Tag:2012 BCS Bowls
Posted on: January 5, 2012 4:51 am
  •  
 

What's to be done about 'rogue' AP voters?

Posted by Adam Jacobi

A report came out Wednesday night that some AP voters were prepared to vote LSU as the national champion even if Alabama beats the Tigers at the BCS Championship on January 12. There are conditions, of course; if 'Bama wins handily, there's not going to be much doubt who the deserving national champion is. But still, if the title game is another close, unconvincing affair that this time tilts in favor of Alabama, there are people on record who are at the very least open to the prospect of sticking with LSU.

"Awarding a championship to a team that loses its final game is beyond counterintuitive and may be un-American," said David Teel of the Daily Press in Hampton Roads, Va. "But if LSU loses narrowly, I will absolutely consider (voting the Tigers No. 1). That's how good the Tigers' regular season -- five wins over the top 25, four away from Death Valley, including at Alabama -- was." Another voter in Albuquerque told CBSSports.com's Dennis Dodd that Alabama's win "would have to be like 63-0 or something" before he'd consider voting for the Tide over LSU.

[Doyel: Splitting BCS national championship 'stupidest idea ever']

The conundrum Teel raises along with his supposedly "rogue" compatriots is a real one, and one that cuts to the core of polling as a college football institution. At the end of the day, though, Teel is not only well within his right to wonder aloud about this game's effect on his final ballot -- if the conditions are right, he should follow his gut and go with LSU to win the title.

First, it's important to understand why polling even needs to exist in college football (which it does!) in the first place. The validity of determining a Top 25 in college football is dramatically hindered by two factors:

1) We just don't have much data to work with. Assuming one of the central maxims of college football and the BCS is correct -- that the most important determinant in whether one team is better than the other is what happens when they play each other -- then in order to justify a two-team playoff out of a 120-team league, we would likely need way more than 12 or 13 data points for each team (especially with two-thirds of nearly every schedule dedicated to common games with a highly consolidated group of conference opponents). Baseball uses 162 games in a 32-game league, and this year, it needed all 162 just to determine an 8-team playoff setup.

Now, the point can be made that MLB didn't actually need all 162 games to determine its playoff participants -- nobody was screaming about major league baseball's illegitimacy when the season was 154 games long (or less) for the first 85 years of the league's existence, after all -- but if we extrapolate college football's rate of missing opponents to the MLB, the season would be four games long, three of the games would be dedicated to intra-division play, and the fourth game would be for one non-division opponent. And then two title game participants are chosen. If MLB commissioner Bud Selig proposed this, he would be fired. He would be quadruple-fired. Then the riots would begin.

2) The data we do have is highly contradictory anyway. Even if we had a season with dozens upon dozens of games, upsets are so prevalent that the rankings would still be a relatively poor predictor of future games. We all like to believe that if one team beats the other, it's better than the other team, but here's the full list of the Associated Press Top 25* teams that have not lost to a team ranked below them: LSU, Alabama, Oregon, Arkansas, Virginia Tech, Georgia, and Penn State. In other words, even among what voters have determined to be the best 25 teams, 76% are ranked ahead of a different team that beat them during the season, and it took only 12-13 games to get to that point. For the next 25 teams, the ones with even more losses than 1-3 on the year, there would be utter carnage in trying to only rank teams ahead of the ones they beat. Consider that the next time somebody makes the all-too-prevalent argument of "How can Team X be behind Team Y in the rankings when Team X beat Team Y?" 

Now, even though college football is filled with game-changing factors that hinge on chance (weather, injuries, fumbles) this pattern of teams routinely losing to worse teams is not a phenomenon unique to the sport. Going back to baseball, losses are so prevalent that even the best teams rarely win more than two-thirds of their games. In professional football, the teams with the best regular-season record are barely more likely to make the Super Bowl than the average playoff-bound team. But those two leagues (and every other professional team sport) feature multi-round playoffs, so the contradictions are rendered meaningless through the process of the playoffs -- even as said playoffs routinely eliminate teams that would take a BCS Championship bid if such a system existed in the league.  

College football does not have the luxury of expanding its schedule to adequately address either of the the above factors, especially in light of the FBS' mammoth number of programs -- football is debilitatingly brutal as it is, plus the prospect of trying to turn a profit in the postseason is prohibitively difficult for athletic departments even with a one-week schedule -- so it has to make do with its small, weak set of data in order to determine championship participants. In must step pollsters to interpret that data in their own way, and generally, those pollsters do a very good job of contextualizing the data and putting together a (temporarily) coherent Top 25 -- at least in the poll's weekly aggregations. So given the limitations of college football scheduling, there's really no other way to delineate between specific programs than by subjective ranking.

The rankings are each pollster's individual interpretation of the entire season, and if there's any doubt about that, regard the amount of teams that find themselves ranked second in the season's very final poll without playing in the BCS Championship because they won their bowl games while ranked third while the BCS Championship loser was thumped so soundly it couldn't hang onto the second-ranked spot. Those votes as No. 2 aren't protest votes to suggest that the BCS took the wrong team to challenge the top-ranked team or that a plus-one needs to be enacted immediately, they're reflections of each team's work on the season as a whole.

So given that, it's particularly backwards of the BCS and Coaches Poll to require that the winner of the BCS Championship be voted as national champion while allowing the loser to be ranked lower than second if need be. The season as a whole is what it is, and if AP voters determine that a potential slim Alabama victory over LSU at a (semi-) neutral site in the BCS Championship doesn't constitute enough of a reason to like Alabama's season more than LSU's, those voters should absolutely rank LSU first in their final ballots. They should be prepared to defend the decision, of course, but they should do it; otherwise, what's the point of being granted a vote in the first place?

*The AP Top 25 was chosen because the Coaches Poll and BCS exclude Southern California for reasons that are not germane to this particular topic.
 

Keep up with all the latest results and preview the rest of the bowls at CBSSports.com's Bowl Pregame. 

Keep up with the latest college football news from around the country. From the regular season all the way through the bowl games, CBSSports.com has you covered with this daily newsletter. | Preview

 
 
 
 
The views expressed in this blog are solely those of the author and do not reflect the views of CBS Sports or CBSSports.com