The College Football Bowl Picture

Over at Inside Higher Ed, they have a piece looking at the state of college football as we enter bowl season. This is dominated by two large tables of numbers, one good, and one bad.

The first table is the good one, as it explains why the college football "championship" is so messed up. It lists the 32 bowl games that will be played over the next month, and the per-team payout for each. The five major BCS bowls pay each team $17 million, which neatly explains why the college football elite are unwilling to put in a playoff-- in any real championship system, they might end up having to share that money around more evenly.

(This table also give the lie to the common claim by college footbal fans who say that there would just be no way to choose teams to go into a playoff-- you have thirty-two bowl games. You're already picking 64 teams for the "postseason"-- explain to me again how a playoff system is going to be the death of Western civilization?)

The bad table is the second one, which lists all the schools involved in those bowl games, and their graduation rates for black football players, white football players, and all athletes. The problem I have isn't with the idea of making this information public, but rather that they're doing the wrong comparison, again.

As I've complained before, if you want to use graduation rates to complain that minority athletes are ill served by colleges and universities, you need to compare similar groups. Yes, it's sad that there's such a large gap between the graduation rates for black athletes and white athletes, but the relevant number isn't the comparison between different groups of athletes, but between athletes and non-athletes of similar backgrounds.

If you want to use graduation rates to make a case that Athletics Are Evil, you need to show not only that there's a gap between the graduation rates for black and white athletes, but that the gap is bigger than the gap between black and white non-athletes. Otherwise, all you've shown is that colleges and universities in general do a bad job educating students of different racial groups, and, well, we knew that already.

To put some numbers on this, the IHE table shows big discrepancies for the two teams in the "championship" game-- Ohio State's black football players have a 43% "Graduation Success Rate" compared to 74% for white football players, while Louisiana State's split is 42%/70%. That looks pretty damning.

But the correct comparison is between atheltes and non-athletes of similar background, and there things don't look quite as clear. The numbers above are the "Graduation success Rate," which is a bullshit NCAA statistic, and not calculated for non-athletes, but you can compare athletes and non-ahtletes in general using federal graduation rates from the NCAA web site, and what you see is that athletics doesn't really seem to make much difference.

Using overall graduation rates, the black/white split for atheltes at THE Ohio State University is 42%/60%, while for non-athletes it's 40%/64%. For LSU, the athletes are split 36%/52%, while non-athletes are at 42%/56%. That doesn't look like a big change-- the racial divide is very slightly worse for athletes at LSU, but it's slightly better at OSU.

(To head off an obvious possible objection, the number of black athletes is around 10% of the black student population at both schools-- 577/845 for OSU, and 75/680 for LSU, so it's not that the athletes are pulling the numbers down for everybody else. White athletes are a tiny fraction for both schools.)

This isn't an exact comparison-- it's possible that the football team has a much worse graduation rate gap than the general student body, but this is being concealed by the high graduation rates for fencers or wrestlers or somebody else. I doubt it, given the size of college football teams, but it's possible. The only way to know for sure would be for somebody to pull the graduation rate numbers for the football team specifically, and compare them to the same numbers for the student body as a whole. (I can't find those at the moment, because the NCAA's web site is a world of suck, and I have other things to do today. I do have those number for Maryland and Duke, from the previous post, and the football graduation rate for those two schools is not significantly different from the overall rate for athletes, but I don't have a sport-by-sport breakdown by race.) Say, maybe, somebody sitting down to write an article about racial divisions in graduation rates of big-time college football programs...