Jewstin
11/23/2008, 08:55 PM
I got really, really tired of having to read the stupid numerical system the computer polls use, so here are the top 5 teams of each computer poll (as per http://cfn.scout.com/2/582114.html ) with a breakdown of the net differences between us and Texas in all three polls in the BCS calculation, along with their BCS contribution.
Computer Polls:
(the lowest and highest is thrown out in the calculation, leaving 4 scores on a 25 point scale ... #1 is awarded 25 points, #5 is awarded 21).
Anderson and Hester
-------------------
1.) Alabama
2.) Utah
3.) Texas - 23
4.) Oklahoma - 22
5.) Texas Tech
Billingsley
-----------
1.) Alabama
2.) Oklahoma - 24
3.) Texas - 23 (thrown out)
4.) Texas Tech
5.) USC
Colley Matrix
-------------
1.) Texas - 25
2.) Alabama
3.) Utah
4.) Florida
5.) Oklahoma - 21 (thrown out)
Massey
------
1.) Oklahoma - 25 (thrown out)
2.) Texas - 24
3.) Texas Tech
4.) Alabama
5.) Utah
Sagarin
-------
1.) Texas - 25 (thrown out)
2.) Texas Tech
3.) Alabama
4.) Oklahoma - 22
5.) Utah
Wolfe
-----
1.) Alabama
2.) Texas - 24
3.) Texas Tech
4.) Oklahoma - 22
5.) Utah
In summary:
Texas - 23, 23, 25, 24, 25, 24
Oklahoma - 22, 24, 21, 25, 22, 22
There are four polls where either Utah or Texas Tech is ranked ahead of us (2 for the Utes, 2 for the Raiders).
BCS Point Contribution
Harris Poll: .0074 point advantage, OU -> + .002467 BCS contribution
Coach's Poll: .0275 point advantage, OU -> + .009167 BCS contribution
Computer polls: .060 point advantage, Texas -> - .020 BCS contribution
Sum those three, and you have the net difference between us and Texas in the BCS average - roughly .008363 points.
It's interesting to see the variations in the computer polls. Why the hell have they not implemented more of them is my question. Having six polls, then throwing out a third of them seems silly. If nothing else, these calculations need to be standardized and pooled into a single computation, or there needs to be more of them. In statistics, the more, the better, especially when the results appear so incredibly cracked-out. :D
Barring a major drop in the human element of the calculation, I think the computer situation will iron itself out next Sunday.
Computer Polls:
(the lowest and highest is thrown out in the calculation, leaving 4 scores on a 25 point scale ... #1 is awarded 25 points, #5 is awarded 21).
Anderson and Hester
-------------------
1.) Alabama
2.) Utah
3.) Texas - 23
4.) Oklahoma - 22
5.) Texas Tech
Billingsley
-----------
1.) Alabama
2.) Oklahoma - 24
3.) Texas - 23 (thrown out)
4.) Texas Tech
5.) USC
Colley Matrix
-------------
1.) Texas - 25
2.) Alabama
3.) Utah
4.) Florida
5.) Oklahoma - 21 (thrown out)
Massey
------
1.) Oklahoma - 25 (thrown out)
2.) Texas - 24
3.) Texas Tech
4.) Alabama
5.) Utah
Sagarin
-------
1.) Texas - 25 (thrown out)
2.) Texas Tech
3.) Alabama
4.) Oklahoma - 22
5.) Utah
Wolfe
-----
1.) Alabama
2.) Texas - 24
3.) Texas Tech
4.) Oklahoma - 22
5.) Utah
In summary:
Texas - 23, 23, 25, 24, 25, 24
Oklahoma - 22, 24, 21, 25, 22, 22
There are four polls where either Utah or Texas Tech is ranked ahead of us (2 for the Utes, 2 for the Raiders).
BCS Point Contribution
Harris Poll: .0074 point advantage, OU -> + .002467 BCS contribution
Coach's Poll: .0275 point advantage, OU -> + .009167 BCS contribution
Computer polls: .060 point advantage, Texas -> - .020 BCS contribution
Sum those three, and you have the net difference between us and Texas in the BCS average - roughly .008363 points.
It's interesting to see the variations in the computer polls. Why the hell have they not implemented more of them is my question. Having six polls, then throwing out a third of them seems silly. If nothing else, these calculations need to be standardized and pooled into a single computation, or there needs to be more of them. In statistics, the more, the better, especially when the results appear so incredibly cracked-out. :D
Barring a major drop in the human element of the calculation, I think the computer situation will iron itself out next Sunday.