Staff Blog
Berri on Brand, cont.
 |
To catch up start HERE then click HERE and HERE
Be warned that we’re veering dangerously off course here. After this sentence I don’t type word “Elton Brand” at all this post, despite the misleading tag. What I’m after now is the future of basketball and maybe a better understand of the current game. Anyway …
When we left I asked Berri if, because big men produce more wins, if more value should be placed on finding, signing, and developing elite post players and filling the guard positions with role players than searching for talent regardless of position. In other words: are big men inherently more important than guards?
He responds no; one position is not any more valuable than any other:
When we consider position played, it comes out that a slightly above average center is not better than an elite point guard. An elite point guard — relative to his position played — gives you much more than a slightly above average center (relative to his position played).
That in and of itself makes some sense, but I think here am running up against a point that I’m working hard to understand: if no position is inherently more valuable than the next, why are centers “worth” more?
The answer and plenty more ATJ
The standard criticism of Berri’s work is that too much value is placed on rebounds. For example, according to the WoW scores last year Joel Przybilla produced 9 wins and had a win score of .239. Berri ranked him as the 15th best center in the league. I’m fine with accepting that he is a serviceable big man, but according to the rankings he outperformed Tony Parker (7.6 wins, win score of .157), Allen Iverson (9.5, .134), Kevin Martin (8.4, .183), Ray Allen (8.1, .148), Joe Johnson (6.9, .100), and Rip Hamilton (6.6, .131), not to mention a mostly injured Dwyane Wade.
Does that mean that Przybilla is better (or at least had a better year last year) then the all-stars? If so wouldn’t it make more sense to try to add players like Przybilla instead of players like Parker or Martin?
Well, no. Przybilla is not “worth” more than better players, he simply has a higher score because centers score higher on Berri’s ranking. For his career Anthony Mason had a .509 shooting %, which is over ten points higher than Larry Bird (.496) and almost 40 points higher than Reggie Miller (.470) – does that mean that Mason is a better shooter than two of the greatest shooters who ever lived? Of course not. For a similar reason we shouldn’t be bothered by the fact that Joel Przybilla produced more wins than Ray Allen last year. JP (I have two good stories about him, but now isn’t the time or the place) is an average starting center and all the mentioned guards are top 15. Comparatively they’re better.
Berri gives an example:
Eddy Curry gets about five rebounds per game. If he were a point guard, this would be pretty good. As a center, it is awful. If you don’t consider position, you might think that Curry is an okay player. He hits his shots and he scores. But his lack of rebounds and turnovers — which would be okay for a point guard — kill his team. And this is because Curry is not a point guard. He is a center.
His takeaway is that two gaurds should only be compared to two gaurds, SFs to SFs, etc. While I agree that there are inherent problems with comparing an elite SG to an elite C, certainly a comparison of how those two elite players differ from the mean should be fair game.
So lets do that: according to WoW, last year Kobe was worth 16.4 wins and Bonzi Wells, the 15th ranked shooting guard averaged 5. This means that Kobe was worth 11.4 wins more than an average starting shooting guard. That seems reasonable, 11 wins is a lot. That was the difference between 9th place and 1st place in the West last year.
At the same time Marcus Camby was worth 21 wins while Przybilla (the 15th ranked C) was worth 9 – a difference of 12 wins, which, if you’re paying attention, makes Marcus Camby better more valuable even when position is accounted for than Kobe. Now there are a couple explanations here.
The first is that there could just be more servicable shooting guards – certainly there are more talented 6′5″ basketball players than there are 6′11″ basketball players, for no other reason than there are far more 6′5″ people out there. Przybilla may be an average center, but an average center may just be worse than an average two guard. Kobe is really good, but maybe 18 other guys are pretty good. Sure it deflates Kobe’s value a bit, but it also makes the really good centers stand out
Or perhaps the fact that Kobe had to do so much decreased his efficiency. Marcus Camby was never called upon to bail out his team with a low-percantage shot with 3 seconds on the shot clock, like Kobe had to do a half dozen times a game. Miss more shots = get a worse rating, but no one is blaming Kobe for taking those shots.
Still, at the end of the day Marcus Camby was not only worth more wins than Kobe in abstract, but it seems off that he seems more valuable then Kobe when adjusted for position than Bryant. It could be a case, like the Iverson example in his book, where a player is overvalued because of our emphasis on scoring, but I watch basketball and it is hard for me not to see how good Kobe seems, and he is the top rated shooting guard in the league.
From this it seems to follow that a Camby-Gasol led team have been comparably successfully to the Kobe-Gasol team one. Intuitively that seems crazy, but I guess so did a Sixer playoff run last year. I’m interested in his take.
But what really interests may not be any of this – beyond what it means in the narrow sense for the Sixers – is where the league is headed, and what Berri can tell us about the future of basketball.
If every team in the league adopted his insight as gospel and built their teams according to predicted win scores, what would the league look like? Lets say for a second that Marcus Camby is better than Kobe, what does that say about slashing two-guards without Kobe’s talent? What does it mean for a guy who may not have traditional SG skills (jump shot, great athleticism) but won’t get killed on defense and can really rebound for a guard?
It seems like there has been a recent push for a more European-style, where players have skill sets that go beyond traditional positions (power forwards who can shoot, big guards who can rebound). If Berri’s system ruled the world, would that continue or would teams go towards very defined roles?
We’ll try to get his take, but let us know what you think – and if you’re buying Camby’s value. Remember, this was the one guy who got the Sixers might last year.
Hi, I’m a regular reader of Dave Berri’s journal. I’m not pretending to understand his calculations, but I do read all his columns. What I take from it, in short, is that pure scoring (as opposed to scoring efficiency) is overvalued by anyone interested in wins, although it is rightly valued by players interested only in getting paid. A team following Berri’s principles would look very much like the Celtics last year, when everyone agreed that the scoring would come naturally and they should focus on just about everything but scoring.
In some ways Michael Jordan was bad for the NBA, I judge, because he was a scorer and a champion. He really may have been the best ever — although Berri has made a case for Magic Johnson. But Jordan needed Pippen and Grant and Rodman, who are players Berri would rank more highly than most experts. And Jordan is, I believe, still the only player to lead the NBA in scoring and win a championship in the same season.
Since Jordan retired, everyone has been looking for the next Jordan. GMs should be looking for the next Jordan if they can find him, of course, but they have a better chance finding the next Pippen or Grant or Rodman, because those players will be undervalued. And coaches should do everything they can to reward more than just scoring.
Kevin Durant was voted rookie of the year last year. Under Berri’s system, he wouldn’t even be in the running, although his college stats were phenomenal. For whatever reason, Durant has focused on scoring while ignoring other parts of the game, including scoring efficiency. And he seems to be getting rewarded for his efforts, in every way except wins.
The problem with Camby, by the way, is that he has often been injured, and now is getting old. If he stays healthy playing starter minutes for an entire season he can be great, but that’s a big if.
This whole deal of center vs point guard is similar to the value of quarterbacks vs tight ends in fantasy football drafts.. It’s all relative to other players at the position, hence Value-Based-Drafting (VBD) was invented. So, a center should be compared to the average of all centers, point guard to the average of all point guards etc
Geoffrey, I hear you with VBD, where it would seem silly to compare a quarterback to a tight end, but I would argue that in the grand scheme of football most would agree that a quarterback is inherently more valuable than a tight end.
While there would presumably be times when an excellent tight end would be preferable to a mediocre quarterback, it seems as though a football team’s time and resources would be best spent finding, training, and developing quarterbacks because, more often than not, they control more of the game than a tight end does.
I’d prefer Antonio Gates to Tavaris Jackson, but I might rather have Jake Delhomme than Gates, and I would certainly rather have a top QB like Brady or Manning.
Likewise, it seems to me that Professor Berri’s stats suggest that big men, because they produce more wins, are like quarterbacks while shooting guards are more like auxiliary role players. No one would want to pick someone like Samuel Dalembert over Kobe, but perhaps Dwight Howard does have more value.
The problem, also, is that there are so few great or even good big men. That makes them all the more valuable when you find them. It’s not necessarily that they produce more wins than the guards, it’s that they produce way more wins than the stiffs playing center on the other teams. Guards are easier to replace with at least an adequate substitute.
Tim – the one response that I would have to your point would be that, according to Berri, big men do in fact produce more wins.
I know I’m repeating an example, according to the WoW scores last year Joel Przybilla produced 9 wins, which was more than Tony Parker, Kevin Martin, Ray Allen, Joe Johnson, etc.
This wasn’t because Przybilla was secretly an elite center – he wasn’t, Berri ranked him 15th – it seemed to be because big men tend to produce more wins.