Wednesday, October 13, 2010

Andy Staples Shouldn't Be Allowed To Vote In Polls

Sports Illustrated used to be the gold standard of sports writing. It's amazing how many great writers they've had over the years. But as with all magazines, they're struggling in the age of the Internet, and their writing quality has really gone down hill. And Andy Staples wrote a horrendous article for them today (read it here), which was so bad that I feel the need to tear it apart.

It's one thing for a writer to be wrong, but it's another for them to be wrong while simultaneously being so self-righteous about it. Andy Staples writes about a problem he calls a "disgrace" by some people who have "shamed" their alma maters. He says that people have "kill[ed] our credibility". What's he so furious about? Alabama being ranked ahead of South Carolina in the AP college football poll.

I already wrote about this a few days ago, and I'm going to recommend Staples read this essay that I wrote. The essence is, head-to-head is meaningful, but not the be-all end-all. Upsets happen. And the thing is, people know this. In the same article, Staples gives his own Top 25 rankings. He does put South Carolina ahead of Alabama... yet he also puts USC ahead of Washington even though Washington has beaten USC. Heck, he has Virginia Tech listed but leaves out James Madison. How is that fair? It took me about five minutes to find to his AP voting record from last year. It turns out he put a 10-3 Georgia Tech team 12th, and didn't rank at all a 9-4 Miami team that beat Georgia Tech 33-17 earlier in the season. Any time Team A beats Team B which beats Team C that beat Team A, which happens anytime sample sizes get large enough, we see the idiocy of the ideology that any team must be ranked behind any team in loses to.

As I've written many times, it's because deep down the voters understand what I'm saying. They understand that upsets happen, and they're supposed to be ranking teams by how good they are. Yet while they understand that fact, they somehow turn their minds off when two teams that are close in quality to each other play. They get all self-righteous about it: If Team A beats Team B then by golly Team A must be better!.... unless we already know Team B is better, of course.

And in this case, almost everybody agrees Alabama is far better than South Carolina. Sagarin still ranks Alabama the second best team in the country, while South Carolina is back in 12th. If those two teams played again on a neutral field, Alabama would be favored by 5-to-7 points.

I don't want to linger on this issue too long, but I do need to quote what might be the dumbest paragraph I've ever read from a Sports Illustrated writer:

-------------------------------------
The poll forces us into a lot of silly hypotheticals. Would Boise State beat Ohio State on a neutral field? Who really knows? Sometimes, though, we have actual on-field results to inform our voting. We don't have to wonder what would happen if Alabama and South Carolina played. We know.
-------------------------------------

Unbelievable. In the first three sentences he is saying that he shouldn't be allowed to vote. If he has no idea whether Ohio State is better than Boise State, why exactly is he qualified to vote on that, again? We only need to come up with 60 voters for the AP Poll in the entire country... they couldn't find somebody else? And in the latter part of the paragraph, he appears to be asserting that if Alabama and South Carolina played again, the exact same game would happen. That's a pretty bizarre claim. In fact, if those two teams played again we can take a pretty good guess what will happen: Alabama will win, because they're the better team. Even great teams sometimes lose to inferior teams, so South Carolina might win again, but almost every expert would pick Bama.

If Andy Staples really wants to dispute that last paragraph, I have a bet for him. There will be thousands of different match-ups that will happen twice during the college basketball season. For example, Duke and North Carolina will play twice in the regular season. Andy is free to pick any of those match-ups. I will bet him that when those two teams play a second time, the final score will be different from the first game. Any amount of money, Andy. Come on, we already know what happens when those two teams play!

I apologize to those who feel like I harp on this issue too much, but hey, we're in the silly season right now! It will be a couple more weeks before we even have exhibition games. Once the season gets going I'll be back on task. But with not much in the way of college basketball news today, I couldn't resist tearing apart that Andy Staples article.

No comments: