This morning – while mowing my lawn – I received a call from David Biderman of the Wall Street Journal. David wished to know if LeBron James, Dwyane Wade, and Chris Bosh would actually be the most productive trio employed by a single team. Obviously the question presumes that LeBron is going to Miami, which may or may not be true. Regardless of LeBron’s final destination, though, Biderman’s question is an interesting thought experiment.
So I stopped mowing my lawn and went to my computer to see if I could provide an answer. Stumbling on Wins already provides the productivity numbers of each team that won a title from 1978 to 2008. All I needed was to update the analysis seen in the book. Now if I was Arturo Galletti or Andres Alvarez, this task would have taken about 15 seconds (Arturo and Andres are amazing!!!). However, it took me about an hour to provide David with an answer. And you can see this answer in the following short story:
What If LeBron Went to Miami, Too?
Speaking of LeBron…on Thursday I am scheduled for the Hays Advantage (with Kathleen Hays) on Bloomberg Radio. Kathleen and I are scheduled to talk at 1:30pm EST (and you can listen live on-line) about LeBron and what he means to the team that ultimately gets to employ him next season.
And finally… Jonathan Weiler – author of Authoritarianism and Polarization in American Politics – has written a column for Huffington Post examining the relative treatment of Amare Stoudemire and David Lee. While we wait for LeBron, this post might be worth discussing.
– DJ
Wages of Wins Journal 
nerdnumbers
July 7, 2010
DJ,
Thanks for the compliment. As a nice segway, I have a post on the top three players based on last year’s play for some of the free agent movers and shakers: http://nerdnumbers.wordpress.com/2010/07/08/24/
Also, Robbie’s post on the Lebrononomics on various teams is really cool. Awesome post by Biderman! It is amazing to see how good Rodman really was. It is also fair to say he was underrated given the number of times he was traded.
todd2
July 8, 2010
Nice work prof. It looked as if Ainge had struck a template for success with the Allen/Garnett/Pierce project, but in this light he may have just caught lightning in a bottle. There is probably some merit for assembling a big three (i.e., the Spurs) but it looks like some balance needs to be made with the rest of a squad. The Cavs were an extreme example of an uneven talent distribution this past season—what would a bell curve look like for the average championship squad since 1977, in terms of WP48? That might give us a better indicator of the makeup of a contender.
Chicago Tim
July 8, 2010
nerdnumbers
Rodman wasn’t traded that much. He demanded a trade from Detroit, and he clashed with management and struggled with an injury in San Antonio.
arturogalletti
July 8, 2010
Prof,
Thanks for the complement and congrats on another WSJ mention. I liked the HuffPost piece but I thought they obviated one of the Central points in undervaluing David Lee: Selection Bias. Under the white men can’t jump corollary, NBA scouts,pundits & gm’s are simply not conditioned to accept 6’9″ White Guys who can play. His stats in the most common used numbers 20 pts a game & 11 rebounds would seem to scream max player (and are better than boozer’s) . I do think this is an important factor in understanding why he was excluded from the top FA list in most peoples minds.
reservoirgod
July 8, 2010
If Stephen A. Smith is right and Lebron goes to MIA without any of the other top wins producers from last season going to ORL or BOS then I think KG will be the only thing standing between the Heat and the NBA Finals (of course, this doesn’t take injuries into account). I think Bosh will have to outplay KG for the Heat to beat the Celtics.
Alvy
July 8, 2010
How were the projections done? It just seem a little off for me. Like, will Wade see a fairly large decrease in wins, while LeBron, although a year older, will increase his wins? w/e
Anyway, I think LeBron James goes to the Chicago Bulls.
marparker
July 8, 2010
Byron Scott
Danny Ainge
Joe Dumars
Michael Jordan
Mario Elie
Kobe Bryant
Manu Ginobili
Richard Hamilton
Dwade
Ray Allen
These are the shooting guards who played the most minutes for championship teams since 1985. Thats near 30 years and only 10 guys. You go to other positions and the results are all over the place. At 2 guard the list only includes solid players who had above average seasons. I think there has go to be some cut off where a team can’t win a championship without a certain caliber shooting guard. It’s clear that Lebron doesn’t have one.
By my calculations all of those shooting guards either had an offensive rating(bballref) 7 points except Rip Hamilton.
Tangent Alert: I’ve been studying Bref’s ws/48 metric and I believe that to more accurately depict how players are effected by playing on good teams. For instance espn auto wp48 has Hamilton as below average that season. However, ws/48 has Hamilton as .15 for that season. I’m inclined to believe that a player like Hamilton is fine as long as he’s playing with teammates like Ben Wallace and Chauncey Billups who cover for his negatives. But I’m not inclined to believe that Hamilton would be useful in many other places circa 2005.
This all leads me back to Lebron James. He doesn’t play with anyone I could argue would belong on that list. Meanwhile 4 other teams have guys on that list though Ray Allen is old and isn’t the same player he once was some really old shooting guards have been valuable to championship teams.
I believe the Bosh move puts Miami in the driver seat. A guy like Dwade gets any help, and the rest of the league is in trouble. San Antonio is always lurking and Dajuan Blair should play a larger role this season. Boston is old but we’ve seen what that crew of old guys can do in June. The Lakers are the 2 time defending champ. However, if Lebron gets to play along side a worthy wing we could get a very interesting post season. However, none of his free agent suitors currently offer him that luxury.
marparker
July 8, 2010
edit: All those 2guards have offensive ratings 7 points above league average except Rip Hamilton.
marparker
July 8, 2010
sorry for the troll style posts….
Of course I turn on the Espn and it’s telling me that Lebron James is going to Miami. How fitting would that be after what the Lakers did to get back to championship level. Miami sees your Memphis faux paus and raises you 3 max contracts.
simon
July 8, 2010
marparker/
That works for other positions as well. If you count the players who got most minutes at C and PF, you also get around 10 players. After a quick&dirty count, I got around 12 players for C and 7-8 players for PF although (I could be wrong and there are some grey areas.)
Thus trying to argue that a certain position is worth more than suggested because there’s a limited selection of players who played for the championship teams apply for the positions.
It’s pretty logical too when you think about it. There have been only 10-13 different “squads” who won the championship since 1985, and after winning championships they are not likely to make big changes to their rosters.
tywill33
July 8, 2010
Tremendous article on Stoudamire and Lee, with one exception. The author perpetuates the myth that Win Score “weights” possession statistics more heavily than other metrics.
Dave, maybe I’m wrong. But as understand the term “weighting” it refers to augmenting something to make to magnify its importance. In other words the statement “The Professor weights performance on final exams more heavily when determining final grades” means to me he ADDS to its value to artificially magnify its importance.
But that’s not what Win Score does with possession statistics. The statistics themselves are considered without any augmentation at all. Therefore, to me, its incorrect to say they are “weighted”.
Why am I making such a big deal out of a small point? Because it causes misconceptions in the minds of casual fans that undermine fair consideration. It makes them dismiss Win Score out of hand as a metric that artificially enhances rebounding, as though you created it the metric while on commission from the “National Rebounding Awareness” lobby or something.
Zack
July 8, 2010
This comment is very out of place, but I didn’t really know where else to put it. Basketball-reference has a statistic called win shares, which, similarly to wins produced, calculates the number of wins a player contributed to his team. I looked at the manner in which the statistic was calculated, and it seems very different from wins produced. Still, despite their very different origins, both win shares and wins produced correlate very well to actual wins. I love the wins produced statistic, but the fact that a completely different stat can also correctly predict wins is very troubling to me. Can anyone restore my faith in wins produced?
dberri
July 8, 2010
Win Shares and Wins Produced both explain current wins equally well because both employ the same basic relationship (i.e. offensive and defensive efficiency explain wins). The difference between the two metrics is how the components of the efficiency metrics are allocated across players. Dean Oliver did not use a statistical model to do the allocation. He essentially used the “difficulty theorem” — or the idea that what looks more difficult should be weighted higher– to allocate the stats. And then adjustments are made to make it all work out. We talk about the probles with the “difficulty theorem” in The Wages of Wins.
Beyond the problem I have with the “difficulty theorem”, the Wages of Wins measures are more consistent across time. So if we consider a) explanatory power of current wins and b) consistency across time; then Wins Produced is the preferred measure.
That being said, it the universe of models was Adjusted Plus-Minus, PERs, and Win Shares (which it was before the Wages of Wins was published), then Win Shares would be the preferred model. I have an article coming out in academic collection that makes this point.
Aon
July 8, 2010
This is hilarious and so true. Completely puts berri in his place.
http://mgoblog.com/content/sports-economists-always-wrong-about-everything?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+mgoblog+%28mgoblog%29
nerdnumbers
July 8, 2010
Aon,
Article kind of reminded me of the Scary Movie franchise. The first several movies assumed because they could make a reference to a movie that they had made a joke. Similar here, the fact that they can label Berri’s theories does not make it hard hitting analysis.
Also congrats on getting into Princeton DJ, I think it’s a step down from CSU but it certainly shows the hard hitting research the author did.
arturogalletti
July 8, 2010
Zack,
To continue in the prof. line, Both models produce similar results when you line up actual wins vs predicted wins (They both actually have the same correlation:97.4% for 2009-2010 win totals). The upside of Wins Produced is that by using a statistical model to allocate the division of wins among player one can more accurately predict performance over time.
I’ll point you to ta few articles:
http://www.basketball-reference.com/blog/?p=6859
B-ball reference article where they admit that everybody got it wrong with Garnett,Pierce and Allen
https://dberri.wordpress.com/2007/11/30/three-is-the-magic-number/#more-651
https://dberri.wordpress.com/2007/11/20/will-the-boston-celtics-be-the-most-improved-team-in-nba-history/
https://dberri.wordpress.com/2007/11/30/three-is-the-magic-number/#more-651
Some of the articles on the last big trade that sold me on the predictive power on WP.
kevin
July 8, 2010
“It is amazing to see how good Rodman really was. It is also fair to say he was underrated given the number of times he was traded.”
Rodman was like a mini-Russell. He didn’t need the ball to dominate.
Rodman was fascinating. His public personna was at complete variance with his on-court game. Off the court, he tried to be flamboyant, narcissistic and out-of-control. But on the court, he was the epitome of control. He was unselfish to a fault, never tried to force a play and always seemed to be in the right place at the right time.
robbieomalley
July 8, 2010
Aon,
Joe the troll puts Berri in his place to that degree about once a week. Certainly is hard hitting analysis.
And were you going for “Anon” but failed? If so, it essentially sums up the rest of your comment.
simon
July 8, 2010
nerdnumbers/
People still laughed at Scary Movie because they thought “Hey, I recognize the cultural references, so it must be funny!” I see the same pattern in the comments section of that article posted by Aon. The quality of criticism seems to have gone down since 2006-07. I blame the FreeDarko article.
arturogalletti
July 8, 2010
If Lebron signs with Miami, We get a real life live test of the difference in the predictive models.
dberri
July 8, 2010
Apparently my place is at Princeton. So on one level I appreciate being put in my place there.
Then again…no offense to people from New Jersey, but I really prefer living in the Rocky Mountains. So I think I will just keep my place here.
marparker
July 8, 2010
Simon,
The only problem is that the other positions
include Derek Fisher, Ron Harper, old Gary Payton etc. The shooting guard position is full of actual career producers.
marparker
July 8, 2010
Some great 60 win team failures were missing shooting guards…Seattle in 94 played Kendall Gill more than Nate McMillan…Dallas in 06 only had Jason Terry…etc etc
szr
July 8, 2010
Aon, I’m not sure how this analysis touches WoW conclusions. I mean, if Berri points out an interesting trend in the data about QBs taken in the draft, you can’t turn around and say its obviously wrong unless you can show where Berri is wrong. Was the database wrong? Was there a missed calculation?
His complaint about the NHL goalies argument might have some more teeth, only it amounts to expanding the data set so much that the average becomes meaningless, since there are plenty of crappy goalies who barely play a game before its obvious they’re not good enough to be in the NHL. Maybe Berri should say “The very top NHL goalies is only marginally better than the average NHL goalie. But both the top and average are very very good.”
In fact, in Stumbling, Berri DOES make that point about coaches. He explicitly states that his argument isn’t that the coach is irrelevant to an NBA team, only that the very very best (think Phil Jacksons) NBA coaches can only eek out a few more wins than the average NBA coach. I don’t think anyone would contest that claim. I mean, even an average NBA coach is going to be one of the greatest basketball coaches alive today.
In short, comparing someone to their peer group, and finding they’re only a little better, isn’t at all the same thing as saying they’re only a little relevant. This is especially true when your peer group are the greatest players of your generation at your position in your sport.
simon
July 8, 2010
marparker//
I still think that’s misleading. For instance Ron Harper was very productive for the Bulls. I guess what you’re trying to say is scoring shooting guards are more valuable than numbers say compared to point guards, and who know, maybe it’s true, but reaching that conclusion based on positional similarity in team construction to past championship without taking other factors into account seems a little dubious.
simon
July 8, 2010
szr//
Great point. I sometimes wonder if they read the articles. Both studies concentrate on relative value in performance of coaches and goalies in comparison to their peers, not the coaching performance of inanimate rod, chairs, high school coaches, Whoopi Goldberg, etc.
marparker
July 8, 2010
Simon,
I’m saying I took a hard look at all the positions and shooting guard seemed to have really important players playing there. All teams have to “choose” a position where that player doesn’t have to be productive. For the Lakers and San Antonio it was Artest and Bowen. For the Bulls it was Wennington and Longley. I can’t find the team that won with let the 2guard be our weak position theory. But my general knowledge only goes back about 25 years.
dberri
July 8, 2010
simon,
I don’t think you should suggest a critic of my work didn’t actually read the work. I am almost positive that could never, ever happen :)
arturogalletti
July 8, 2010
Getting back on topic. I’m hearing buzz about Miller going to Miami as well.If Miami does that and plays Lebron at PF and Bosh at C, that would make for a scary,scary team.(Click my name for a pretty table on this):-)
robbieomalley
July 8, 2010
Arturo,
I heard about Miller to Miami as well. I don’t know where they’re getting the money though. I’ve heard they don’t have enough to offer the big three all the max and now they have enough to offer Miller 30 million (the reported offer)? That would be a good pick up though.
arturogalletti
July 8, 2010
Rob,
The offer is the MLE and he supposedly he has a deadline to accept today.
robbieomalley
July 8, 2010
I thought you weren’t allowed to offer the MLE unless you were already over the cap at the beginning of free agency. I guess that’s wrong though.
marparker
July 8, 2010
Lets say Miami gets Lebron and Miller and fills the rest of its roster with middlers and all these guys put up average numbers. What would that mean?
Secondly, I am openly routing for Lebron to go to Miami. All I’m hearing from sports talk is that a team built like this can’t win. I would love to see the experiment play itself out.
arturogalletti
July 8, 2010
Marparker,
It’s the Pareto principle vs. conventional wisdom in an old-fashioned texas deathmatch. I know who I have.
Scondren
July 8, 2010
Robbie, I’m almost certain you are correct about the MLE only being for teams over the cap.
simon
July 8, 2010
Holy crap. If Miami does manage to fill its rosters with average players plus the big 3, that’ll be an amazing team to watch. Unlike others I want to see that happen.
bags fly free
July 8, 2010
arturo, how do you predict player’s performance over time with WoW? Do you assume they get older and will play less minutes?
marparker
July 8, 2010
Also, has anyone else been paying attention to golden state? They may have a potential lineup of Curry, Ellis, Lee, some small forward and Biedrens. I don’t think they will play any defense but it would be fun to watch.
arturogalletti
July 8, 2010
bags,
The prof. posted something on this earlier (see the last post) but past performance and age can be used to predict future Adjp48 fairly accurately (Minutes and thus wins are a little harder). However it’s fair to say that there is an high probability that Joe Johnson and Dwayne Wade have already already had their peak season (I’d say more but, spoilers).
Mike
July 8, 2010
By my calculations, if the GSW players keep their last year’s WP48 and add Lee, the starting lineup of Curry-Ellis-ReggieWilliams-Lee-Biedrins would produce about 44 wins if using a reasonable minute distribution (about 36 min per game per starter).
If Curry improves (likely) and Biedrins reverts back to his pre-injury form from the past 2 years (0.3 WP48), then the lineup would produce about 50 wins.
I hope LeBron doesn’t screw this up tonight by signing with NY. UNSTOPPABLE BABY!
robbieomalley
July 8, 2010
LeBron James, Wade, Bosh – Three Hours Away
For a preview:
marparker
July 8, 2010
hahahaha
robbieomalley
July 8, 2010
This just in,
Multiple sources say if LeBron sees his shadow, it means six more years of Cleveland.
robbieomalley
July 8, 2010
This just in,
“The Decision” predicted to have more viewers than the last episode of M*A*S*H.
robbieomalley
July 8, 2010
“Cleveland, If you like it then you should have put a ring on it.” – LeBron James
robbieomalley
July 8, 2010
http://www.ustream.tv/channel/cleveland-com-live-broadcast
Live feed from a bar in Lakewood, Ohio. This stream could be priceless.
robbieomalley
July 8, 2010
Who is everyone picking before this thing gets started?
I’m going with the Cavaliers.
dberri
July 8, 2010
If it is Miami… am I the only one who is reminded of th 1996-97 Houston Rockets? Hakeem, Drexler, Sir Charles, and a collection of players making relatively low wages.
Alvy
July 8, 2010
ha! good point, d. berri. i hope lebron goes to chicago (because they would be able to compete right away) but i’m okay with heat because of the experiment (which apparently we’ve seen something like this before)
reservoirgod
July 8, 2010
Except Barkley & Drexler were at the end of their careers when they got to Houston… And they didn’t have 2 of the best players at their respective positions.
robbieomalley
July 8, 2010
I was expecting a Shyamalan twist that didn’t come.
Alvy
July 8, 2010
how many of you guys are heat fans now? lol :P