Monday, February 05, 2007

The AP top 25 poll

I used to be really into college basketball from around 10th to 12th grade, and oddly enough this year, I’ve gotten into it again. That means I’ve never followed college basketball closely pre-March Madness while I was in college, but I just never went to schools with good basketball programs and I just shifted my interest to the NBA.

This year, I started following college basketball again. I don’t know why, exactly. I think it’s an interesting time because people who would have been in the NBA are now forced to do at least a year in college, so that makes it all very interesting and we know that at least some of the players are NBA-worthy. Anyway, one thing that used to annoy me back in the day and still annoys me now is the AP Polls.

It seems like the AP poll, operates under a fairly automatic system that works without much thought: Take last week's poll, move down 2-3 spaces the teams that lost, and move up the teams that won. It seems like little thought is given to the caliber of the opponent or much else. Ohio State and Wisconsin are even in the Big 10, yet Ohio State is sure to move up on them in the polls because they didn't lose last week, but that doesn't neccessarily mean they're the better team. Oregon is ranked 9th and while USC and Washington State are tied for second in the Pac-10 ahead of Oregon, they wouldn’t be able to move ahead of Oregon in the polls, because Oregon is ranked too high to drop that far even though USC demolished Oregon twice.

Even if a top 25 team loses to a team ranked higher than them, which would be the expected outcome, (i.e. Arizona losing to UNC, Clemson losing to Duke, or Washington State losing to Oregon) they move down in the rankings while mid-major teams like Butler, Air Force, or Nevada move up for beating mediocre teams. I think there are a lot of advocates for allowing the mid-majors to have a chance, but it’s ridiculous to think that because Butler beat Southwest Missourri and Southern Mississippi while Arizona split games against North Carolina and Oregon, Butler should move ahead in the polls because it didn’t have any losses. The AP Polls also discriminate against the mid-majors anyway because they’re basically based on the preseason. If you were the preseason #1 and you keep losing, you automatically have a much better chance of staying in the top 25 two months later than preseason #20, because the voters will keep moving you two spaces down each time you lose. Preseason #20, meanwhile will be out of the polls in a couple weeks. Preseason #1 is usually going to be a team from a major conference and the mid-majors, if they make it all, will likely fill out the lower ranks.

Fortunately, the seeding for March Madness isn’t done by the polls but by a committee and by the RPI, which is probably something like the BCS system. I think any computerized system is going to be much better than the AP Poll because it doesn’t suffer from human problems of having short-term memory and the like. I don’t know much about college football, but if the BCS system places teams mathematically, that’s gotta be better than human judgement in this case. People tend to complain when the AP Poll disagrees with whomever the BCS assigns as #1 and #2 but if the AP Poll in Football is anything like the AP Poll in Basketball, I’d much rather trust a computer.

No comments: