Last October this was quite the debate. Just to recap:
- The media and coaches have always liked Durant. He was a top pick in the draft and voted to the All-Rookie team by the coaches. And the media said he was Rookie of the Year in 2008.
- John Hollinger’s Player Efficiency Rating has also always said Durant was “good”. Both PERs and the media are heavily influenced by a player’s point totals, so it’s not surprising to see PERs and the media agree (and the same can be said for coaches).
- Wins Produced said Durant was below average as a rookie, but improved dramatically in his second season. How did he improve? His shooting efficiency, rebounding, and steals went up while his turnovers went down. In other words, Durant improved with respect to the primary box score statistics that increase wins.
- Despite the improved box score numbers, though, adjusted plus-minus insisted that Durant was below average in both his first two seasons. The difficulty with this measure is that it never tells you why a player is “good” or “bad”. But since Durant had very good box score numbers, it was argued that Durant was actually very bad at on-the-ball defense.
Of course, there was another potential explanation. Adjusted plus-minus is a very inconsistent measure. The year-to-year correlations are very low (there is a great deal of “noise” in the model). So it could be Durant’s adjusted plus-minus was due to his supposedly awful defensive skills. Or it could be noise in the model.
After 15 games this year it is beginning to look like “noise” was the culprit. According to BasketballValue.com, Durant’s adjusted plus-minus score is 16.09. This is the second highest mark on the team. So it looks like everyone is now in agreement. Durant is officially a “good” player (ain’t that a relief?).
Fortunately for the Thunder, he is also not alone. Here are the top producers of wins in Oklahoma City [WP48 = Wins Produced per 48 minutes].
Kevin Durant: 579 minutes played, 0.204 WP48, 13.5 projected Wins Produced
Thabo Sefolosha: 491 minutes played, 0.182 WP48, 10.2 projected Wins Produced
Russell Westbrook: 512 minutes played, 0.138 WP48, 8.0 projected Wins Produced
James Harden: 296 minutes played, 0.219 WP48, 7.4 projected Wins Produced
This quartet are on pace to produce 39.0 wins. As a team, Oklahoma City has an efficiency differential (offensive efficiency minus defensive efficiency) of 2.4, a mark that projects to about 47 wins across an 82 game season. So the four players listed above are responsible for about 83% of the team’s wins.
Missing from this list is any player at power forward or center. Jeff Green remains a below average player. And Nick Collison and Etan Thomas have yet to produce much in 2009-10. D.J. White, though, has posted a 0.423 WP48. Before Thunder fans get too excited, White has only played 44 minutes and has yet to play more than 15 minutes in a game (or more than two games in a row).
Despite the problems in the frontcourt, it does look like Durant is a “good” player. And because he has three other “good” teammates, it looks like the Thunder will be “good” this year.
Let me close by noting the play of James Harden. So far Harden is only averaging 19.7 minutes per game. But if his WP48 of 0.219 continues – and his minutes per game don’t change – he will produce 7.4 wins by the time the season ends. Such production could be what pushes the Thunder past the 41 win mark; or the mark that divides “good” teams from “bad” teams. Harden probably may not play enough minutes to warrant consideration for Rookie of the Year (he plays the same as Thabo Sefolosha). But in a discussion of “good” rookies in 2009-10, Harden’s name should be mentioned.
– DJ
The WoW Journal Comments Policy
Our research on the NBA was summarized HERE.
The Technical Notes at wagesofwins.com provides substantially more information on the published research behind Wins Produced and Win Score
Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:
Simple Models of Player Performance
What Wins Produced Says and What It Does Not Say
Introducing PAWSmin — and a Defense of Box Score Statistics
Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.
Ray
November 25, 2009
I thought Serge Ibaka would be high on this list, despite his limited minutes. It seems like he’s killing with rebounding. Also, I’m surprised to see Westbrook above average. I know his turnovers are high, and it seems like his shooting efficiency is only medicore. But it’s good to see that he’s actually more productive than I though.
So according to the numbers, with healthy players, the Thunder should get to the 47 win mark?
Ray
November 25, 2009
I just checked the numbers and Westbrook is only shooting 42% and Harden is under 40%. Westbrook is among the leaders in turnovers at 3.8. Where are these two making up ground to be “good” players? It seems WP would kill these guys for turnovers and poor shooting.
ilikeflowers
November 25, 2009
Westbrook has 230 points on just 190 shots for 1.21 pps, he’s a very efficient scorer.
Harden has 123 point on 106 shots for 1.16 pps, also good.
Durant has 1.31 pps.
ilikeflowers
November 25, 2009
Here are the top 10 PG’s by PPS
Paul……….1.66
Nash……….1.46
Bynum………1.4
Billups…….1.35
Calderon……1.31
M. Williams…1.25
Evans………1.25
Westbrook…..1.24
Jennings……1.23
Ridnour…….1.22
sportsfanatic613
November 25, 2009
There have been numerous articles written about the Oklahoma Thunder, but no mention has ever tried to analyze why their GM, Sam Presti drafted Jeff Green over Joakim Noah. For all the good that is written about Presti, had he drafted Noah instead of Green, a lot of the Thunders problems would have been solved. Playing Noah at either power forward or center instead of Green at either small or power forward would been the wiser decision. Why was Presti so high on Green? Didn’t Presti realize that having drafted both Green and Durant that there would be duplication of their talents? Noah would have been a much better compliment to Durant.
Italian Stallion
November 25, 2009
I apologize for the hijacking the thread, but this article is too good to not post.
It’s about how the standard position assignments don’t always reflect how the players play on offense. This is what I’ve been saying about Bargnani and others for awhile.
http://www.hoopdata.com/recent.aspx?aid=49
Rob O'Malley
November 25, 2009
I’ve always been a huge fan of James Harden. Many draft people thought he had the game of Manu Ginobili, now I can see why. If his WP48 is .200 now, imagine how high it will be when he begins to hit shots. Also Durant hasn’t been shooting as well as we saw last season. If the Thunder big time players begin to make their shots to the level they are capable (Durant and Harden in particular), they might be able to win 50ish games.
Italian Stallion
November 25, 2009
Ilikeflowers,
Can you calculate the average PPS for the league by position? For starters at each position?
Rob O'Malley
November 25, 2009
Also, because Greg Oden is playing so well, I think an Oden/Durant comparison would be a good post.
I haven’t looked into this much but off intuition the Oden/Durant comparison could be similar to the Jordan/Rodman discussion. In that while Odens WP48 is higher than Durants, if we look at SD about the average then we see that Durant is better. Therefore popular opinion is vindicated? I’m not positive which player is more productive, I am just speculating that this may be the case.
In any case I think it would add an important dimension to posts in which you do comparisons between players who play different positions to mention their difference in SD. It makes things make a lot more sense in some cases.
ilikeflowers
November 25, 2009
IS,
You can get a good idea from this link
http://sports.espn.go.com/nba/statistics?stat=nbafgpct&season=2010&seasontype=2
you can filter it by position, rookies, and internationals
Italian Stallion
November 25, 2009
IMO Green is playing a level or two below his level of last year after making a lot of progress between year one and year two. Perhaps, we should give him a few more games before making too hasty a judgment.
There’s another potential problem.
I assume he’s being categorized as a PF (correct me if I am wrong), but his game is nothing like the traditional PF’s game. He’s only 6′ 9″ and plays more like a SF than a PF. He spends a lot of time on the perimeter, doesn’t board as well, or block shots as well etc…
http://www.82games.com/0910/09OKC9.HTM
You can argue that using him as a PF hurts the team, but I would argue that that is an issue of coaching and team personnel and not an issue with Green’s game. I’m not arguing that he’s a top player, but I think he’s better than the rating he gets here.
ilikeflowers
November 25, 2009
Top 10 league-wide
Howard…….1.9
M. Gasol…..1.75
Paul………1.66
Hilario……1.58
Oden………1.58
James……..1.57
Pierce…….1.55
Stoudemire…1.54
Bosh………1.53
Speights…..1.51
ilikeflowers
November 25, 2009
IS,
EstWP48 for Green:
0.020 as PF
0.120 as SF
ESPN just has him listed as a Forward. Sounds like the Sonics ;0) are in bad need of a PF.
Tom
November 25, 2009
First of all,
“After 15 games this year it is beginning to look like “noise” was the culprit. According to BasketballValue.com, Durant’s adjusted plus-minus score is 16.09.”
The stats linked from that site are not through 15 games. Which those two sentences seem to imply from their proximity.
Second of all ‘large but not large enough sample size A is probably misleading, just look at much smaller sample size B” really?
Celtzfan8617
November 25, 2009
Any chance we can get a post on the struggles of the Boston Celtics, maybe a response to the ESPN article “What’s wrong with the Celtics?” (http://sports.espn.go.com/nba/columns/story?columnist=sheridan_chris&page=celtics-091122) They focused on KG, Sheed, Rondo, and Ray.. does that measure up with what wins produced sees as the problem?
Ray
November 25, 2009
I hate to have you guys answer this for the 100,000th time. But I want to understand this model, so I don’t dismiss it. Since this is regressed from team statistics, team statistics that are most condusive to winning are allocated to each player. Statistics that correlate to wins on the team level are mainly good FG%, low TOs, and rebounds. So, is that why this model has a lot of high ranked centers and power forwards? Is it fair for me to say that a guy like Kobe is still a more skilled than a guy like David Lee, but Lee plays a more efficient position? And didn’t Delonte West and Kobe have similar WPs last year? Where does Kobe slip up in this model. I thought his FG% (usually in the 46-47% range) is actually good for a SG.
My other question would be, how would the model rate a guard who shot 15 times a game and made 48% (very good) versus a guard (or whatever) who shot 5 times a game for 55%. Is volume of efficient scoring factored in?
Ray
November 25, 2009
By the way, I really dislike Kobe, so I’m not complaining. He’s just the superstar that is always rated lower than he’s perceived to be. And I genuinely want to know where in his statistics he falls behind LeBron Wade and Paul so much.
Rob O'Malley
November 25, 2009
A Kobe and Wade comparison
todd2
November 26, 2009
I’m glad to see Harden having some success. I’ve always been a fan of ‘tweeners like him because they can cause matchup problems; bigger than a lot of 2’s and quicker than a lot of 3’s. The issue with players like him is how well they adjust. They’ll thrive if they acquire skill sets for multiple positions and their coaching staff puts them in positions where they can succeed. I think Sam Young comes from the same mold but may have less upside because he’s four years older.
ilikeflowers
November 26, 2009
EstWP48 Says that the issue is KG, Wallace, and House. Rondo is their most productive player.
Paul Pierce……..0.25
Ray Allen……….0.08
Kevin Garnett……0.17
Rajon Rondo……..0.29
Kendrick Perkins…0.14
Rasheed Wallace….0.01
Eddie House…….-0.02
Marquis Daniels….0.1
Shelden Williams…0.28
khandor
November 26, 2009
1. There is little to no duplication of skills in the form of Jeff Green and Kevin Durant.
2. The collection of players which the OK City now has on its roster is substantively different than last year’s group [which begin the season]:
Starters
Westbrook/34.1 MPG, 15 Gs
PG and OG
Sefolosha/32.7 MPG, 15 Gs
OG and SF and PG [new add]
Durant/38.6 MPG, 15 Gs
SF [Def] and OG [Off]
Green/33.9, 15 Gs
PF and SF
Krstic/23.5, 15 Gs
C [new add]
——————
Key Subs
Ollie/14.8 MPG, 14 Gs
PG [new add]
Harden/19.7 MPG, 15 Gs:
OG and PG and SF [new add]
Collison/17.8 MPG, 13 Gs
PF and C
Thomas/16.1 MPG, 15 Gs
PF and C [new add]
Ibaka/12.8 MPG, 8 Gs
C and PF [new add]
——————
Extras
Livingston, PG [new add]
Weaver, OG/SF
White, PF-SF [new add]
Unfortunately, trying to understand how exactly the game works, strictly by-the-numbers, won’t really provide you with a great deal of valuable insight to the dynamics of a successful TEAM.
simon
November 26, 2009
Ray//
The rationalle for that has been explained in dberri’s book. The gist is: talented big men are in a short supply and that’s why even after adjusting for the positional averages, there are usually more big men ranked high in this system. There are more gap amongst big men because, well, big men are harder to find.
One way to adjust for this is to use the standard deviation above average instead of just raw numbers and once you do that, the bias toward big men lessens. It’s explained here: https://dberri.wordpress.com/2009/09/24/two-perspectives-on-the-top-10-across-the-past-decade/
But to make it short, Kobe just isn’t that great in terms of all time SG by any statistical measure, and the WoW system isn’t the only approach saying that. The systems mostly agree that he is one of best recent SGs , but fall short of historical greatness. He’s certainly no match for Jordan or Magic Johnson in this system.
Ray
November 26, 2009
Thanks simon, that pretty much answered everything I wanted to know.
jbrett
November 26, 2009
Dr. Berri,
I know that Iverson has announced his intention to retire; however, I want to caution you against a post putting a bow on his career just yet. I believe that, now that he has informed the world he plans to take his ball and go home, we will discover there are still people in the NBA who can find a spot for a lead guard with no fear of a 41% FG percentage, and whose determination to prove no one can stop him from launching a shot results in an offense with all the movement of a foosball table. I’m confident that in a few days—what’s that? He’s already un-retired? Um, nevermind; I’ll save the rest for next time.
Italian Stallion
November 26, 2009
Ray,
IMHO, one of the reasons that Cs and PFs are ranked higher than SFs/Gs is that more of what big men contribute to winning is captured in the box score.
If that wasn’t the case, I think we would see successful teams made up of all PFs and Cs. lol
Among good small men, I can think of several possible contributions to winning that have reasonable value that don’t get captured.
1. They get assists, but we don’t capture the value of a great initial pass that eventually leads to a score (kind of like the hockey assist)
2. Some shoot well from the OUTSIDE and help space the floor for big men and others that like to drive to the hoop. Unfortunately, outside shooting tends to LOWER efficiency because the shots are harder and you don’t get fouled as often. So something that can be a net positive and that is often a requirement actually impacts them negatively statistically relative to big men even though they have the superior shooting skill.
3. They dribble the ball, get it to the other end of the court, break presses etc…
These are the kinds of things that positional adjustments try to correct for.
However, as a stated earlier, what do you do with a PFs that play like SFs?
What do you do with a guy like Dirk Nowitski?
Dirk is a big man that can score inside or outside, from beyond the ark, handle the ball, pass well etc… And he does some of those things even better than most Gs. To me, he is almost certainly underrated.
dks
November 26, 2009
I want to echo the comment from Tom on 11/25 at 5:30. Why do you decide, after 15 games of one season, to discard two years worth of data? You certainly can’t justify it by saying that 15 games is LESS NOISY than two years! I feel like I’ve seen this reasoning before on this blog, where recent performance is weighted extremely high and lots of past data is ignored.
A full season of this may start to sway me that Kevin Durant is better than we thought. Or it might just tell me that, after a couple years in the league, he’s figured some things out and improved a fair bit this year. I haven’t seen much on basketball aging curves, but my impression is that large, discontinuous jumps in performance are somewhat common among young basketball players (much more common than in baseball, for example).
ilikeflowers
November 26, 2009
After 15 games this year it is BEGINNING to look like “noise” was the culprit. According to BasketballValue.com, Durant’s adjusted plus-minus score is 16.09. This is the second highest mark on the team. So it LOOKS like everyone is now in agreement. [BEGIN SARCASM] Durant is officially a “good” player (ain’t that a relief?). [END SARCASM]
todd2
November 27, 2009
Durant needs to have a conversation with Josh Smith. 3 for 20 vs Portland? Still too many 9 for 21’s and 10 for 24’s. Maybe there’s a pattern for new players: get your points and new contract and then worry about winning.