The USA Today and Harris polls make up two-thirds of the formula to determine the BCS rankings. So while computers get the brunt of the bashing from college football pundits for their roles in the BCS controversies, it's now the polls that have the most bearing on who plays for the title. Polls used to make up a quarter of the formula for the BCS rankings from 1998-2003, but they still played a large role in determining who went to the BCS title game.

Let's take a look at what role the polls have played during the biggest controversies in the BCS era:

1998: Ohio State is rated ahead of Florida State in the preseason poll, but loses later in the season in a shocking upset against Michigan State at home. Florida State lost at home to an N.C. State team that went 7-5, but since the Seminoles' loss was earlier in the season, they're ranked higher in the polls and play Tennessee for the national title instead of Ohio State.

2000: Miami loses to Washington in the second week of the season, but the Huskies lose to Oregon on the last Saturday in September to quell their excitement. Later in the season, Miami defeats Florida State. All three of those teams end the season with one loss, so which one do you think ends up playing Oklahoma in the national championship game? Florida State, who was ranked the highest of those three teams in the preseason poll. Disclaimer: Miami was ranked higher than Florida State in the polls before the title game. However, Washington was lower than Miami. Tell me how that made sense.

2001: Despite getting trashed in the Big 12 championship game by Colorado, Nebraska earns a berth in the national championship game where they also get destroyed by Miami. Nebraska earns the berth over an Oregon team that won the Pac-10. Guess who was ranked higher in the preseason poll? Nebraska.

2003: The polls seem to have no effect this season, as USC is shut out of the national championship game despite finishing on top of both the USA Today and AP polls. The Trojans defeat Michigan in the Rose Bowl and finish the season No. 1 in the AP poll ahead of BCS title game winner LSU. It was this debacle that led the BCS to tweaking its formula to the current setup. It also eventually led to the Harris poll replacing the AP poll in the formula.

2004: USC and Oklahoma start the season ranked first and second, and they run the table to meet in the national championship game. Auburn also finishes the season undefeated, but doesn't get a shot at the title game because of the two teams ranked ahead of them all season. USC whips Oklahoma in the title game, and everyone is left wondering if Auburn would have given the Trojans a better fight. Granted, this is an extreme case and hardly happens, but it exemplifies the problems with the system.

2006: Despite not playing the week before, Michigan drops below Florida in the polls and misses a trip to the title game to play Ohio State. While Michigan lost to Ohio State in their final game and Florida indeed proved to be a worthy titleholder, dropping a team that didn't play the week before doesn't make a whole lot of sense. What did they do to get dropped?

These examples point out the two major problems with the polls: they are released too early in the season, and teams are penalized for losing later in the season.

There's no doubt that a team's preason ranking makes a difference. Georgia head coach Mark Richt admitted as much on Monday when he said the higher you start in the polls, the better chance you have of ending up high in the polls.

Why not wait until three or four weeks are gone in the season to release the first poll, as the Harris poll does? It makes no sense to release a ranking of teams that haven't even started practice yet. A perfect example of this was last season, when Michigan was ranked in the top five in the preseason poll, yet lost four games, including the season opener to Football Championship Subdivision (or Division 1-AA as I like to call it) Appalachian State. No one really knows how good teams are going to be until they take the field, and there is no point building up a team's reputation in the eyes of fans and the media by giving them a lofty ranking before they have even hit the practice field.

There is also no reason why a team should be penalized for losing later in the season. While the current system has done a better job of eliminating this problem in recent seasons, teams still often drop in the polls simply because they lost, even if it's a close game against a team ranked higher than them. Sometimes, teams fall below schools they have beaten earlier in the season with the same number of losses.

If the polls are going to be a large part of deciding who plays for the national championship, these are issues that are going to have to be fixed. Releasing the first poll later in the season is a good start.