Wednesday, October 26, 2005

Death to Humans

It's almost amusing watching the BCS writhe around in its death throes. Year after year, we participate in the great tradition of finding out how the newly-revamped BCS system will crash and burn and who the unfortunate victim of the debauch will be. While it really should have no effect on Notre Dame's season this year, it's still interesting to note that the Irish are yet another bemusing piece of roadkill from the runaway train that is the BCS's latest rankings.

As we all know, the good-hearted, God-fearing, only-slightly-corrupt individuals whose lives are devoted to this sport have determined Notre Dame's position in the pantheon of college football is between 9 and 10. However, the evil, sadistic computers that are hell-bent on world domination and enslaving humanity feel the Irish belong down in the muddle of bottom-feeders that unceremoniously round out the top 25.

Now HRB is not condemning all computers. It would be very hypocritical of us to do so considering Notre Dame is led by an evil robot genius. But the system of utilizing computers is fatally flawed. Here's HRB's take:

1. There is no necessity for computer rankings. Every year, the BCS hamstrings and tweaks the rules for computer rankings as a result of a disappointing result from the previous year. Why? Because the computers went haywire and threw a monkey wrench in the conventional wisdom that is wrought out in the human polls. It is clear that the BCS wants the computer rankings to be little more than a way to objectively validate what the polls dictate, so why even have them?

2. You cannot make a matrix more accurate by limiting its input. Hello. This should be obvious. This is like the AP telling its voters they can only watch one football game a week, otherwise they might draw upon too much information to come up with their rankings. Yet, somehow, the BCS mandate that Margin of Victory be removed from polls miraculously passes the straight-face test. The company line is that it discourages running up the score, but coaches still run up the score. If we're going to have the indignity of computer rankings, they should at the very least be accurate, or at least not have their accuracy actively reduced by the BCS committee. An accurate ranking system is an accurate ranking system, even if it's based its formula on the astrological signs of the players.

3. There is just no substitute for blind speculation. Look, ranking teams is an imperfect, almost purely subjective task. So let's quit this charade of pretending it's a science. Sure, the Colley Matrix can predict that 40 midgets would beat one lion in a fight based on the inverse square of their opponents' opponents records, but how much more credibility does a computer have when it comes to blind speculation? At least I can provide a litany of vague arguments about the favorable matchups and strategems that would give the edge to the lion.