Dear Jeff Anderson, Richard Billingsley, Wesley N. Colley, Chris Hester, Kenneth Massey, Jeff Sagarin, and Dr. Peter Wolfe: Don’t take this the wrong way or anything, but I have some issues with the way y’all do things. Shall we chat?
Exhibit A: the ESPN.com BCS computer rankings. You’ll see the normal columns, the total %, and the avg. rank, but I also added a column for disparity between the highest and the lowest computer rankings for a team (assuming 26 for all zeroes), as well as a column beside each computer ranking (the green columns) showing the disparity between that specific ranking and the overall BCS ranking.
1) 7 of the 25 ranked teams, or 28%, have a disparity of 10 or more. Let me repeat that: Twenty. Eight. Percent.
2) By comparison, 7 teams also have a disparity of less than 5. If you take out WVU and Southern Miss (which only have 1 non-zero ranking), then there are 5 teams. 3 of them are the top 3 teams in the country. The other ones are located in the Palmetto State (South Carolina and Clemson).
3) The average disparity is 6.6 – easily a wide enough discrepancy to take a team out of BCS contention (or put them there).
4) This also shows the disparity between the computer and human rankings. For example, the overall 5th-ranked team in the country, Virginia Tech, is ranked #18 by the Jeff Sagarin system. On the other side of the spectrum, the overall #17 team, Baylor, is ranked #7 by Jeff Sagarin’s and Kenneth Massey’s systems.
5) Out of the 136 overall computer rankings (150 total slots – 14 zeroes), 5 of those computer rankings differ by 10 or more from the overall BCS rankings. 4 of those 5 are from Jeff Sagarin’s system.
6) Richard Billingsley and Dr. Peter Wolfe seem to have the closest systems to the overall rankings: both have 9 rankings that are identical to the overall BCS rankings, and both are +/- 4 or less on each ranking with the exception of one +/- 6 for each system.
“But Armondo”, you start, “all this is just mathy mumbo-jumbo. I like stats and figures just as much as the next man, but what does it mean?”
It means that the current system is far from perfect. It means that the fate of several schools, including the amount of money, recognition and accolades that they have a chance at receiving, is partially determined by computer algorithms, or complex formulas, with the names of 7 individuals attached to them. Why them? Why do they get to choose? I concede that there is no “perfect” computer ranking system, and by “perfect” I mean one that would make everyone happy. Folks would whine and complain regardless of the system. But why multiply that by 6?
Here’s my suggestion: one system, right-or-wrong. Take the system of the 6 that, over time, has come the closest to the overall BCS rankings, and accept that system as gospel. Even open it up to the world to do better – if you can write an algorithm that can come closer to the overall rankings than the current 6 systems can, then you’re in. You get a fat check, a hearty handshake, and the knowledge that your system is making all the decisions now. I guarantee that something better and verifiable would surface.
Also, no more of this black-box crap – the algorithm would be completely open, for all the world to see, including coaches and fans. Every year, the algorithm would be reviewed and revised, if necessary, via a majority vote by a group of ALL FBS coaches and statisticians; that way, everyone can still have their say. Any changes would not go into effect until the season AFTER next, so teams have a chance to adjust accordingly.
Finally, human pollers would also be 86’d. Again, the computer system of choice would not be perfect, but it would at least be consistent, and able to evolve with the game via the yearly review. There’s no question that there’s human bias – it’s only natural – and it’s difficult to operate outside of those biases. This way, it would be a non-issue altogether. The AP and USA Today Coaches Polls can still exist alongside the BCS rankings, and cna still have their own awards; they just would have no bearing on the BCS anymore.
See? An entire solution that doesn’t even mention a playoff. Easy-peasy. Now it’s off to go devour some highly-ranked cake. (And don’t worry about disparity; it’s #1 across-the-board.)