This post uses the Wins Produced metric. For longtime fans we’ve made a few changes in the offseason, you can read up on them in the walkthrough. For all fans you can find up to date numbers at the NBA Geek.
I listened to a very enjoyable video cast over at Truehoop yesterday with Haralabos Voulgaris (@haralabob) and he had a great line that I can’t agree with more:
You shouldn’t call a player that’s not currently playing at an All-Star level an All-Star regardless of what he did in the past
It is common to consider a player’s legacy when discussing their skill. This season Kobe has not been playing well. However, I have consistently ended up in twitter fights where Kobe’s legacy is included to justify his current performance. The idea that experience helps a player is a bit off.
In Stumbling on Wins Berri and Schmidt showed that players do not age like wine, they age like milk. This especially matters to Kobe, who is 33 this season. Of the 405 players that have suited up this season only 41 of them are 33 or older. In that group of veterans there are some players with some great legacies. In fact if we jump back to Kobe’s MVP season we can see how close these players are to greatness.
| Player | Age | MPG | WP48 | WP | League Rank |
|---|---|---|---|---|---|
| Chauncey Billups | 31 | 32.3 | 0.287 | 15.1 | 4 |
| Marcus Camby | 33 | 34.9 | 0.257 | 14.8 | 6 |
| Steve Nash | 33 | 34.3 | 0.253 | 14.6 | 7 |
| Kevin Garnett | 31 | 32.8 | 0.298 | 14.4 | 8 |
| Jason Kidd | 34 | 36.3 | 0.236 | 14.3 | 9 |
| Kobe Bryant | 29 | 38.9 | 0.194 | 12.9 | 14 |
| Tim Duncan | 31 | 34.0 | 0.233 | 12.9 | 15 |
| Shawn Marion | 29 | 36.7 | 0.253 | 12.2 | 18 |
| Paul Pierce | 30 | 35.9 | 0.195 | 12.7 | 22 |
| Ray Allen | 32 | 35.9 | 0.181 | 9.9 | 30 |
| Dirk Nowitzki | 29 | 36.0 | 0.162 | 9.4 | 34 |
| Vince Carter | 31 | 36.8 | 0.151 | 9.3 | 35 |
| Ben Wallace | 33 | 32.5 | 0.174 | 8.0 | 49 |
Using the WP48 metric a player is generally considered a star if they have a WP48 above 0.200 (twice as good as an average player). They are considered a good player if they have a WP48 above 0.150 (50% better than an average player). In 2008 seven of our thirteen greats were playing at star level. All of them were playing much better than average and all ranked in the top 50 in the league for production. How big of a difference does four years make?
| Player | Age | MPG | WP48 | WP |
|---|---|---|---|---|
| Ray Allen | 36 | 34.9 | 0.334 | 1.7 |
| Marcus Camby | 37 | 25.2 | 0.264 | 1.3 |
| Steve Nash | 37 | 29.3 | 0.222 | 1.2 |
| Ben Wallace | 37 | 12.9 | 0.178 | 0.5 |
| Jason Kidd | 38 | 28.1 | 0.168 | 0.8 |
| Shawn Marion | 33 | 27.1 | 0.153 | 0.9 |
| Kevin Garnett | 35 | 30.4 | 0.126 | 0.6 |
| Kobe Bryant | 33 | 36.4 | 0.115 | 1.0 |
| Paul Pierce | 34 | 30.4 | 0.099 | 0.3 |
| Chauncey Billups | 35 | 30.6 | 0.089 | 0.3 |
| Vince Carter | 35 | 20.8 | 0.066 | 0.3 |
| Dirk Nowitzki | 33 | 31.0 | 0.059 | 0.4 |
| Tim Duncan | 35 | 25.2 | =0.002 | 0.0 |
Excluding Ray Allen, Marcus Camby and Ben Wallace all of our former greats have degraded since their old days. Even Camby and Wallace are hard to give much credit to as their production is around the same on far fewer minutes. All of our players have decreased their playing time (I have no doubt the insane lockout induced schedule has contributed to that). In 2008 over half of our players were stars. Four years later less than half of them are much better than average. In fact, a few have fallen off a cliff.
Here’s a reminder from Stumbling on Wins how we expect players to change year to year as they age.
| Age change | Expected performance change from previous season |
|---|---|
| 23 to 24 | +2% |
| 24 to 25 | no change |
| 25 to 26 | -2% |
| 26 to 27 | -4% |
| 27 to 28 | -6% |
| 28 to 29 | -9% |
| 29 to 30 | -11% |
| 30 to 31 | -17% |
| 31 to 32 | -22% |
| 32 to 33 | -35% |
| 33 to 34 | -57% |
| 34 to 35 | -146% |
Players peak around 25. Up until they’re around thirty their decline is slow. Once they hit thirty-two though their degradation is very swift. All of our greats are now in that range and some of them were already there four years ago.
When players age in sports it’s not a question of if they’ll stop being great it’s a matter of when. Sure some players can defy expectations. However, when we look at our current crop of older players with good resumes in the NBA it doesn’t look good to think that as a group they’ll even stay a shadow of their glory days. People can quote player awards, that they’re clutch (even if they’re not), or that they have experience. If they’re in their thirties though they’re a ticking time bomb waiting to fall apart and that even includes great players. So when someone reminds you that Kobe or Dirk was an MVP just a few seasons ago, it’s not out of line to ask what they’ve done for you lately.
-Dre
I’d love to see Kobe’s WP48 without the 2 Denver games included.
I love this article because it shows just how much WP48 prefers certain positions over others. Who honestly thinks that today Kobe is just barely better than an average player? Or that Billups is actually a below average player in this league? Shows yet again that advanced stats don’t do anything but give a false sense of security in the NBA at this point because they are inadequate at explaining a large portion of the game. There’s no credit for the assist leading to the assist (i.e., hockey assists) or off ball movement or even what type of role a player is expected to play, since many teams ask different things of their players. This wouldn’t be such a problem if people didn’t look at these stats as the definitive word. The stats are helpful, but they’re still in their infancy and people need to realize that.
Tundra,
WP48 is actually one of the few metrics that actually does directly adjust for position.
In terms of defending these players I’m confused. I actually agree that all of these players have been very very good in the past. I believe they are not as good this season.
In terms of choosing stats over anecdotes by the way I stick with stats.
Tundra,
Reading the FAQ before commenting will help you make better arguments.
Also, you want to argue on the basis of anecdotes? OK:
“Who honestly believes X” is an argument that goes back to things like “the earth is round”, “the earth revolves around the sun, not the other way around”, etc. How’d that work out for them?
Dre,
I would argue that using 2010-11 stats is more helpful. It’s entirely possible that Kobe is just having a bad couple of weeks (and that Ray Allen is having a pretty awesome couple of weeks).
Your point remains the same, since Kobe’s production last year was much lower than in 2007-08, but you won’t have to fend off sample-size arguments.
Patrick,
True. However what has been happening with me lately is I say “So and so is not playing well this season” and the retort tends to be “RINGZ!” or “How many All-Star games?” etc. So I figured I’d at least show that how our former stars are playing this season relative to their old selves is far off (with the exception of Allen, what’s up with that?)
“So when someone reminds you that Kobe or Dirk was an MVP just a few seasons ago, it’s not out of line to ask what they’ve done for you lately.”
How can you say something like this? Dirk just won the Finals MVP. ;)
I know, you think that Chandler deserved it, however if you not just count the Finals but the entire Playoffs, Dirk has been the best Mavericks player even by your WP48 measure. and btw. Dirk has been sick in the Finals series (finger + fever) and was still very efficient in the 4th quarter. I know, that doesn’t matter to you either, but asking what he has done for the Mavs recently sounds a little disrespectful to me.
btw… since you recently analyzed the clutch performance of Kobe, maybe you can do that for Dirk as well. Would be interesting.
Is it just me, or is Mike Brown the worst coach for the Lakers? He seems to have given Kobe free reign to do whatever he wants, and that is a recipe for, well, mediocrity.
Compare the big three – http://www.thenbageek.com/players/compare?utf8=%E2%9C%93&player_ids%5B%5D=138&player_ids%5B%5D=137&player_ids%5B%5D=142 – and Kobe is taking over twice as many shots per 48 as Gasol, despite Gasol being a more efficient scorer.
Kobe, if you believe the rhetoric, supposedly beat the Suns single handedly by scoring 48, but he didn’t shoot lights out – he simply shot a LOT, except for one stretch where he hit 7/8 against the Suns second unit.
The Suns made the mistake of rushing Kobe on 3s. The key to beating LAL is to let Kobe shoot threes because he will (51 in 12 games or over 4 a game) and he’s only hitting 19% (meaning he makes between zero and 1 a game). When he gets inside the arc, he’s deadly, but he’ll willingly settle for a three pointer.
I watched the Clippers a bunch and Billups does not look good at all. His WP48 looks spot on, to me. Or maybe even too high.
I’d still like to know Kobe’s WP48 without Denver. He had 2 bad games, with 1 of those being the worst game of his career as a starter IMO. But in every other game, he’s been really solid.
With Denver, he’s shot 46% on the year. Without it, it’s 50%. FT% up too. TOs down. Reb/Assist almost unchanged.
Point is, his WP48 in the 2 denver games were atrocious while in all other games, it’s been much better than his season average, I bet.
sorry for a 2nd post. To respond to the other poster.
Kobe is shooting a bit too much at the end of games. However, Pau isn’t shooting enough by his own choice. It’s frustrating, but he’s overpassing, again. Kobe can’t make Pau shoot. Only Pau can. Bynum is also forcing too many shots.
I’d love Pau to shoot more. But if you are watching the games, you should see him hanging out at the 3 point line or on the weak side not doing much. Other times, it’s Fisher ignoring him. Fisher is useless at this point. Kobe’s been passing it a lot this season. Highest ast% and AST48 of his career.
Paul,
http://wagesofwins.com/2011/06/15/dont-blink-a-review-of-the-nba-finals/
Last season Dirk was ranked 66th in the league with a production akin to Paul Millsap.
In the playoffs he was the second best player behind Jason Kidd for the Mavericks. In the finals he was the 5th best Maverick. I am not discounting what Dirk has done. (I do disagree with both his MVP and Finals MVP for what it’s worth) I am merely stating Dirk is not the player he once was. The fact that Dirk was the best player in the playoffs in 2006 does not change the fact he has degraded and is playing poorly now (which was my point)
I will admit I am not as big of a clutch fan in terms of research. Abbott and the Truehoop folk did all of the hard work for Kobe, I just looked Youtube and Basketball-Reference for perspective. My point on clutch is that it does matter but it is far far less than people think. Kobe is not clutch but that hasn’t stopped him from being a very productive player on some very good teams. Melo has been clutch and he has been an overpaid player on some mediocre teams.
Again, I agree that Dirks overall performance is decreasing, but I still think he has been the most important player for the Mavericks even in the last 5 years. Statistical Plus/Minus shows exactly this and it is very robust for him. So the question remains why Dallas is loosing without Nowitzki on the court and winning with him, if he is only slightly above average.
Is it maybe the confidence fairy? I guess it is not his leadership ;)
So it has something to do with what he is doing on the court. My best guess is, that defenders are afraid of leaving him alone… Yesterday in Boston Dirk was standing at the sideline and Garnet between him and the basket when Haywood was running to the basket completely alone. Guess what Garnet did… NOTHING (which was of course an defending error, but still). So Haywood good an easy basket which looks good in the statistic and Dirk got no credit at all.
I guess your argument is that these things have no large effect since your model explains wins so good. However, if you distribute the wins wrong within a team, it is not surprising that your estimates still work. And if players change teams, you often see substantial changes in your WP measure (even if it is far more stable than statistical +-). Lets see how WP superstar Lamar will perform in the course of the season.
I would also argue that most boxscores just signal good performance which might translate quite well on average, but there can be huge deviations of that.
Like I mentioned earlier, having to shoot in the last seconds of the 24s clock might have an negative effect on your scoring efficiency. Trying to contest a 3-Pointer might decrease your chance for a rebound (and will almost never lead to a block which is also unnecessary)? What is the box score for setting a nice screen? …
I can think of many ways of helping your team on the expense (or with no effect) of your WP score that I wouldn’t dismiss players just because they don’t fair well in the statistic.
Paul,
You seem to be backing up my point. The Mavs have been a great team since Cuban bought them. Dirk is a huge reason why. Last season he was still a big reason why, just not the primary reason. Guess what that resume buys him this year? Nothing. It doesn’t matter that he has been good. It matters that he’s old and decreasing and can no longer carry a team to a championship.
The plus minus stuff frankly doesn’t work. You are really masking many many variables behind one and attempting to explain it. When Dirk gets off the floor there tends to be a reason. Other substitutions take place. Who else goes on the court? All of these matter. It is not enough to say “The Mavs are better with Dirk on than off” without controlling for all the other measures. The second issue with that is many of those other measures just don’t happen enough to get a consistent value. Let’s say Lamar replaces Dirk and covers Steve Nash for 2 minutes. That may happen once. . . ever. That’s too small and too noisy to use as a consistent measure.
In terms of your anecdote there’s a clear problem there. Unless you watch every game for similar stories for every such player it doesn’t really prove anything. What’s worse is that this is an impossible way to judge players. There is simply not enough time to watch every player all the time.
All I am saying is great players stop being great and using their career to explain why they still deserve lots of shots or playing time is not an acceptable excuse to me.
I’m sorry, I haven’t read this blog before, but I had to chime in.
Saying that Kobe or Dirk are anything bug top-10 players is naive. Perhaps Kobe and Dirk aren’t as efficient as other players, but that doesn’t mean they’re worse players. Sure Paul Millsap and Dirk have equivalent production using the WP metric, but Dirk is the focus of the defense when he’s on the floor. Same as Kobe. Great teams always have “Top Dog” scorers that can get their own shot and make plays for others. These players might have lower efficiency ratings than role players like Ryan Anderson, but this is an indictment of the measurement, not an indictment of the player.
Ask the Mavs if they would trade Dirk for Millsap.
dcrowgamer,
Don’t apologize for chiming in! We love debate here! That said calling me naive without much backup doesn’t help too much.
What’s interesting is again I agree there was a time both Dirk and Kobe were very good. They’re not at that level any more. Saying “They help their team because other teams focus on them” doesn’t really mean they’re great players. What matters is if they produce out of that focus. Kobe just decides to take the shot. Dirk actually is pretty good at shooting but he doesn’t do enough of the other roles required by a PF to make up for it.
Also if the Mavericks wouldn’t trade Dirk (slated to make 60 mill over the next three seasons), who is 33 for Millsap (slated to make 15 mill over the next two season) is 26 and putting up great numbers. . . . well I’d laugh very hard at them if they justified it saying “Dirk is a top 10 talent because other teams focus on him”
I see that statistical +- has some issues, especially for short periods of times… But seriously, 5 years in a row buy far the best statistical +- value in Dallas? I can see that not all combinations of players are played together, so you might hide a player who would do quite well if he would play in Nowitzki’s “main” lineup. However, he usually plays in a lot of different lineups for quite some time over the years, so that he has contributed only average can be ruled out from my point of view. Further does statistical +- include the strength of teammates and opponent, which is ignored by boxscores… (which maybe explains why Wright and Williams are so efficient, since they only played in “garbage” time)
And about the anecdote, I didn’t want to prove with it that Nowitzki is definitely producing more points than others like this, just that there is a good chance that this things are very important. And only because it is impossible to get statistics like this, doesn’t mean that it wouldn’t be more precise. And following from that you shouldn’t be so confident about your BoxScore based measure, if you know that it ignores a vast amount of relevant information.
If you are talking salary you should acknowledge that star players are not payed their real value in their prime. So many stars are actually getting long max out contracts to make up for that. From today’s perspective Kobe is already too expensive for what he is doing right now, but he is to a substantial part payed for his past performance. As for Dirk I want to believe he hasn’t passed that peak yet. :)
@ronartest (see what I did there?) the issue, against the Suns (as a Suns fan, the only LAL game I’ll watched), was that Kobe just ran a one man show. He is taking too many shots, and he is turning the ball over a ton. Check his career stats (Patrick you legend for that) and http://www.thenbageek.com/players/compare?career=1&player_ids%5B%5D=137&utf8=%E2%9C%93 show 5 TO per 48 (a career high) and a career high in shots, and his usage rate of 37.0% is nuts (see ESPN http://insider.espn.go.com/nba/player/hollinger/_/id/110/kobe-bryant).
The bigger issue is that Mike Brown seems, and this is purely anecdotal, to be a ridiculously indulgent coach when it comes to stars. Just like King James, Brown seems to LET Kobe take an inordinate amount of shots, and I’m not sure that is the wisest course of action for LA, and the worst Laker’s seasons have been when Kobe has a usage rate that is high, like ’04-’05 – 06-07.
Maybe it is Gasol not being assertive, maybe the coaches’ play calling. All I know is that no one, and particularly Kobe, should have a usage rate this high.
motherwell, i agree and I don’t. Kobe’s usage is too high. true. But this is the result of the Odom trade. Odom was the primary ballhandler outside of Kobe. Right now, there is no real 2nd ball-handler. Blake can do it, but plays only half the game (and often without Kobe). Fisher is terrible. Pau and Bynum or post players or jump shooters, and Barnes is a jump shooter/slasher who cannot dribble. McRoberts is the same as Barnes and Mr. Peace shouldn’t touch the ball, period.
So what is Kobe to do? Outside of Blake, there is no one to handle the ball. This is why the Lakers tried to trade for CP3. Look, I don’t want Kobe’s usage to be this high, but I look at the team and am confused as to who should take some of that away, outside of Pau being more assertive. Kobe is passing more than ever right now, too. The ball is in his hands by default. It’s not Brown’s fault, either.
If you can tell me who on the roster should have the ball to initiate offense, let me know. He should be closer to 34% if Pau would get aggressive, again. But that’s as low it can go without another initiator.
Re- Dirk +/- and stuff. Why couldn’t Dallas win a game without Dirk last season? What is the WP values of players as they play with Dirk and without. For instance, what is Kidd’s WP when Dirk is on the bench? Same with Marion and Chandler?
I am pretty certain if you run this analysis with most players of Dirk’s ilk, you’d find they are much better alongside that player. If true, why is that? And what does that tell us about Wins Produced individual distribution?
regarding Kobe’s turnovers – in his last 6 games he’s averaging under 3 per 36minutes. He screwed up his hand and it took him a few games to adjust, including 8 in the opener. His 5 per 48 is very misleading. This is the problem of looking at stats without context. 2 terrible games can skew the stats for the other 10.
“Saying “They help their team because other teams focus on them” doesn’t really mean they’re great players. ”
I think the point being made was that if opposing defenses chose to stop another Mavs player (or focus on them all equally) they could shut down the role players and reduce their efficiency, but Dirk would go hog wild and his efficiency would skyrocket. So instead, they live with some role player(s) achieving a higher efficiency just to keep Dirk under control because they feel the net is a positive. That has the effect of making the role players look better than they are and Dirk worse than he is.
Whether this impact exists and to what degree it impacts the stats is the tougher question to answer, but if it does, it’s probably true of most of the elite high usage scorers with only slightly above average efficiency.
One thing is 100% certain. Most of the players on the list are beginning to noticeably decline. Some of it may be due to the lockout and players being out of shape, but the rest is father time.
Pingback: Are Celtics fans disappointed in Garnett and Pierce? | The Wages of Wins Journal
Pingback: Is Melo mad that Amaré stole his playbook? | The Wages of Wins Journal
Pingback: Losing your star in a bad trade | The Wages of Wins Journal
Well, this chart is pretty silly right now, but it’s unclear why: you could say that most everyone’s improving their numbers, or that most everyone’s regressing to the mean.
Of the guys performing sub-par, all but Kobe are doing better, and of the 3 doing better than the mean, Wallace and Camby are doing a little bit better still and Ray Allen is doing a lot worse.
Pingback: Podcasting with Chris Yeh on missing Jeremy Lin | The Wages of Wins Journal
Pingback: The Nuggets defy psychology | The Wages of Wins Journal
Pingback: Steve Nash in Rise of the Phoenix | The Wages of Wins Journal
Pingback: What’s up with the Clippers? | The Wages of Wins Journal
Pingback: How the San Antonio Spurs are Moneyballing the NBA | The Wages of Wins Journal