The problem with the majority of college football polls is that they are comparing apples to oranges. Voters are forced to pick one over the other, even though there is no empirical, head-to-head way to compare the two.
Let’s imagine that the Big 12 South is a barrel of apples, and the SEC East is a barrel of oranges.
So, amongst the barrel of apples there are three that are all equally sweet, one that is just a little more tart than the first three, and two that might be useful as pie filling at best.
In the barrel of oranges, there is one that is extremely sweet, another almost as sweet, and four that are barely worth squeezing.
Which is better, one of the three sweetest apples, the sweetest orange, or an entirely different fruit altogether, like a California grape?
I think each of us would have our own favorite. Do we even know if the worst orange is better that one of the best apples?
We don’t really.
So how could we compare the shiny apples of the Big 12 to the ripe oranges of the SEC?
One way would be for the teams to play each other, head-to-head. Another way would be if teams in the two conferences played each other, and the outcome of those games could be used as the common denominator.
Unfortunately, there was only one such game this year, a 52-10 beat-down of the Arkansas Razorbacks when they visited the Texas Longhorns. What did that tell us?
Well, nothing really.
The week before the inter-conference beat-down, the Alabama Crimson Tide visited Arkansas and won 49-14, and the week after the Florida Gators visited Arkansas and left with a 38-7 win.
Factor in home field advantage, and this basically tells us that Arkansas is not very good, and Texas, Alabama, and Florida are all pretty good.
There is no way to convincingly determine which team is the best based on this one inter-conference meeting.
In contrast, there are 10 SEC vs. ACC games this year.
Even though the conferences have split the series 3-3 with four games remaining this season, these games serve as a strong basis of comparison for the two conferences and have a significant impact on the championship picture and bowl selections.
One inter-conference matchup is clearly not enough. 10 is definitely overkill.
I’m convinced that there is a magic number of inter-conference games that would produce a strong and reliable basis of comparison between the conferences, and if major conferences were required to play a minimum number games against each other, a computer poll utilizing a resulting comparison matrix would result in a “perfect” ranking system.
The Colley Matrix ranking system would be a good example of such a poll and could become that perfect ranking system if this magic number is achieved.
Even non-computer polls like the AP and Coaches polls would become more accurate—giving voters more information for comparing the different conferences.
This would eliminate the need for playoffs and perfect the BCS ranking system without sacrificing ticket sales, academic needs, or that special thing that makes college football great.
Traditional rivalries such as Florida-Florida State, Georgia -Georgia Tech, and South Carolina-Clemson could remain, while new rivalries between the ACC, Big 10, Big 12, Pac-10, SEC, Big East, and others would grow.
This year, the Texas Tech Red Raiders opened with wins against the Eastern Washington Eagles, Nevada Wolfpack, Southern Methodist Mustangs, and Massachusetts Minutemen.
They rose as high as No. 2 in the country, even though there was no way to compare Texas Tech to schools in the other major conferences. The undefeated Utah Utes are having a related, yet opposite, problem and could well be the best team in the country.
We may never know.
I do not know what this magic number should be, so let’s make an educated guess. The average team has four non-conference opponents. What if we require that for each school at least one game must be against a team in one of the other six BCS conferences (ACC, Big East, Big 10, Big 12, Pac-10, and SEC)?
For the SEC, this would result in at least 12 non-conference games against the other five BCS conferences. Now, instead of allowing 10 of these games being SEC-ACC matchups, cap that series at four games, and spread the remaining eight non-conference games equally (two games each) against the Big East, Big 10, Big 12, and Pac-10.
Does this sound far-fetched?
The SEC has 15 games against other BCS conference teams this year, and would only need to swap three ACC games for one Big 12 and two Big 10 games to meet this arbitrary requirement. This season, that one additional SEC-Big 12 match-up could become the key to perfecting the ranking system.
Based on the previous two BCS Championship games, the two additional SEC-Big 10 matchups might not have made any difference. However, the Big 10-Pac 10 games seemed huge and definitive this year.
If the USC Trojans and Ohio State Buckeyes had not met earlier this year, OSU could still be ranked as high as No. 2 and still in the title picture for the third year in a row.
Imagine the impact if the Texas Tech replaced Massachusetts with just about any team from another BCS conference. That one game could have had a major impact on the championship picture.
“One” and “two” are the magic numbers. Each team in a BCS conference should play at least one non-conference game against a BCS conference team, and each BCS conference should have at least two games against each of the other BCS conferences.
To complete the entire picture, the same requirements could be applied between the BCS conferences and the mid-majors/independents.
We are not that far away.
In college football, six degrees of separation is too many.
I say let the conferences go head-to-head, creating more basis for comparison. Only then, will we be able to have a perfect ranking system and know who truly should be playing for that 2009 Bowl of Oranges.