Behind the Numbers: The Thin White Line and Foul TheoryPosted by KCarpenter on December 1st, 2010
Kellen Carpenter is an RTC contributor.
Moneyball isn’t the founding document of sabermetrics; that honor probably belongs to Bill James’ Baseball Abstract. That said, the reason that most baseball fans know about the advanced, modern approach to baseball statistics is because of Michael Lewis’s nearly 300-page story about the Oakland A’s. Well, they know either because of the the book itself or the uproar and debate around it. In any case, Moneyball was and remains a cultural phenomenon, a true breakthrough into the mainstream. Because of Moneyball, baseball and the dialogue around baseball has gotten smarter.
Basketball hasn’t had a single watershed moment like Moneyball. Progress in the advanced stats movement has come in fits and starts. Unsurprisingly, the closest thing to a mainstream breakthrough for advanced statistics in basketball came from Michael Lewis. “The No-Stats All-Star” was published February 13, 2009 in the New York Times, and focused on the world of advanced stats in basketball through the microcosm of the Houston Rockets player Shane Battier and his general manager, Darryl Morey. The article introduced the larger world to lots of fun ideas like offensive and defensive efficiency and adjusted plus/minus. It had a few interesting smaller nuggets too, and today we’ll be taking a look at one of those.
According to Morey (and the research), the worst possible outcome of a defensive play is to foul. In fact, Morey mentions that they identify other teams in the NBA that make use of the modern numbers-based thinking by looking to see which teams make a consistent and radical effort to avoid fouling. It’s a simple check, but one that makes sense: of the Four Factors that contribute to defensive efficiency, opponent’s free throw rate is the easiest to control. Telling a team to not foul is an easier instruction than “rebound better!” or “reduce your opponent’s effective field goal percentage!”
In any case, Morey’s observation made me curious as to which teams in college basketball make use of that mantra and consistently avoid fouling. To make sure that I didn’t wrongly consider a fluke year, I looked at the average team opponent free throw rate from 2006-10. For those keeping score at home, free throw rate is calculated by dividing attempted free throws by attempted field goals (and multiplied by 100 to get a slightly friendlier percentage). The average free throw rate for Divison I schools over the past five years comes in at 37.1%, while the fouling-est team registered a 54.3% and the best, least-fouling-est team managed a mere 23.3%. [the complete and fully sortable list is located here as a Google Doc] Now, to name names: The top four teams who foul the least in order: Ohio State, Siena, Connecticut, and North Carolina. Three of these teams are perennial championship contenders and Siena is one of the most successful mid-majors of the past few years. The rest of the teams near the top of the list are a little more scattered. The top thirty teams on the list are an odd admixture of regular contenders (Syracuse, Florida, Arizona), mid-major spoilers (College of Charleston, George Mason), and middling teams of all sizes (Central Connecticut, Samford, Notre Dame, Boston College). So that’s interesting, but inconclusive.
But, what about the bottom of the list? Is there any discernible trend amongst the teams that foul the most? Well, here, the picture is clearer. The bottom seventy-five or so teams are almost exclusively small conference schools that have had middling success at best, though the true booby prize goes to Central Michigan, who, from 2006-10 averaged an opponent free throw rate at an astonishing 54.3%. So while it’s unclear that rarely fouling is the true mark of a great team in college basketball (though it does appear to have helped some very good teams), fouling a lot seems to be the mark of pretty bad to thoroughly mediocre teams.
Except, of course when it isn’t. While the bottom of the list is filled with the names of March afterthoughts, a few names stand out. Villanova. Kansas State. Georgia Tech. Each of these schools has fielded some very good tournament-ready teams in the past five years. The question then becomes why they are towards the bottom of this list. It would be easy to dismiss one year of frequent fouling, but five years points to a coaching philosophy or theory behind the fouls, and more interestingly, a deviation from the semi-conventional wisdom of avoiding fouling.
One of the answers seems to be a Four Factors trade-off. Kansas State and Villanova have excelled in forcing turnovers in recent years and the fouls seem to be an unfortunate side effect of such overzealous defense. In this scenario, Jay Wright and Frank Martin have accepted a higher rate of fouling as an acceptable byproduct of forcing lots of turnovers, which they have decided is far more important to their defense. This is potentially a very reasonable thought. Unfortunately, it might not hold up to scrutiny. Last February, John Gasaway at Basketball Prospectus puzzled over Villanova and its foul-happy ways and made a pretty persuasive case that fouling so frequently was destroying Villanova’s defensive efficiency. Furthermore, the existence of many schools that foul less than Villanova and Kansas State and force even more turnovers seems to undercut the idea that fouling is necessary to maintain a high rate of forced turnovers. At this point, it seems like Villanova and Kansas State have passed the point of diminishing returns on their trade-off: they are giving up far too many free throw shots to justify the marginal increase in opponent turnovers. If either school can manage to foul less, they are looking at an easy increase in defensive efficiency, and so far, so good: this season, both teams are above-average in opponent free-throw rate. So congratulations are in order for the two sets of Wildcats for joining the ranks of the hack-less.
The case for Georgia Tech is potentially different. While the Yellow Jackets may be fouling in an effort to induce turnovers (at which they are good but not great), there is a potentially more interesting scenario in play. The folks at From the Rumble Seat did an analysis last season where they noticed that Tech seemed to be fouling the worst free throw shooters on opposing teams in what seems to be a fairly systematic way. Now, while fouling bad free throw shooters at the end of the game is the standard come-from-behind tactic in basketball, routinely and deliberately fouling bad free throw shooters throughout the game is a fairly novel and odd practice (except to Jimmy Valvano). This is particularly interesting, because recently I learned (from the folks at the Hang Up and Listen podcast) about a similar tactic described much more fully by MIT’s Dr. Walter Sun in a paper that explains how a team can gain an edge over other teams by systematically fouling the opposing team’s worst free throw shooter during the bonus (but not double bonus) in an effort to drastically decrease the opponent’s average points per possession for these few qualifying possessions. It’s a clever strategy and almost certainly true, but the question is whether anyone in college basketball is actually using the tactic; and more specifically, if Paul Hewitt and Georgia Tech are doing so. I can’t say that for sure, but I will say this: Dr. Walter Sun got his undergraduate degree from Georgia Tech.