The conference computer ratings

[Blue Hen]

Just looking over some of the better known computer systems that offer conference strength comparrisons....and thinking about AUfan, my SEC buddy.

The SEC....as rated by :

James Howell - 3rd
Sagarin - 3rd
Wolfe - 2nd
Anderson - 3rd
Massey - 3rd
Colley - 2nd

The SEC's OOC SOS is pretty much a consensus dead last among all 11 FBS conferences. The B12 seems to be a consensus # 1 in strength right now.


Anyway, Aufan, you should contact these computer geeks and explain the superiority and dominance so that they can fix their systems.

[EvilVodka]

no doubt the SEC is having a less than sterling year...but they're still better as a whole than the PAC 10 and Big 10

anyone that truly knows football knows this is all cyclical...the Big XII and ACC conferences are pretty good this year...

the ACC is tricky, they don't have top teams, but they've got alot of good teams in the middle

What has hurt the SEC the most IMO has been the play of Tennessee and Auburn

[Blue Hen]

yeah, in the very very SEC friendly coaches poll , LSU was the 6th best team in the land before any FB was played in August. Auburn was the 11th best team and Alabama was unranked. Georgia was the very best among all CF teams ( #1 ranking ), Tennessee the 18th best CF team and Florida the 5th best.

Florida has lived up to its hype and Alabama has exceeded expectations but the other SEC brand names have been disappoinments with AU & UT being mega flops. No other league, year in and year out, gets inflated rankings quite like the SEC...imo , of course. USC seems to get unearned high pre season rankings also. Oh well.

I say the first , last, and only 'poll' ( if you have to have these things ) should appear in mid January.

[Hail to the Victors Valiant]

Just like in the fcs, some poll is needed to rank teams for a playoff. The fcs ranks its teams as well but for a different reason.

[aufan59]

Just looking over some of the better known computer systems that offer conference strength comparrisons....and thinking about AUfan, my SEC buddy.

The SEC....as rated by :

James Howell - 3rd
Sagarin - 3rd
Wolfe - 2nd
Anderson - 3rd
Massey - 3rd
Colley - 2nd

The SEC's OOC SOS is pretty much a consensus dead last among all 11 FBS conferences. The B12 seems to be a consensus # 1 in strength right now.


Anyway, Aufan, you should contact these computer geeks and explain the superiority and dominance so that they can fix their systems.

In those systems, teams are judged by their schedule instead of their strength. Texas Tech would be ranked higher than the New York Giants if the Giants had played Alabama's schedule, but a subjective measure of strength allows us to fix the computer polls' mistakes.

Yes, I agree, strength is subjective. However, the SEC will dominate bowl games yet again, including a 3rd National Championship in a row. This will be plenty of proof that the SEC is the best conference, strength-wise.

[JamesHowell]

> In those systems, teams are judged by their schedule instead of their strength.

Not at all true. They are judged by the schedule AND HOW THEY PERFORM AGAINST THAT SCHEDULE.

If you play a weaker team, you should, by definition, perform better than a team that plays a relatively tougher team. SOS does not determine where conferences or teams rate. It is how they perform against that SOS. If two teams have the same score, the team that played the tougher opponent is, by defintion, stronger. If however, one team dominates a team and another team squeaks by a slightly tougher team, the team that played the weaker opponent might rate out higher.

[ZOOMBAG]

In those systems, teams are judged by their schedule instead of their strength. Texas Tech would be ranked higher than the New York Giants if the Giants had played Alabama's schedule, but a subjective measure of strength allows us to fix the computer polls' mistakes.

Yes, I agree, strength is subjective. However, the SEC will dominate bowl games yet again, including a 3rd National Championship in a row. This will be plenty of proof that the SEC is the best conference, strength-wise.

Sagarin and a load of others have regularly squashed the misconception that Strength of Schedule is an overwhelming statistical factor, at least in a predictive system that attempts to determine who would beat who by how much. It matters, but just not as much as most college football fans think. Average margin of victory actually matters more. And any system that rates a one point loss to the #1 team at the #1 teams home higher then a one point win over the #119 at your home...is irreparably BROKEN. The weakest win possible is ALWAYS more valuable than the strongest loss possible.

[aufan59]

Sagarin and a load of others have regularly squashed the misconception that Strength of Schedule is an overwhelming statistical factor, at least in a predictive system that attempts to determine who would beat who by how much. It matters, but just not as much as most college football fans think. Average margin of victory actually matters more. And any system that rates a one point loss to the #1 team at the #1 teams home higher then a one point win over the #119 at your home...is irreparably BROKEN. The weakest win possible is ALWAYS more valuable than the strongest loss possible.

I had always thought that the formulas used by the BCS couldn't include MOV. Polls without MOV give every team a predetermined maximum value, as schedule is the only determining factor.

Just curious, but aren't the sample sizes too small to be accurate in predictions? There aren't that many BCS vs BCS games. I don't see how the SEC can really be compared to the Big 10 so because they've yet to play a game against each other.

[Blue Hen]

Fun only ...the polls in the NCAA championship divisions, although the selection committees might give them a look for 'at large' consideration.

[Blue Hen]

Glad you understand how those computer systems work, auFan, because I sure don't .

The SEC is 12-13 vs the B10 in PSEGs of the BCS era , all played in SEC backyards. I've been hearing for all 10 BCS seasons that the SEC would dominate, just as you said but , so far, the only dominance is in words. When the games are actually played there's no dominance. So, you're guaranteeing that this exhibition season will be different ? Maybe....we'll see.

Also, the SEC has posted exactly 4 winning bowl seasons in the last 10 post seasons (the BCS era). What kind of 'dominance' is that ?

[Jim]

I guess I better rank teams better in my polls since I set the standard !!! I need to get it done Saturday night or at the very latest, Sunday morning so these guys can take their next breath. I will try to be more diligent in my polling placements so as to keep the others on the straight and narrow so they can feed their familys. Thanks for the "heads up" guys. I appreciate it and I, if I may speak freely, know those around the nation do as well.

You'll be in my prayers.

[ZOOMBAG]

The sample by which Sagarin's models are derived from go back at least 50 years. I' sure our Jame's Howell's model is backed by several decades of data as well. It's pretty large.

His BCS rating is separate from his main rating. His BCS rating is "neutered" by not having any MOV calculations.


Big 12 is a pretty solid #1 this year....

[aufan59]

The sample by which Sagarin's models are derived from go back at least 50 years. I' sure our Jame's Howell's model is backed by several decades of data as well. It's pretty large.

His BCS rating is separate from his main rating. His BCS rating is "neutered" by not having any MOV calculations.


Big 12 is a pretty solid #1 this year....

At the end of the season it only uses data from this year.

In statistics, relevance of the results increases as sample size increases. This season there is a sample size of zero SEC and Big 10 teams playing each other.

The elo-chess system works much better when you have thousands of players that can distribute themselves along a bell curve after hundreds of games in consistent environments. It just seems silly to apply it to something like a football season.