This is part one of a two-part look at the assumed worthlessness of the pre-season poll.
Yesterday we investigated some of the biggest problems with the polling system. Between pollsters who don't watch the games, unclear voting instructions, and the challenge of digesting all of the results it's quite a shock there aren't more problems like the one we saw in 2004.
Opponents of pre-season polls would have us start the poll in week four, eight, or whatever. They'd claim that this, somehow, would fix the problems. How is that going to work? The coaches and media can't even faithfully evaluate one single week of games, how are they supposed to evaluate four, six, or eight all at once? One of two things would happen: they'd either go with their instincts in week eight, or they'd keep their own personal poll from week to week.For those in the former category, the problems would be many: how on earth can you evaluate eight weeks of football with any consistency? They'd end up relying mostly on records and margin of victory, something computers can do better, faster, and more accurately. This baseline would be altered by the slant of the media covering college football, which wouldn't eliminate pre-season polling, it would just change its form.
Besides, no action of the BCS will eliminate pre-season polls. Pretty much every news outlet and blog does their own pre-season top 25. Why? Because the off-season is long and boring, that's why. This will not change. There will still be polls out there to crib off of, and the voters will. Week eight would show all manners of "research" from voters in the form of taking a gander at the polls and seeing which ones they like best. Does this really solve the problem?
Then, of course, we have the folks who keep their own polls to make their life easier in week eight. I have a feeling a lot of voters would choose this method.
Either of those is functionally identical to a pre-season poll and, in some cases, worse. What happens to the team that gets no media attention, goes undefeated through the first eight weeks, and then winds up in the lower half of the top 25 in the first poll because they're from a non-BCS conference, didn't end up on ESPN very often, or otherwise flew under the radar? Answer: they're going to have to struggle, even if they go undefeated, to even make it into the BCS standings, let alone play for the title. Is that a better solution?
Sure, if we assume that waiting until week eight has all of the benefits that its proponents claim, this can be helpful for the teams that are already in the national spotlight, but what are the odds that Boise State would've played Oklahoma in the 2007 Fiesta Bowl? Does this change really help anyone EXCEPT the big boys?
And wouldn't that just increase the emphasis on cupcakes? When week eight rolls around, would you rather have one close loss to a great out of conference opponent, or be undefeated with some token wins over cupcakes? Will anyone outside of your conference or region even remember that epic week one match-up? Doubtful.
Still, it's clear that there are issues with the pre-season poll. But are they necessarily alleviated by just moving the poll back. What happens to Auburn in 2004 if the polls started in Week Eight? In the first eight weeks of the season, Auburn played two teams of note: LSU and Tennessee. They beat LSU by one point. By Week Eight, LSU had two losses.
"But Pete," you might be saying, "LSU was #4 and Tennessee was #10 when Auburn beat them! That has to count for something!" You'd think that, but you'd be wrong. In our hypothetical world, nobody is ranked anything. All we know is that Auburn went undefeated through their first seven games, beating some pretty good SEC opponents and trashing cupcakes like Citadel. Would they REALLY have been ranked better than #4 going into week eight against Kentucky? It's certainly not a foregone conclusion.
After all, in order for Auburn to have a shot to play for the BCS title in 2004, your delayed polling has to find a way to place Auburn over Oklahoma in the final polling. Were Auburn's first seven games that much more impressive than Oklahoma's? Would people who, at the beginning of the season thought Oklahoma was a better team suddenly change their mind in Week Eight when the "first" poll were to come out? Keep in mind that, should records have such a strong impact, there were five teams in 2004 that ended the season undefeated and more who survived through eight weeks. How do those get sorted out?
It seems tremendously unlikely that just moving the polls back would have had such a huge impact on the final standings in 2004.
So what would?
In short: more agility in the polls. Getting rid of the notion that it's taboo for a team to win and drop or lose and move up. Eliminate the infrequency with which teams jump other teams when both have won. Make the polls measure something specific. Set a goal and specific question, since "who are the 25 best teams" is truly unanswerable.
The pre-season poll is an easy target, but getting rid of it isn't going to solve the problems.