Who is the Smartest Rocket? Introducing OALABII and Memorializing Shane Battier

The other day, Xiane suggested that the Rockets largely look for "high-IQ" players. And then I got to thinking about basketball intelligence when I remembered that Jeremy Lin was on the Knicks' roster. I don't think Lin has an exceptional "Basketball IQ" or anything, but he's obviously a pretty clever guy, and that probably translates to some level of intelligence on the court, I guess. But it really made me think about who on the Rockets has the greatest basketball sense on the court.

Of course, I could just go with my gut, right? I could say it's Lowry or Parsons or Scola or someone because they just seem like the type. But, instead, I'm going to take a page from Bill James and try to come up with a system to describe who, exactly, is "smart." It might not be quite scientific, but it's better than nothing.

I think the prototypical "smart" basketball player, at least for the Rockets, is Shane Battier. We all remember (I'm sure) Michael Lewis's article on the "No Stats All Star," in which Battier's game as a guy who does all the little things necessary on the court was lovingly detailed. But Dave Berri (the guy behind Wages of Wins) took issue with one of the critical points of Lewis's piece: Battier's value was, contrary to what Lewis (and the Rockets) contended, expressed very well through the box-score. Why? Because, while most fans look to the points column near-exclusively when evaluating players, the box score has a lot more than that in it: it records field goals and attempts, steals, turnovers, rebounds, etc. And, looking at those statistics, we can see Battier's value. He shoots efficiently (or at least he did up until this year), he doesn't turn the ball over, he gets steals, he doesn't commit many fouls.

There's a distinct value in not doing something bad -- not missing shots, not turning the ball over, not committing fouls. We tend to look at those players who do a lot of something good (scoring, for instance), but our attention doesn't always turn to those who simply don't do things wrong. Battier was one of those players who rarely did anything wrong on the court.

So much of Battier's value lay in his ability to simply not give the ball to the other team. To me, that's what "smart" basketball players accomplish, and fortunately for us it's something pretty easily quantifiable through the box score.

So I'm going to model OALABII (the Only_A_Lad Adjusted Basketball Intelligence Index, pronounced "wallaby") on Battier's contributions to the game, or rather his lack of negative contributions. We'll thus be looking at the five statistics that really encompassed how not-bad Battier was (and how good he was at not giving the other team another possession): minutes played, field goals, field goals attempted, turnovers, and personal fouls (later we'll incorporate steals, but at first it will be just these).

Minutes played is important because it will allow us (in an imperfect way) to adjust for possessions and time on the court. Field Goals Attempted minus Field Goals will give us how many shots a player has missed in his time on the court. Turnovers and fouls are obvious. Missed shots, turnovers, and fouls are all effectively possessions gifted to the other team: they represent a shot that could have gone to an open teammate, a bad pass, and bad defense.

Of these, I suppose fouls are the ones most likely to be disputed as necessarily "bad." Suffice it to say that while not all fouls are terrible, they represent a failure of the defense to adequately contain a player (or they represent a turnover on offense), and while this is not always the fault of the man who picks up the foul, the guy who fouls could have performed better defense and not fouled, usually. Fouls, at the very least, aren't good. Good defenders rarely foul.

OALABII is really simple: we take missed shots per minute, fouls per minute, and turnovers per minute and add them together. A score of zero is best (a player who never misses a shot, never turns the ball over, and never fouls -- a guy who never gives the other team the ball back), and so higher scores are worse. Essentially, it tells you how many "mistakes" a player tends to make (on average) in a minute.

I've included Battier's Houston career for comparison. Battier's score across his Rockets career is 0.21, better than anyone on the team getting significant minutes. You might expect the big men -- Dalembert, Patterson, Hill, Scola -- to do the worst in this formula, if only because personal fouls are included, and big men seem a lot more likely to pick up a lot of fouls. However, being big guys might hurt them when it comes to fouls, but their more infrequent misses makes up for this. Indeed, Thabeet (in admittedly very limited minutes), Patterson, and Dalembert are the top three Rockets in OALABII, and these results don't appear to be based on simply having low scores in any one area (though Thabeet's lack of any missed shots really helps) -- all three just aren't really doing much to give the opposing team any help.

Instead, it's the guys who have been the focus of the offense -- guys like Martin, Scola, and Lowry -- who look the worst. This is partly because all take plenty of shots and miss many (this is particularly true of Martin this year), but it's also because they're turning the ball over a lot (not as true of Martin, but definitely true of Scola and Lowry).

Looking at these numbers, it's kind of amazing how rarely Battier did anything bad. He did roughly one bad thing every five minutes on the floor. That's crazy, when you think about it. He missed a shot, turned over the ball, or fouled someone only maybe six or seven times in a game. In contrast, Martin misses, turns it over, or fouls once every two minutes or so. And it's not like Martin misses, fouls, or turns it over that often.

It occurs to me, however, that this first version of OALABII weights missed shots too highly. Not all missed shots are bad ones, but (more importantly given the framework we're working with here) not all missed shots go to the other team -- some become offensive rebounds. And, fortunately for us, we know how many become offensive rebounds on average. The league-wide offensive rebounding rate this year is about 22.8% (this is actually pretty consistent from year to year. It was 22.8% last year, a little less the year before, a little more the year before that). So we can adjust for that, noting that 22.8% of these missed shots aren't total misses:

This rearranges some stuff, of course. Budinger moves up in value now that some of his misses go away, taking third-place, it doesn't really help out Lowry, Martin, and Scola. And I don't really think of these guys as dumb or anything on the court, either, so maybe we should look a little further for some adjustments.

Battier's value didn't just come from not turning the ball over. Net turnovers matter, and I think turnovers probably correlate with "Basketball IQ." Knowing how to get to the ball, predicting passes, etc. are all things that smart players are probably good at, right? So let's change "Turnovers/Minute" to "Net Turnovers/Minute."

This final version of OALABII yields perhaps the most predictable answer: Parsons is one of the most intelligent players on the team (he's one of only two players to have a negative turnover differential --more steals than turnovers), though not really much smarter than Patterson, Dalembert, or Budinger. The biggest shift is for Lowry, who moves from 0.32 OALABII to 0.27. That seems appropriate -- it was weird to see him at the bottom of the list.

Martin still lies at the bottom of the Rockets getting significant minutes, as does Scola. This is primarily because Scola fouls quite a bit (poor defense) and Martin shoots a lot. OALABII doesn't totally examine offensive efficiency (only part of it). But it is also a reflection of Martin's struggles this season -- his career OALABII is 0.25, which would put him just between Dalembert/Williams and Lowry this year.

Problems aside, what does this little experiment tell us?

First, Battier's days in Houston were astoundingly mistake-free. Battier is a monument to good decision-making on the court. His near-zero (it's like .0013 per minute) net turnover rate, his miniscule number of missed shots, and his very rare fouls made him somebody who almost never gave the ball away to the other team. Put simply, he was an amazingly smart player, someone whose ability to minimize his mistakes made him a spectacular asset to the team.

Second, there are a few Rockets who play that same way without anyone really ever calling them "smart." Samuel Dalembert is a smart player this year. He doesn't miss very often, he doesn't foul, and he doesn't turn the ball over. In other words, his hands are a relatively safe place for the ball. The same is true of Parsons (though he has already had the "smart" label applied to him) and Patterson, and even in his limited time on the court this year, Hasheem Thabeet has managed to not do much that is counter-productive.

Dalembert's intelligence on the court (would "judiciousness" be the better word?) is a lot of why he is so good this season, and while Worrell, Drexler, and Bullard tend to label him "a hard worker," I think he probably warrants a comparison to the smartest of the smart, Battier. That seems like an odd comparison to make, but in terms of minimizing failure, Dalembert is a lot like Battier (though, you know, not nearly as good at it).

Last, Xiane's statement that the Rockets draft players who have a high IQ seems to be right (the Rockets' draftees are at the top of the list, after all), though it's impossible to be sure without looking at OALABII across the league. Parsons, Patterson, and Budinger all are ranked near the top of the team, tightly clumped together in the index. Morris is ranked last, but that is in very limited minutes (only 17 so far!), and his college numbers (unadjusted for rebound rate) translate to 0.29 OALABII, which wouldn't be near the top, but we're not dealing with a great deal of spread between the numbers here, anyways.

Anyways, I thought this was kind of neat, though it shouldn't be taken very seriously. As I said, the only way to have any sort of confidence in this would be to look at a much larger sample of the NBA population, which would be time consuming. As it stands, there is a standard deviation of .08 in the Rockets' OALABII, so it's not like there's a vast difference between these guys' values -- it's very clumped together. Still, I guess the smartest guy on the team is Parsons, and so the data seems to fit. Patterson and Dalembert aren't very far behind. I've said Parsons is "Horry-like" in his do-everything-ness, but if he works on his jump-shot (and thus starts missing less), he might just be the next Shane Battier, and that would be pretty awesome.