October 18, 2010

A Rant about Rants about the BCS

I just saw This article on the front page of ESPN, written by Mark Schlabach. Feel free to read it, but the gist of the article is that The Computers Just Don't Understand and certain teams (i.e. LSU) just don't pass the mythical Eye Test.

Obviously, those hard drives and keyboards haven't watched the Tigers try to operate their offense in the closing seconds of a game.

Those computers were apparently rebooting when the Sooners struggled to beat teams like Utah State, Air Force and Cincinnati earlier this season.

You know what? Go fuck yourself with your computer ranking strawmanning. As if all the poll voters have watched the LSU games. Or all the other games for that matter. Or take strength of schedule into account in any meaningful or uniform way.

I'll be the first to admit that I would rather the BCS ONLY used computer rankings, but I'm in the minority here. Look, the computers have biases too (i.e. whatever model was used to build them), but the important thing is that THE BIASES ARE CONSISTENT. I can't stress this enough. There's nothing more arbitrary than an "Eye Test" especially when one of the main polls used is a coaches poll that is often foisted off on assistants or used for political gain within the system. No one at any of these schools watches all the games.

Besides, the reason why we do a weighted average of the computers and the bullshit polls is exactly the reasons why he's complaining - that computers can't see OU's and LSU's struggles, just as human beings can't watch every play of every game and adjust for strength of schedule with any sense of order or consistency.


AK said...

Totally agree. I wish they'd stop referring to them as computer rankings. Yeah, they use computers to calculate the totals, but it's not like a computer does the ranking, it just runs a statistical formula of some person's/people's own devise. They should label them subjective and objective rankings, because that's exactly what they are. There are advantages to both methods, I suppose, but the chief difference is the matter of objectivity vs. subjectivity, not man vs. machine.