Ed Carson of Investors. com reported the following trend a few weeks ago (HT Freakonomics):
The best kept secret of the past 20 years has been this: When the Los Angeles Lakers won the NBA championship, the market would almost always fall that year. When the Lakers lost, the market would usually rise. The Laker Indicator only steered investors wrong in three years during the entire span, and not once from 1995-2007.
An investor who put down $1,000 into the Nasdaq at the start of 1987 and stayed fully invested through 2007 would have ended up with $7,604. But an investor who bought the Nasdaq in years the Lakers lost and stayed in cash when the Lakers won would have finished with $21,189. This strategy would have kept you in the market during the 1990s bull market, avoided the 2000-2002 bear and then got back in as the market uptrend resumed.
Such numbers speak for themselves. In fact, I can’t imagine anyone looking at Carson’s numbers and coming to another conclusion.
Okay, if you read a little further in the column you see that Carson disagrees with Carson’s numbers. Here is Carson again: Correlation does not always mean causation. Psychological ‘secondary’ gauges may appear to work for a time, then suddenly stop. And it’s easy to look for excuses, a la the Lakers in 2008, for why your special indicator really still works.
Focusing on Jamal Crawford
The Carson story about the Lakers came to mind as I watch a bit of basketball on Christmas day. While watching I heard (and I didn’t write this down so this may not be exact) Hubie Brown briefly discuss the contenders in the East. After noting Boston, Cleveland, and Orlando, Brown noted that people shouldn’t ignore the Atlanta Hawks.
So far I agreed with Brown. The Hawks currently rank second in the NBA in efficiency differential (offensive efficiency minus defensive efficiency). So any discussion of NBA contenders has to include Atlanta.
But after noting that Atlanta is a contender, Brown decided to offer an explanation for Atlanta’s surge. And that explanation focused on Jamal Crawford. Essentially Brown argued that the off-season addition of Crawford is the reason why Atlanta has improved.
I should note that I am probably being unfair to Brown. Yes, I did hear him say this. But in the course of a game – where the talking heads have a job to do (i.e. keep talking) – one is bound to find something said that doesn’t make complete sense or that the person talking wouldn’t take back (or at least explain better). That being said, though, I want to proceed as if Brown really meant what he said (i.e. Jamal Crawford is the reason why the Hawks have gotten better). And I want to proceed in this fashion because Brown’s focus on Crawford illustrates a larger point about correlation and causation.
Explaining Atlanta’s Surge
Let me start by offering a defense for Brown’s comment. It’s easy to see why someone might focus on Crawford. Of the eight players who have logged at least 400 minutes for the Hawks this season, Crawford is the only one who was not with Atlanta last year. Last season the Hawks won 47 games without Crawford. This year the team’s efficiency differential suggests Atlanta will win more than 60 games. Such evidence seems fairly self explanatory. Atlanta only has one new player getting any minutes, so that one new player must be the reason why the team is much better.
Once again, though, correlation doesn’t prove causation. Just because Atlanta has done better since it acquired Crawford, it doesn’t mean Crawford is the reason why the team has improved.
Back in November I offered an explanation for why the Hawks have gotten better. That explanation centered on the play of Josh Smith. Now that Atlanta has played 29 games it seems like a good idea to re-check the numbers.
We can find Josh Smith’s box score numbers at Basketball-Reference.com. A check of these numbers reveals – just as we saw in November – that Smith has improved with respect to shooting efficiency, rebounds, assists, steals, and turnovers. His per game scoring is down, but with respect to the statistics that drive wins (both theoretically and empirically), Smith has improved tremendously.
We can see this improvement when we turn to Wins Produced.
Table One: The Atlanta Hawks after 29 games in 2009-10
Table One reports the Wins Produced for each player Atlanta has employed this year. And it reports what we could have expected had each player maintained his performance from last season. The numbers from last season indicate the Hawks should be on pace to win 46 games in 2009-10. Again, that’s about what this team did in 2008-09. So given the team’s roster moves – and what the players on this roster did last year — Atlanta shouldn’t be any better.
But Atlanta clearly is better. When we look at how performance has changed, we see that the primary reason for this improvement is the play of Josh Smith. Had Smith maintained his 2008-09 productivity level this season, he would be on pace to produce 6.2 wins. Instead, his improved box score numbers translate into more than ten additional wins. In sum, both the empirical evidence found in the box score numbers – and how these numbers theoretically and empirically are linked to wins – indicates it’s not the addition of Crawford that has transformed the Hawks (Crawford – as I noted last April – has generally been a below average shooting guard in his career and he is once again below par in 2009-10). The key is really the improved play of Josh Smith (Al Horford and Joe Johnson have also helped some).
Correlation Stories
Let’s say, though, you really wanted to stick with the Crawford story. One could argue that somehow Crawford’s presence has led Smith to hit more shots, grab more rebounds and steals, and commit fewer turnovers. Such a story is tempting, especially if you begin your analysis with a correlation.
And this is a great example for why we tend not to begin our analysis with a data-mining search for correlations. Once you find a correlation it’s too easy to start inventing theories. A better approach – and the approach we teach our students – is that good empirical analysis begins with some sort of theoretical model, and then moves on to the data. It’s only through theory that we can actually argue any causation at all. Or in other words, without a theory all we have is a correlation.
The Lakers-stock market story and the Crawford-Hawks stories are good examples of correlations in search of a theory. Other examples in basketball can be seen whenever you see people argue that when player X is added to team Y, team Y does better (or worse). Again, one can show that such correlations exist. But without a theoretical structure, it’s often hard to believe that a causal relationship has actually been uncovered. And without a causal relationship, you really don’t have much of a story.
– 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.
Wages of Wins Journal 
brgulker
December 27, 2009
I’m waiting for khandor to come brag about how he was right ;)
Joe
December 27, 2009
I know this doesn’t have anything to do with your post, but in the wages of wins you noted that soccer is competitively balanced compared to basketball and baseball. I was just wondering how this could be if it seems as the the top 4 teams are always the same in the 3 major leagues. eg: MANU Chelsea arsenal inter ac milan real madrid barcelona etc… Even the world cup final and semis seem to feature the same teams like brazil france spain germany argentina italy etc..
stephanie
December 28, 2009
Maybe I’m wrong, but it seems like you’re working the dumb media angle here. I think in this case they have agreed with your analysis quite a bit. Have you missed all the talk about how Josh Smith has “matured,” how he’s “focusing on his strengths” and so on? The fact that he isn’t chucking threes and playing closer to the basket has been brought up repeatedly in games I’ve watched, half time shows, etc. Heck, even Jalen Rose gets it.
brgulker
December 28, 2009
Heck, even Jalen Rose gets it.
That just made my morning :)
mrparker
December 28, 2009
The Hawks team defensive efficiency is 105(106.4 league average). That is nothing to write home about. They have an entire team of guys having a career year up through this point. Can they really keep up that pace?
And another, the league average offensive efficiency was 108.3 last year and 107.5 the year before that. It seems like the league is getting back to some old habits of employing deplorable basketball players.
The Bulls, Bobcats, Timberwolves, Nets, and Pacers all have offensive efficiencies below 100. In the last two years no team employed an offense that bad. Last year the worst team clocked in at 100 and the year before that 102. This year we have 5 teams below 100. That’s ridiculous.
Here’s the list guys with more than 500 minutes
that are 100 or below offensive efficiency. I know this isn’t a wow stat but it has the same inputs.
Devin Harris
Courtney Lee
Douglas-Roberts
trenton hassel
terrence williams(rookie)
raferalston
johnny Flynn(rookie)
corey brewer
dantay jones
brandon rush
stephen jackson
tyson chandler
john salmons
kirk heinrich
brad miller
There are at six players by my count who have had an average wp48 season and they are playing dreadful. Are they stealing money? How come they all play for the same teams? Help me out dberri, I feel like I have 3/4 of a beautiful league. Can we blow up the front office of those teams?
Rob O'Malley
December 28, 2009
“Heck, even Jalen Rose gets it”
Really funny, and is actually a really powerful statement. It really hammers home your point.
stephanie
December 28, 2009
I shouldn’t pick on Jalen. He’s cool. Maybe if Mark Jackson had said it…
mrparker:
I think a case can be made that the league is more interesting when there are a surplus of super teams and the rest is worse than usual. It makes the playoffs more compelling I think. Failing that, there’s always the wonderful Basketbawful.
khandor
December 28, 2009
STARTERS
PG – Bibby
OG – Johnson
SF – Williams [new add? … given the GP’s he missed last yr]
PF – Smith-Jos
C – Horford
——————
KEY SUBS
PG – Teague [new add]
OG/PG – Crawford [new add]
OG/SF – Evans
PF/C – Pachulia
C/PF – Smith-Joe [new add]
——————
HEAD COACH
Mike Woodson
Perhaps, what Hubie Brown should have said, instead, is something like this:
“The Atlanta Hawks have improved this season due to the ‘inter-action’ of the following:
i. Josh Smith’s continued maturation, as a player and a person … much of which is predicated upon the existence of ii. and iii., below;
ii. The good health, thus far, of Marvin Williams;
iii. The additions of:
A. Jamal Crawford, i.e. a good sized, combo guard who can create his own shot, score, and is a multi-dimensional player;
B. Jeff Teague, i.e. a solid back-up who has the right character and skill-set to fit in with this team;
C. Joe Smith, i.e. a solid back-up who has the right character and skill-set to fit in with this team, at his specific position;
which has siginificantly increased the Hawks’ versatility and effectiveness, in terms of Offense, Defense and Rebounding … when you consider that their Core Group of players has now been together for a number of consecutive years, while continuing to improve their individual and collective performance.
Simplistic answers … like, e.g. Team X is better this season because Player X has improved in this specific area, and that’s all … are very rarely accurate when it comes to explaining properly the reasons why a specific team is now playing better in the NBA.”
——————————————————–
brgulker,
In this case, bragging isn’t necessary, as the facts speak for themselves. :-)
dberri
December 28, 2009
Khandor,
I keep looking for “facts” in your comments but always come up empty. Opinions, yes. Informed opinions… well, not often. Fact… can’t find them.
Unfortunately, your analysis speaks of someone who has watched a great deal of television and is now adept and impersonating the talking-heads on television.
khandor
December 28, 2009
David,
It is my expert opinion that what some/many consider to be fact is not fact at all.
Fact 1: The current standings in the NBA, http://www.nba.com/standings/team_record_comparison/conferenceNew_Std_Cnf.html
Fact 2: The NBA analysts who claimed, in advance, that this year’s Atlanta Hawks would likely perform as well or better than last year’s Atlanta Hawks were, in fact, few and far between.
As I said above … in this case, at least … the facts speak for themselves.
dberri
December 28, 2009
Khandor,
Again, you really don’t have facts. Just opinions. And you seem to think (and I suspect you learned this from watching television) that if you state your opinion emphatically then somehow it will persuade someone. Perhaps in your local bar or someplace else on the Internet this is a successful strategy. But in an academic forum, you need some kind of substance behind what you are saying. You never, though, seem to have any testable hypothesis (this would be “substance” in an academic forum). Just opinion after opinion. Again, this might work for you elsewhere. But here, all of your words (and often there are too many) fall flat.
Isaac
December 29, 2009
The idea that player combinations are important is not trivial or unfounded. Good teams are balanced in the skills that they employ. Explaining a team sport with observations about individuals is not good science. There are interaction effects at play that are difficult to capture but they should not be ignored. I get that you have a developed a rigorous formula (and it is a good one) that you want to apply, but I think you are dismissing potentially useful criticisms just because they aren’t presented in an “academic” manner. Khandor’s theories are not as strong as yours, we all get that, but maybe you could explain why WoW accounts for things like interaction effects instead of claiming that he doesn’t have any hypotheses. Then again, maybe you did at some point and I just missed it.
Also, here is an interesting fact:
The best Hawks lineup (by any statistical measure I have seen) is Bibby-Crawford-Johnson-Josh.Smith-Horford, and it isn’t a close race. It is the starting lineup with Williams replaced by Crawford. This lineup scores .23 more points per possession and allows .02 less than the starters and leads the team in +/-, eFG%, and Win%.
khandor
December 29, 2009
Isaac,
re: “Khandor’s theories are not as strong as yours, we all get that, but maybe you could explain why WoW accounts for things like interaction effects instead of claiming that he doesn’t have any hypotheses.”
The fact of the matter is that “the theories I have and present are actually ‘stronger’ than David’s” … at least, in some specific instances.
Exhibit A
David’s Hypothesis about this year’s Atlanta Hawks team, provided during summer/09:
The Hawks will not be an improved team next season because they have failed to add new/more productive players to their team.
khandor’ Hypothesis about this year’s Atlanta Hawks team, provided during summer/2009:
Part 1 – The Hawks will be a slightly improved team next season because they are maintaining a stable roster of Core Players who are continuing to improve, individually and collectively, while adding “good fit” minor pieces to complement their Core Group.
Part II – The Hawks will not actually be trying to win the NBA championship during the 2009-2010 season … if they do, that would be terrific, but it’s not their primary goal this year. Rather, the Hawks will be maintaining a slow-and-steady build-up of a team they hope will eventually be able to retain the services of Joe Johnson next summer – during his UFA period – and then become a serious challenger for a NBA championship during the next phase of development for their group of ‘high end’ Core Players [e.g. Joe Johnson, Josh Smith, Al Horford and Marvin Williams], when they’re in their peak years [approximately 27-32 years of age], as NBA players.
——————————
David,
Point 1. Others are always free to judge for themselves whether the opinions [?] and theories [?] and observations [?], etc., which I provide about the game are [eventually?, subsequently?] proven to be accurate [correct?], or not.
Point 2. As I’ve mentioned in the past … when “your numbers” and “my acumen” converge on the same point, then, there’s a strong likelihood of a future sports-related event occurring in the way you’ve suggested in your article. When your article and my acumen do not converge, however, I have yet to see presented any “evidence” to suggest that anyone else’s perception of what is likely going to happen is, in fact, more accurate than mine.
Point 3. Is the primary goal you have for this blog really to create “an academic forum”, exclusively, where there is little room made for the role of vibrant opinions, theories, observations, etc.?
… because, based on what I’ve seen from your own writing in this space for the last year plus, I don’t think it truly is.
IMO, you are very good … one of the best, in fact … at what it is you happen to do.
All the best and,
Keep On Truck’n :-)
Isaac
December 29, 2009
Khandor, you’re killing me. Berri’s theory was not that the Hawks are not going to get better, in fact his theory has very little to do with the Hawks at all. His theory is that wins produced is a useful statistic that can be used to explain basketball outcomes. He took this theory and applied it to the Hawks, with less than perfect results (as you can expect with any less than perfect theory). Your theory is that you think the Hawks are getting better because sometimes players get better with age and that Jamal Crawford and the other subs are “good fits.”
See how your theory is specific to one team and time and more than a little vague while Prof Berri’s is generalized and concise? That is why his is a better theory. Not because it was more or less correct this time, but because it is testable, generalized, repeatable, and concise.
Anyway, I should have kept myself out of your argument and just said that I think interaction effects are important and that this Hawks team is a good example of a player who usually makes teams worse (Crawford) making a team better. Sometimes a team needs a player who is not afraid to take less than terrific shots, force the issue offensively, and shoulder the scoring load. This is a bit more opinion based than I would like, but there are some statistics backing it up and I think it makes way too much sense not to be explored further.
dberri
December 29, 2009
Isaac,
We have discussed interaction effects before in this forum. The argument offered is that such effects — contrary to what we see in football — are quite small. This conclusion is based on the fact (not an opinion) that NBA players — relative to what we see from baseball and football players — are quite consistent from season to season.
Now players are not perfectly consistent, as the Josh Smith story illustrates. And that is what makes the Josh Smith story so interesting. Players typically don’t make such leaps this far into their career.
One could argue — as I note in the post — that Smith’s leap is due to the team adding Crawford. But that seems like a good example of confusing correlation with causation.
Unfortunately, this happens frequently in basketball analysis. People observe a team does better or worse with a specific player and conclude that player is the cause of the team’s outcomes. But without some sort of theoretical model — and supporting empirical evidence — such conclusions can’t be supported. It doesn’t mean that it isn’t “true”. It simply means that you don’t know that it’s true (even if you really, really believe it).
dberri
December 29, 2009
Isaac,
I missed your second post (posted as I wrote my comment). You did a better job saying what I was trying to say.
As you can see, we disagree on the importance of interaction effects. But that is not uncommon. Theories and empirical evidence don’t settle all debates or provide all answers. It is just a better approach than just making up stuff as you go along.
khandor
December 29, 2009
Isaac,
re: “See how your theory is specific to one team and time and more than a little vague while Prof Berri’s is generalized and concise? That is why his is a better theory. Not because it was more or less correct this time, but because it is testable, generalized, repeatable, and concise.”
Although I understand what you’re trying to get at with this comment …
1. A theory which fails to work properly is not a “better” theory than a theory which actually works.
2. Understanding basketball properly is about being specific to one team and one time, etc.
——————————————————
David,
re: “This conclusion is based on the fact (not an opinion) that NBA players — relative to what we see from baseball and football players — are quite consistent from season to season.”
1. According to what specific measurements?
Italian Stallion
December 29, 2009
D. Berri,
As much as love most of your system, IMHO you keep allowing your numbers to blind you to obvious realities.
You state that the interaction of players doesn’t matter much and point to the consistency of players from year to year as evidence of that. But you fail to acknowledge that GMs and coaches are specifically trying to build teams of players whose talents and skillsets compliment each in ways that will make them all successful. That is their job!
So to a large extent what you may really be measuring is their success rate and not the fact that player interactions don’t matter.
There are examples of failures (some that were widely predicted beforehand). So it is not unheard of. When that happens, one of the players is usually traded.
IMO, I would have no problem putting together a team of players that are all highly rated as individuals that would not be a very good team. But I guess if I did it 10 times in a row I still wouldn’t prove it to you without the math to back it up.
I understand the standards of academia, but at some point you have to get past that and deal with reality and common sense until the math catches up.
This all reminds me of stock market, efficient market theory, and academia.
I’m sure the research and ideas behind efficient market theory taught in our best institutions are quite impressive.
The problem is that anyone that’s been on Wall St for more than a few years knows that it’s a boatload of crap! The stock market is often wildly out of whack and there are loads of people with decades long records of outperformance using similar methodologies to prove it. They don’t have the scientific and mathematical research required of academia to prove their ideas, they just have the billions of dollars. To me that’s enough proof. I’m willing to mimic them, wait for the math, and get rich in the mean time.
dberri
December 29, 2009
IS,
Perhaps you should read all of the academic articles explaining the problems with the efficient markets hypothesis. Those articles follow the basic rules of systematic inquiry. They do not rely on such statemements as…
In my opinion…, or,
You have to understand common sense….
This is the difficulty you have in explaining your ideas. Stating that your ideas simply follow common sense doesn’t persuade. I can find people who disagree with you who also claim “common sense” is on their side. Again, can you offer an argument that doesn’t rely entirely on “IMO” and “common sense” statements?
brgulker
December 29, 2009
Here’s a passing thought (in the most literal sense) that might have something to do with Smith’s increased efficiency:
Josh Smith is shooting less this year and shooting better. Perhaps hat isn’t his better judgment, and perhaps it’s not at the instruction of his coach.
Perhaps, an alternative explanation is that Josh Smith is shooting less (and taking higher percentage shots) because the coaches have instructed Crawford to take the type of shots that Smith was shooting last season (i.e., low percentage jumpshots). Or similarly, perhaps Crawford is simply just shooting a lot, which means there are fewer shots for guys like Smith.
Obviously, this doesn’t explain anything other than shooting, even if there’s a hint of truth to it.
In sum, I guess I’m wondering if Smith’s increased offensive efficiency could actually be accidental rather than intentional.
dberri
December 29, 2009
brgulker,
Josh Smith has improved across the box score. Hard to see how Crawford does this for him. And if Crawford’s bad shooting has such an amazing benefit, why can’t we see this on any other team that employed Crawford? As the article I link to in my Crawford post notes, Crawford has always played for losers.
Italian Stallion
December 29, 2009
D. Berri,
I totally acknowledge and understand your point of view.
We are not all blessed with equal mathematical skills, intuition, or powers of observation. We all learn in different ways and at different rates.
What’s the old saying?
“Opinions are like assholes, everyone has one”. :)
However, when someone expresses a series of opinions that turn out to be right way more often than other supposed experts, it perks up my interest.
I really don’t care if if he can prove his ideas or present them in a way that others can duplicate. I’m still interested. At a certain point it stops being a potential string of good luck etc…
You don’t know me from a hole in the wall. I don’t claim to be an expert. I’m just an opinionated guy with a handicapping background that watches and thinks about basketball a real lot. However, I bet that if you paid close attention to all the people on this forum you’d find a few people whose opinions consistently turn out to be very insightful even when they disagree with yours.
When you find that, I think it might be worth paying close attention and potentially tweaking your own research and methodology even if they can’t satisfy the standards of academia.
brgulker
December 29, 2009
Dr. Berri,
Yeah, I get that.
I wasn’t trying to “credit” Crawford with “making his teammates better.” What I was suggesting is that perhaps because Crawford takes so many (bad) shots, perhaps there are fewer shots for Smith to take.
But your point about Crawford’s and his teammates history pretty much dismantles my idea anyway :)
khandor
December 30, 2009
Although this specific series of statements/observations might seem to be quite reasonable to an educated person:
* Jamal Crawford has always taken a high percentage of bad shots.
* Jamal Crawford has always played for losing NBA teams.
* Jamal Crawford’s bad shots have contributed negatively to the poor performance of the losing team’s he has played for thus far.
* Jamal Crawford will continue to take a high percentage of bad shots AND continue to play for losing teams … because Jamal Crawford’s performance as a basketball player remains essentially constant over time, according to a series of specific measures.
The fact is … they do not work in that way.
If Jamal Crawford is transported into “a different situation” than what he’s been in before … i.e. with a different and improved [i.e. a better fit with his specific skill-set] group of teammates, coaching staff, system of play, GM, owner, level of maturity, etc. … then, it is a simple fallacy to think/expect/suggest to others/etc. that he [and, therefore, the team he’s playing for] is destined to perform in the same way as he has prior to this specific point in time.
An exclusively mathematical-based theory which attempts to explain precisely how the game of basketball actually works will fall short of accomplishing its specific goal … at least, until an exponentially greater number of factors are included in the measurement equation.
Basketball, understood properly, is about the individual and collective effectiveness of specific match-ups, at a given point in time.
John
January 8, 2010
Why is everyone dismissing Crawford? He is shooting 46% scoring 17 points off the bench (averaging 2 less minutes per game). From my observations as a season ticket holder, he is a perfect compliment to the sluggsh Bibby and the one-on-one-half-court Johnson.
james
January 16, 2010
John is because theyve went so far out on a limb to bash Crawford and have totally made fools of themselves with there over reliance on stats that they have to try and keep up the appearance but everyone knows they were pretty much DEAD WRONG about Crawford in Atlanta . He even tries to claim Crawfords play this year has been subpar for a SG which is a complete Joke .
The one thing they cant calculate is chemistry or front office melodrama . The Hawks are the first team that Crawfords played for in which there was not major problems between players/coaches/ management . So its not surprising hes been better all around by being in an environment focused on basketball and winning .
Don Nelson vs everyone including Warriors management , Mullin etc .
Marbury vs Larry brown and isiah
Larry brown vs knicks managment
Hubie was right and the difference between last years hawks team and this years team is Crawford.
The hawks added another guy who can close out a game but one that can also pass dribble and shoot . Which takes pressure off of Joe and changes the way teams can play the Hawks for the better which in turn makes the hawks a better team.
Josh has imprpoved simply by not taking so many long jumpers
Al has become much better on the block against the smaller centers but still struggles against the larger ones
Marvin started slow but is coming around now
Crawford has impacted everything about the Hawks from top to bottom in a positive way.