORtg against a defenders who were (on average) 3.45DRtg better than replacement, hence we could credit Shaq with 116.45ORtg for his efforts? Or is that extrapolation way off?

4. UB Says:
   May 28th, 2010 at 2:07 pm

   Neil, I missed the post from yesterday, but can I make a suggestion when discussing "crucial" games?

   The concept that a crucial game includes "an elimination game for the trailing team," I believe is pretty flawed. That includes all 3-0 and 3-1 games?

   I think your "crucial" stat covered it much better just removing those criteria and simply keeping the rest. To wit:

   "Game 3 or later in a 7-game series; Conference Semifinals or later; series tied, within 1 game either way" = 2-1; 2-2; 3-2; 3-3.

   Adding in "elimination for the trailing team" games simply adds 3-0 and 3-1 games, which are by far the least crucial of the bunch (if you REALLY want to include 3-1 games I wouldn't argue too much, but 3-0 strikes me as a humongous stretch).

5. Neil Paine Says:
   May 28th, 2010 at 6:17 pm

   Sorry, should have included a key to the abbreviations! "oppDPAA" is opponent defensive points above average; it's league defensive rating minus team defensive rating. When there's an underscore, that's what it's weighted by -- "_mp" means it's weighted by minutes played in each game, and "_poss" means it's weighted by the individual possessions used in each game. So if you weight the average opponent Kobe's faced in his playoff career by the possessions he used in each game, their defensive rating has been 2.79 points/100 possessions better than the league average.

   As for a "crucial game" definition, like I said, there's definitely room for argument. But I think facing elimination, even if you're down 3-0 or 3-1, is still a crucial game (it's literally a "must-win"). Also, understand that a game like that only counts as a "crucial game" for the team trailing. So Game 5 of the Celtics-Magic series counted as a crucial game for Orlando, but not for Boston.

6. UB Says:
   May 28th, 2010 at 9:23 pm

   Neil,

   I get what you mean, but one could argue that every playoff game is "crucial" (certainly, it's true in announcer-land, which is the odd and scary realm where all broadcast cliches are true). I don't remember the exact percentage of times the team that wins Game 1 wins the series, but it's something like 80%, correct? So the delta (value) between 0-0 and 1-0 is 30% increased likelihood of winning the series, assuming that there's no inherent advantage in being tied at 0. Meanwhile, obviously as of right now being down 0-3 has a 0% conversion rate to having won the series, though the Magic are certainly trying to change that. But being down 1-3 also still has a very low conversion rate (it's only happened, what, 8 or 9 times ever?), so the delta (value) of winning game 4 there is very low.

   To take a current example...

   Imagine an existing player. We'll call him JaLon Brames. JaLon's an amazing player, but in the fourth game of a series in which his team is totally outmatched, he goes 10/30 from the field, just 2-6 from the line, for 24 points (TS 40% - eugh), with 6 turnovers (though he does get 10 assists and 6 boards to mitigate that somewhat). However, he's not particularly called out for the game, given that it's very unlikely his team will win the series anyway, and their playoff opponents have an unbelievable defense.

   Fastforward a couple of years, with Brames's team now down only 2-3. JaLon now goes 8/21 from the field, 9/12 from the line, for 27 points (TS 64% - quite good), though his 9 TOs hurt his team a ton. But his 10 assists and 19(!) rebounds are certainly a plus for his Javaleers in a losing effort (big coffee conglomerate ownership). After the game he is absolutely KILLED in the media, even though this game was pretty significantly better - simply because the loss of this game was seen as far more damaging given that a win would have set up a game 7 for the Javaleers at home, a much more likely possibility for a victory in the series than still being down 1-3.

   I'm not saying...
   I'm just saying.

7. Neil Paine Says:
   May 28th, 2010 at 11:54 pm

   So what you'd like is basically a "leverage index" within a series for how important each game is based on its location the series score going in. I can actually do that -- look for a post next week.

8. UB Says:
   May 29th, 2010 at 12:05 am

   Neil - that sounds really interesting. Thanks for considering it!

9. Hank Says:
   May 29th, 2010 at 12:37 am

   Very interesting post, but it appears that the Eastern Conference is given a little too much credit defensively, and the Western Conference not enough. At least part of the reason for the high ratings given to Eastern defenses is that they are facing weaker offenses more often throughout the regular season.

10. themojojedi Says:
    May 29th, 2010 at 1:21 am

    Great work in following up this topic Neil, I really like the breakdown weighted by both minutes and possessions.

    If I've understood the definition of terms correctly, this methodology treats an opponent with a defensive rating of 106 in a year where the league rating is 108 as an equally tough defensive opponent as a team with a defensive rating of 100 in a year that the league average was 102. This is clearly not the case. I've only really looked at players in the mold of high-usage perimeter scorers, but those guys absolutely FEAST on defenses with ratings greater than 106 and even the greatest have struggles against the teams with ratings less than 100.

    For this reason, when I was originally having a look at this topic I decided not to assess the opposing team defense against the league average, but rather use the raw defensive ratings. Why? I think you need to directly assess the statistical productivity of a player in the statistical environment in which it was produced, and that environment has fluctuated significantly between 1991 and 2010.

    If we look at a player's Offensive rating as a return on investment, then a 10% return on investment is much more valuable during the lean times of the Global Financial Crisis than during periods of economic prosperity. And for NBA scoring efficiency, 1999-2004 was the Global Financial Crisis.

    During Jordan's first 3-peat the yearwise league average ORtg/Drtg was around 108 in the regular season and 110 in the playoffs. Scoring was cash money, and MJ logged around a 32% of both his career playoff games and minutes during this span. Things remained fairly fruitful in the next 5 seasons, with a yearly average rating of 106.8 in the regular season and 107.1 during the playoffs. MJ played approximately another 38% of his career playoff games and minutes during this stretch.

    From the start of the Tim Duncan championship season in 1999 to the End of the Shaq/Kobe Lakers era in 2004 the yearwise league rating was 103.4 in the regular season and 103.0 in the playoffs. You might as well have hid your money under the mattress, because times were tough. These 6 years were the 6 lowest for ORating/DRating in the last 20 regular seasons and included the lowest 5 for the last 20 playoff seasons. Duncan played around 43% of his career playoff games during this stretch and 45.47% of his total playoff minutes, Shaq around 46.7% (51.5% of minutes) and Kobe 52% (and 56% of his total playoff minutes).

    Since then things have picked up again with average ratings of 106.2 regular season and 107.0 playoffs from 2005-2007 and 107.8 regular season and 108 playoffs from 2008-yesterday. Obviously the modern era of stars like Dwade, LeBron and Chris Paul have played all or nearly all of their playoff games during these years when the points are flowing freely again.

    I really want to write a detailed analysis in full when I find the time, but consider this brief breakdown of the best and worst opponent defenses for just Jordan, Wade, LeBron and Kobe. (These include series prior to 91 and were compiled before the 2010 playoffs, so slightly out of date).

    These 4 guys have played a series against an opponent with a defensive rating better than 102 on 27 occasions in their careers:
    3 times for Jordan - 3 series wins against 93 NY, 97 MIA, 98 IND, includes No Finals, 1 series against ratings better than 100
    3 times for Wade - 3 series losses against 04 IND, 05 DET, 07 CHI, includes No Finals, 2 series against ratings better than 100
    3 times for LeBron - 3 series losses against 07 SA, 08 BOS, 09 ORL, includes 1 Finals, 2 series against ratings better than 100
    18 times for Kobe - 14 series wins, 4 series losses, includes 5 Finals, 13 series against ratings better than 100 including 4 Finals

    These 4 guys have played a series against an opponent with a defensive rating worse than 106 on 29 occasions in their careers:
    16 times for Jordan - Includes 1 Finals (2 other Finals in the 105s, 2 more in the 104s), 6 series against rating worse than 108
    2 times for Wade - Includes 0 Finals (06 DAL at 105), No series against rating worse than 108
    6 times for LeBron - Includes 0 Finals, 3 series against ratings worse than 108
    5 times for Kobe - Includes 0 Finals, No series against ratings worse than 108

    We can also breakdown the defenses faced by alternative statistical measures such as EFG% defense versus the player's EFG%, or look at the quality of defenses as measured by the All-D teams or DPOY awards of the direct man on man perimeter defenders faced (GP, Kidd, Starks, Bowen, Majerle, Christie, Prince, Battier, Artest etc) and the intimidating shot-blocking and help D interior presence faced (Ben Wallace, Duncan, KG, Dikembe, Ewing, Oakley, Robinson, PJ Brown, Zo, The Davis boys etc). For these awards I like to use a 5 year rolling window including awards achieved in the season that the player faced that defender and the 2 years before and after, but that's another story.....

11. hk Says:
    May 29th, 2010 at 3:58 am

    I think Neil was trying to account for players that under perform against soft defenses as well.

    MJ didn't win titles in all top 6 of his post-seasons (going by WS/48). Teammates do carry even the greatest players. "Jalon Brames" shouldn't feel too bad, he's generally over performed in his playoff career.

12. hk Says:
    May 29th, 2010 at 4:15 am

    Solid discussion between the both of you, by the way. These last couple of debates and blog posts have been interesting.

13. Mike G Says:
    May 29th, 2010 at 6:12 am

    If we don't try to measure a player's "offense" and how it stacks up an opponent's "defense" -- rather, check the player's playoff Win Shares (and WS/48) vs his opponents' SRS -- we can avoid a lot of these uncertainties.

    In regular seasons, opponents' SRS total very close to zero, W% close to .500. In playoffs, they are mostly > 0 , equivalent to an avg W% of .600-.620 . We won't get as many WS/48 vs .600+ teams; if we do, we've overachieved.

    WS and SRS seem to go back to the beginning of time, so we could do so with this assessment, too. The only assumption would be equal Min/G vs each playoff opponent.

    I'd tend to agree that all playoff games are equally vital. After Jordan got 63 against Boston (losing), he was pretty wiped out for the finale.

14. Jayson Says:
    May 30th, 2010 at 9:40 am

    Excellent post, themojojedi.. I love the idea for this blog post, but I think the implementation could have been done much better.

    I've made a few postings asking this to be done for a while now, and I'm glad to see it finally being done..

    Not sure I like how it's been done by possesssion though.

15. nimble Says:
    May 30th, 2010 at 10:19 am

    KB showed he has been/will the man,however advanced naphta you throw at him.kb=mj,en of story.

16. nimble Says:
    May 30th, 2010 at 10:20 am

    will be**

17. UB Says:
    May 30th, 2010 at 1:33 pm

    Hk -

    I agree, "Jalon's" performance has been excellent. I was more trying to highlight the difference in reaction to a game that was really quite poor (G4 2006) and one that really wasn't nearly as bad as people seem to be treating it (G6 2010), which I believe has a lot to do with the situation in the series (0-3 vs 2-3).

18. Anon Says:
    May 30th, 2010 at 1:34 pm

    Themojojedi,

    You presented some thoughtful and accurate points. However, it's not like this blog HASN'T visited the "How would a prime MJ do against today's defenses?" topic before, and he has still come out ahead of anybody you want to compare him to in a contextual application of the numbers (such as the points you presented in your post).

19. Neil Paine Says:
    May 30th, 2010 at 5:39 pm

    Themojojedi,

    You can't look at raw defensive ratings w/o regard to the league average, because it completely ignores rule changes, changes in playing styles, etc. that have occurred over time. For instance, our best estimate of the league average defensive rating in 1958 was 89.1. By your logic, that means every team in the NBA that year was significantly better defensively than the best defenses of the past 25 years! But we know that's not necessarily true -- it's just the way the game was played back then, the strategy was to run down the court as quickly as possible and jack up the first shot you could get (pace factor in 1958: 118.9!!). If we compare to the league average, we still see that the 1958 Celtics (DRtg: 84.9) were a good defensive team, but they were more like a team with a 103 DRtg in today's game, not a team with an 85 DRtg.

    Now, it's true that a DRtg of 100 in a league where avg = 102 is slightly more valuable than a DRtg of 106 where avg = 108, simply because when the scoring environment is lower, 2 points of efficiency differential is more valuable. So we could plug the numbers into the pythagorean formula and come up with a slightly more accurate read on the relative value of defense. But you still have to compare to the league average somehow.

20. Neil Paine Says:
    May 30th, 2010 at 5:43 pm

    Really, Jayson, "the implementation could have been done much better"? How would you have done it? (Keeping in mind that I've already explained why it's necessary to compare opponent's defensive ratings to the league average.)

21. Neil Paine Says:
    May 30th, 2010 at 5:48 pm

    In fairness, I should have included the league average ratings in the final table to compare the offensive ratings to as well. But the main emphasis of the post should be on the tables that preceded the final one; I simply added that last one because the data was on hand and I thought it would be cool.

22. UB Says:
    May 30th, 2010 at 5:55 pm

    Neil,

    Just to back themojojedi on one point...

    Your production metric of choice is Game Score, is it not? To quote from the glossary: "Game Score; the formula is PTS + 0.4 * FG - 0.7 * FGA - 0.4*(FTA - FT) + 0.7 * ORB + 0.3 * DRB + STL + 0.7 * AST + 0.7 * BLK - 0.4 * PF - TOV."

    In other words, it's a counting stat. So I can kind of understand his concerns with not normalizing the data. Because a game score of 15 during the 1958 season just isn't the same as a game score of 15 during the 2010 season.

    Also, unless I'm misreading it, he's also claiming that there's a significant kink in the graph at DRtgs of 106 and 100.

23. Neil Paine Says:
    May 30th, 2010 at 7:06 pm

    It was the metric of choice for the post on Thursday, yes. Would I use it for anything but the most simplistic of studies? No. I don't even prefer to use PER, and Game Score is the poor man's version of PER. As for ratings of 106 and 100, I'll have to investigate that. I suspect that there's nothing magical about those benchmarks, except that in the years we have data, the league rating has stayed basically in the 106-109 range, so it just seems like those are important numbers. Basically, he's just saying that good scorers do best against below-average defenses and struggle against the best in the league.

24. themojojedi Says:
    May 30th, 2010 at 7:54 pm

    Neil,

    I think there are many situations when comparing to the league average is necessary. The example you presented of determining how fundamentally "good" a defense is between years and eras is one of them. There it is particularly important to understand the effect of rule changes, the 3-point line, pace, style of play etc. That is, how well was a team able to work within the confines of the prevailing environment to produce statistically tough defense.

    My logic does not assume that a lower team defensive rating means a "better" defensive team. My logic assumes that a lower team defensive rating means that the team has demonstrated over the course of the regular season that it more difficult for the opposing players to score a high number of points per possession against them than a team with a worse defensive rating. This is regardless of the year,era or league average rating of the two teams. We should statistically evaluate how a player actually does perform offensively directly against the statistical standard of defense.

    I don't think the 1958 Celtics (DRtg 84.9) defenses are fundamentally better defensively than the 2008 Celtics (DRtg 98.9). Even if the two teams had a DRtg an equal number of points below league average DRtg for their season I'm still not judging Bob Pettit for shooting 42.3% on field goals (and whatever his corresponding ORtg would have been) in the 1958 Finals series against Boston as harshly as I would judge Kobe for shooting 40.5% or LeBron for shooting 35.5% in their respective playoff series against the 2008 Celtics.

    I'm willing to be convinced otherwise, but I don't think that 'points better than league average' is an appropriate comparative measure of how tough it is to produce numbers against teams with low defensive ratings. People say defense was much tougher during the 1990's hand-check era, and physically it probably was. But is that toughness reflected in the Defensive ratings of the top teams of the era?

    Don't worry Anon, I'm not trying to stir up anything against Jordan. To me he is the best player that has ever played the game, at least that's what the Jordan memorabilia plastered all over my walls seem to reflect.

25. themojojedi Says:
    May 30th, 2010 at 8:05 pm

    Regarding the 100 and 106, no magic numbers there. I had just been dividing the team DRtgs into classes above and below 100 points per 100 possessions with a width of 2 since that's what a field goal is worth.

    And during these playoffs Kobe finally got to play against a DRtg worse than 108 as Jordan had done 6 times and LeBron 3 times. I think I'd classify his series against the Suns (110.2) as an offensive feast, right?

26. DSMok1 Says:
    May 30th, 2010 at 8:40 pm

    Where did APBRmetrics go?

    Neil--you had mentioned some source data for APM/SPM regressions? I'm interested...

    BTW this is an interesting discussion above.

27. themojojedi Says:
    May 30th, 2010 at 9:01 pm

    Should have reported the TS% for the Pettit, Kobe and LeBron example too:
    Pettit: 0.515 TS%
    Kobe: 0.505 TS%
    LeBron: 0.480 TS%

28. Neil Paine Says:
    May 30th, 2010 at 10:07 pm

    Themojojedi- I think you and I are on the same page, it's just that I was looking purely at the strength of opposing defenses without regard to player performance in this post, while you are talking about combining performance and SOS into an adjusted production rating. I can do that too, though, either next week or the week after that, using the pythagorean formula to adjust the player's base offensive ratings for strength of opposing defenses.

    DSMok - Yeah, I have no idea what's wrong w/ APBRmetrics, but I think the gist of my comment was that I'm interested in helping/collaborating with the new advanced-stat-based SPM it looked like you were working on (I think I missed your intro for that). I can send you the APM dataset I used for the most recent regression if you want (it's the Rosenbaum data + 2 yrs. of Lewin data + recent BBValue data; no Ilardi because Steve combined playoff numbers w/ regular-season ones). What was your R-squared on the first pass? Better than what we were getting w/ the Rosenbaum per-40 method?

29. Jayson Says:
    May 30th, 2010 at 10:33 pm

    Neil, when I said I thought it could be implemented better, I meant that performance should be taken into account with the team's defensive quality. Maybe a weighted scale or something where if facing a subpar defensive team - the player's value in terms of ortg or PER should be decreased to account for the defense they were facing, and vice-versa for great defenses. Not sure how it would work out, since I haven't got the greatest grasp on statistics, but I'm interested in seeing what you come up with.

    But I guess you already sorta came up to that point in the last post..

    - J

30. DSMok1 Says:
    May 30th, 2010 at 10:37 pm

    I posted a long reply on APBRmetrics, if it'll come back.

    R^2 are much higher, but I'm regressing against the low-noise 6 year average APM's of Ilardi, so there is no direct comparison.

    I'm using only context-neutral stats in the new regression like AST% and ORB%. I'm liking how that' turning out so far.

    I took the scoring/turnover term in a nonlinear direction, posing it as ((TS%*2*(1-C_1*TO%))-PPP_Threshold)*USG%*C_2, where C_1, PPP_Threshold, and C_2 are terms found by the regression. This contextualizes the turnovers to some extent. PPP_Threshold is in the .88 vicinity in my initial regression, and the C_1 term is about 1.35 (indicating that a turnover is worth 1.35 points). If I can figure out how to include assists both to and from the player within this same term, I think we're seriously on to a good regression.

    I would love to have the additional data for additionally refining the regression.

    A cut-down version of my spreadsheet is at: http://spreadsheets.google.com/ccc?key=0Ah1NfCUslJwxdEhreTd4OGN3SzU1b2U5YXNmTG1JeVE&hl=en

31. themojojedi Says:
    May 31st, 2010 at 12:13 am

    Okay Neil, I have been viewing this post as a follow-up on/extension to the performance in crucial games post which did make claims about player performance using GameScore, as does including the player's ORtg in the final table of this post.

    On an unrelated matter, I'm assuming there has been much discussion here or at APBRmetrics about the topic in years past, but what is the motivation for the Version 1.1 change to WinShares which removes the sum to team Wins adjustment? Is a team level adjustment viewed as too ad hoc and arbitrary in nature? Should Win Shares be titled something less snappy such as Efficiency Differential Shares?

    By the way I love Win Shares, so I'm just trying to understand the motivation and the implications of the refinements made to the methodology.

32. Ryan. Says:
    May 31st, 2010 at 4:22 am

    http://www.basketball-reference.com/blog/?p=3600

    I believe everything you're looking for can be found there, Themo.

33. Ryan. Says:
    May 31st, 2010 at 4:24 am

    Would've helped if I had read that post correctly... FML.

34. hk Says:
    May 31st, 2010 at 10:52 am

    I love Win Shares.

    Coincidentally LeBron has just over 25 win shares (regular and post-season) in a single season, slightly besting Jordan's peak. I don't think he has a problem against # 1 defenses, he destroyed the Magic last year and was still respectable against the Celtics this year.

35. Ryan Says:
    June 5th, 2010 at 1:12 am

    Concerning the recent back-and-forths we've had regarding defense, I'd like to chip in.

    While I agree that adjusting for team-defense is an aspect of the utmost importance, I do believe there will always be a gigantic flaw when measuring the effect they have on players.

    In no way are we immediately able to quantify that during the Laker's first 3-peat, defenses were built and assumed around Shaquille O'Neal, who saw most of the attention and majority of the doubles. They adjusted dependent on approach and performance. These numbers make an adjustment based on the team's defensive performance, but those numbers exist because the defense gave Shaq a hard time than they gave Kobe.

    Likewise, these numbers don't show the Detroit Bad Boys being the best defensive team in the league. However, they were most certainly the best team and defending and containing Michael Jordan. They'd created and adapted the most comprehensive basketball scheme against on player in basketball history. 15 defensive schemes to contain Jordan. Surely, that defense has to have a higher weighting on Jordan's performance than the next.

    Similarly, while Lenny Wilkins coached some pretty good defensive teams, he was unequivocally the worst coach I've ever seen at scheming against Jordan.

    The one team that stands tall, much like the Boston C's, in this regard are the New York Knicks, circa '92. Top defensive team, and probably the best defensive approach I've ever seen against a single player.

    Over the course of an 82-game season, it's quite possible that the '08 Celtics were better than the '92/'93 New York Knicks. But they certainly didn't use the amount of men to stop one man.

    While not the same squad, I've always loved this clip:
    http://www.youtube.com/watch?v=C4kmu4DS6vA

    And Neil, I think you'll like this. Amongst YouTube's laymen, there has been a defensive argument going on quite heavily for quite a while now. One Kobe fan posted this video, following the '08 Finals loss.

    Kobe Bryant vs Zone Defense:
    http://www.youtube.com/watch?v=4kYBeNQdSCc

    While another poster made a sarcastic, satirical reply with this
    Michael Jordan vs. 1-1 isolation defense
    http://www.youtube.com/watch?v=2s9_GKFNL9E

    Things like this make stepping away from the stat-sheet quite fun, and actually keep my somewhat sane.

36. Ryan Says:
    June 5th, 2010 at 1:14 am

    Apologies for the abundance of typos.

37. Ryan Says:
    June 5th, 2010 at 1:42 am

    Also, at the 2 minute mark of this video, Chuck Daily walks you through an aspect of The Jordan Rules

    http://www.youtube.com/watch?v=XcW0JrStspE