Support for that argument came in 2004, when Auburn started the season ranked 17th and went undefeated. The Tigers were shut out of the BCS title game because USC and Oklahoma started the season at No. 1 and No. 2 and also went undefeated.
A closer analysis of the numbers, though, shows that delaying the opening polls might not solve anything.
I studied the AP and USA Today polls from the past five years, and compared the preseason Top 10 to the Week Five Top 10 (or the Week Four Top 10 in 2007). I then measured how many preseason Top 10 teams were still in the Top 10 in Week Five.
The results are recorded as percentages:
Year AP Poll USA Today Poll
2003 50% 60%
2004 80% 60%
2005 60% 60%
2006 70% 70%
2007 70% 80%
Both polls graded out at 66 percent correct from Week One to Week Five. In other words, they were correct on 6.5 teams in their Top 10 before the season started.
In all five seasons, the No. 1 team in the nation stayed on top. No. 2 stayed put in three of the five years.
Michigan gets the title of Most Overrated Team. Four times the Wolverines were ranked in the preseason Top 10, and all four times they dropped out by Week Five. (In 2006, they started at 14 and 15 and climbed to No. 6 by Week Five.)
Three times a team ranked in the Top 10 in the preseason was unranked by Week Five: Michigan in 2007, Oklahoma in 2005, and Auburn in 2003.
My point here?
It doesn't matter when the rankings come out—they're going to be basically the same.
In 2004, it didn't matter where Auburn started—USC and Oklahoma didn't lose. The Trojans and Sooners would have been ranked No. 1 and 2 to open the year whether the polls came out in August or October.
The ONLY solution here is something fans are pleading for—something used by EVERY other sport (including other divisions of college football) to determine champions.