I don't like it, but I'm not a statistician and I see no reason to reinvent the wheel.
My rankings are based loosely on many states' high school playoff qualifiers: teams get 1st level points for beating opponents, and they get 2nd level points for the teams their opponents beat. Teams get more points for beating Power 5 schools (1 point) than Group of 5 schools (0.5 points). And as far as I know, my system is the only one that actively penalizes FBS teams for scheduling FCS teams (-1 point). My opinion is that it gives an unfair advantage to that team in that it's almost an automatic win with no effort while other teams are taking risks. This is the same issue Urban Meyer has currently latched onto: scheduling parity. I happen to agree with coach - the CFP committee needs to set some guidelines on scheduling, and they need to do it now. All P5 conferences should be required to schedule the same number of conference games. Don't really care what the number is.
As an example, Alabama has beaten Duke, NM State, S.Carolina, Southern Miss, Ole Miss, Tx A&M, Tenn, Ark, and lost to LSU (so far).
Remember - this is nothing but numbers. Also, it is not intended to predict future performances. I created this ranking system as a way to measure teams who take risks (think Ohio State vs Oklahoma home and home a couple of years ago) against those who don't (scheduling Wofford or the Citadel in November). While I do have some pretty strong thoughts on the subject, opinions do not matter with the exception of how the algorithm is parsing data.
And no, this isn't just the BCS all over again. First, the BCS unfairly only included 2 teams in its considerations, and so it was biased for top Power 5 teams from the start. But worse yet. the BCS incorporated the human-biased AP/Harris polls and the ridiculously human-biased Coaches' poll. Want to know how much humans ruin this thing? Bama lost at home to a very good LSU and dropped 3 places in the coaches' poll, while Penn State lost on the road by about the same margin to a very good Minnesota and dropped 6 places. Now, you'll say that #3 Bama lost to #2 LSU vs #4 PSU losing to #17 Minn. But I had #3 LSU beating #11 Bama (sorry Tide fans) and #2 PSU losing to #8 Minn.
My rankings are based loosely on many states' high school playoff qualifiers: teams get 1st level points for beating opponents, and they get 2nd level points for the teams their opponents beat. Teams get more points for beating Power 5 schools (1 point) than Group of 5 schools (0.5 points). And as far as I know, my system is the only one that actively penalizes FBS teams for scheduling FCS teams (-1 point). My opinion is that it gives an unfair advantage to that team in that it's almost an automatic win with no effort while other teams are taking risks. This is the same issue Urban Meyer has currently latched onto: scheduling parity. I happen to agree with coach - the CFP committee needs to set some guidelines on scheduling, and they need to do it now. All P5 conferences should be required to schedule the same number of conference games. Don't really care what the number is.
As an example, Alabama has beaten Duke, NM State, S.Carolina, Southern Miss, Ole Miss, Tx A&M, Tenn, Ark, and lost to LSU (so far).
- 9 games
- 6 wins vs P5 = 6 points
- 2 wins vs G5 = 1 point
- 1 loss vs P5 = 0 points
- 6 points for P5 wins, 1 point for G5 win, 0 points for loss is 7 points divided by 9 games = 0.7778 1st level points.
- Add up all of the 1st level points of the teams that the Tide has beaten, which is 15
- Duke=2, NM State=0, S.Carolina=2.5, Southern Miss=2, Ole Miss=2, Tx A&M=3.5, Tenn=3, Ark=0
- Divide 15 by 100 minus the number of missing games (if Bama has played 9 say Duke has played 8, divide by 99) then multiply by 10 to get 1.5152 2nd level points.
- Add 1st and 2nd level points for a total of 2.2929 and 17th in my week 11 rankings.
Remember - this is nothing but numbers. Also, it is not intended to predict future performances. I created this ranking system as a way to measure teams who take risks (think Ohio State vs Oklahoma home and home a couple of years ago) against those who don't (scheduling Wofford or the Citadel in November). While I do have some pretty strong thoughts on the subject, opinions do not matter with the exception of how the algorithm is parsing data.
And no, this isn't just the BCS all over again. First, the BCS unfairly only included 2 teams in its considerations, and so it was biased for top Power 5 teams from the start. But worse yet. the BCS incorporated the human-biased AP/Harris polls and the ridiculously human-biased Coaches' poll. Want to know how much humans ruin this thing? Bama lost at home to a very good LSU and dropped 3 places in the coaches' poll, while Penn State lost on the road by about the same margin to a very good Minnesota and dropped 6 places. Now, you'll say that #3 Bama lost to #2 LSU vs #4 PSU losing to #17 Minn. But I had #3 LSU beating #11 Bama (sorry Tide fans) and #2 PSU losing to #8 Minn.
No comments:
Post a Comment