An "unprecedented six-month investigation" by Sports Illustrated and CBS News into the criminal backgrounds of players at the nation's top college football programs is attracting a lot of attention today, but we can't figure out what to make of it. Of the 2,837 players on the rosters of teams ranked in last year's top 25 preseason poll, 204 "charged with or cited for a crime." That's roughly equal to 7 percent. Is that a lot? It seems reasonable when you're talking about college kids, especially if citations--a large chunk of which were undoubtedly issued as part of underage drinking busts--are included. SI and CBS did not "have access to juvenile arrest records for roughly 80 percent of the players in the study," but one wonders if the findings were skewed by the 20 percent of players they did have data on.

So when Jeff Benedict and Armen Keteyian say the findings lend credence to the "pervasive assumption that college coaches are willing to recruit players with questionable pasts to win," it is a debatable point. What's not up for discussion is the limited impact felonious footballers have on gridiron success--if the coaches are turning a blind eye to troublemaking to win games, they might want to reconsider the strategy. Consider the University of Pittsburgh, which led the way with 22 players charged. They finished 2010 with an 8-5 record and their coach got fired.

We compiled a scatter plot of the data. The x-axis is the the team's preseason rank, the y-axis is the number of players with criminal histories. We're not statisticiains, but this graph appears to show absolutely nothing.


The win-at-all costs coach is a convenient--even intuitive--straw man, but will a coach really bring in a player he thinks will cause trouble? On the recruiting trail, nobody seeks out a wide receiver who will some day beat up his girlfriend in public. One wonders if there are some confounding variables here, driving both the emphasis on athletics and the citations. So we took a very unscientific look at the Greek system, and whether it's having an effect. It turns out 10 percent of Pitt's student body is part of the Greek system. At Iowa, Penn State, and Arkansas the percentages appear to be 11, 12, and 21, respectively. However, commuter school Boise State, ranked fourth for most player citations, is only two percent Greek.

So we remain unsatisfied. Maybe it's the quality of the comparative literature departments.