Part One, today, part two later this week. Thanks to both KJ and Ty for talking things through on this.
Currently, I'm reading Dan Wetzel's marvelous book Death to the BCS. The mid-book review would be like that of a great documentary, it asks many great questions, points out the flaws in the current system flawlessly and offers some answers that are incomplete. This is a good thing, a book with all the answers would come across as dictatorial and anything that commands a billion dollars needs a democratic, conversational solution.
Wetzel et al(I've always said and pals in conversation, it's just more fun to read that way). Wetzel and pals offer the solution of a sixteen team playoff with the eleven conferences getting automagic bids into the playoff and the remaining five teams being drawn at large. They go on to state that all the games except the Championship Game would be played at the higher seed's home field. Which as a Michigan State fan(brr..), still I think if we were playing to be in the elite eight of college football against the best in CFB, my zeal for king and country would keep me warm.
Still, my mind immediately went to two things. First, how could you fairly decide which five teams get in at-large? Obviously human rankings are not reliable and computers while reliable still produce results that a human can look at and know are incorrect. Second, since the home field team would have a distinct advantage over the away team(Can you imagine Texas playing at Michigan State on January 3rd? That's simply not fair.) how can someone ensure that the teams are ranked fairly?
Were someone to draw up this list using the final BCS ranking on December 4th, 2011 they would look like this.
2011 Postseason Rankings - BCS
1.)LSU(SEC Champs)
2.)Alabama(At-Large)
3.)Oklahoma State(Big 12 Champs)
4.) Stanford(At-Large)
5.) Oregon(Pac-12 Champs)
6.) Arkansas(At-Large)
7.) Boise State(At-Large)
8.) Kansas State(At-Large)
9.) Wisconsin(Flopping Champs)
10.) Clemson(ACC Champs)
11.) TCU(Mountain West Champs)
12.) Houston(Conference-USA Champs)
13.) West Virginia(Big East Champs)
14.) N. Illinois(MAC Champs)
15.) Arkansas State(Sun Belt Champs)
16.) La Tech (WAC Champs)
* - Teams 14-16 were unranked, Northern Illinois received 36 votes in the final coaches poll, Arkansas State received 1 vote and La Tech received zero. God rest their souls.
The bracket then looks like this.
This isn't bad, we don't have to abide the annual Boise whining about how they don't get a shot to play for the national title game, after all they play two actual games and 10 tune-ups on their way to an undefeated season every year so it's totally fair that they should get a shot to be mentioned in the same breath as teams who play 8-10 real games and 2 tune up games.
Honestly, my biggest beef with these rankings is that they are in some part based on preseason rankings which are just crap. The rankings then get carried forward into the season and then there is a bias as the season goes on. I want to reward teams for playing and beating quality competition. So I decided to come up with a rating system. This system needed to calculate a score based on wins, losses, quality of opponent and margin of victory. Is this hypothetical ranking perfect? Absolutely not, but is it based on postseason, pre-bowl results when the maximum set of data has been gathered for a college football season? You betcha. Is it a rating system meant to be open for discussion, certainly.
I'm breaking this post into two parts because I'm actually looking to create two or three different kinds of ratings, since it's the offseason and it's time for hypothetical discussion. One common tenet amongst the various ratings will be though that the base Win-Loss total will be calculated as FBS wins - ((FBS losses * 1) + (FCS losses * 2)). Since this system could actually be used to create an actual set of rankings instead of just the college football playoff rankings, I do want to pay attention to things like the effect of an FCS loss.
Currently, I'm reading Dan Wetzel's marvelous book Death to the BCS. The mid-book review would be like that of a great documentary, it asks many great questions, points out the flaws in the current system flawlessly and offers some answers that are incomplete. This is a good thing, a book with all the answers would come across as dictatorial and anything that commands a billion dollars needs a democratic, conversational solution.
Wetzel et al(I've always said and pals in conversation, it's just more fun to read that way). Wetzel and pals offer the solution of a sixteen team playoff with the eleven conferences getting automagic bids into the playoff and the remaining five teams being drawn at large. They go on to state that all the games except the Championship Game would be played at the higher seed's home field. Which as a Michigan State fan(brr..), still I think if we were playing to be in the elite eight of college football against the best in CFB, my zeal for king and country would keep me warm.
 | |
| And beer, so very very much beer could keep me warm |
Still, my mind immediately went to two things. First, how could you fairly decide which five teams get in at-large? Obviously human rankings are not reliable and computers while reliable still produce results that a human can look at and know are incorrect. Second, since the home field team would have a distinct advantage over the away team(Can you imagine Texas playing at Michigan State on January 3rd? That's simply not fair.) how can someone ensure that the teams are ranked fairly?
Were someone to draw up this list using the final BCS ranking on December 4th, 2011 they would look like this.
2011 Postseason Rankings - BCS
1.)LSU(SEC Champs)
2.)Alabama(At-Large)
3.)Oklahoma State(Big 12 Champs)
4.) Stanford(At-Large)
5.) Oregon(Pac-12 Champs)
6.) Arkansas(At-Large)
7.) Boise State(At-Large)
8.) Kansas State(At-Large)
9.) Wisconsin(Flopping Champs)
10.) Clemson(ACC Champs)
11.) TCU(Mountain West Champs)
12.) Houston(Conference-USA Champs)
13.) West Virginia(Big East Champs)
14.) N. Illinois(MAC Champs)
15.) Arkansas State(Sun Belt Champs)
16.) La Tech (WAC Champs)
* - Teams 14-16 were unranked, Northern Illinois received 36 votes in the final coaches poll, Arkansas State received 1 vote and La Tech received zero. God rest their souls.
The bracket then looks like this.
 |
| Russell Wilson finally does get a trip to Manhattan! |
This isn't bad, we don't have to abide the annual Boise whining about how they don't get a shot to play for the national title game, after all they play two actual games and 10 tune-ups on their way to an undefeated season every year so it's totally fair that they should get a shot to be mentioned in the same breath as teams who play 8-10 real games and 2 tune up games.
Honestly, my biggest beef with these rankings is that they are in some part based on preseason rankings which are just crap. The rankings then get carried forward into the season and then there is a bias as the season goes on. I want to reward teams for playing and beating quality competition. So I decided to come up with a rating system. This system needed to calculate a score based on wins, losses, quality of opponent and margin of victory. Is this hypothetical ranking perfect? Absolutely not, but is it based on postseason, pre-bowl results when the maximum set of data has been gathered for a college football season? You betcha. Is it a rating system meant to be open for discussion, certainly.
I'm breaking this post into two parts because I'm actually looking to create two or three different kinds of ratings, since it's the offseason and it's time for hypothetical discussion. One common tenet amongst the various ratings will be though that the base Win-Loss total will be calculated as FBS wins - ((FBS losses * 1) + (FCS losses * 2)). Since this system could actually be used to create an actual set of rankings instead of just the college football playoff rankings, I do want to pay attention to things like the effect of an FCS loss.
No comments:
Post a Comment