Interesting ranking system ...

[Velcro]

here is a ranking system that takes into consideration head to head matchups and tries to prevent a scenario where someone that lost to someone else is still ranked above them.

http://www.unf.edu/~jcoleman/minv.htm

probably not the most accurate, but interesting.

[Velcro]

that system ranks as the #1 system in fairness (assuming the ranking of ranking systems is fair )

http://rsport.netorn.ru/cf/rankrank.htm

[Velcro]

what is funny to me is that the BCS system gets ranked in the mid 20's for fairness   ;D ;D

[Kizarvexis]

that system ranks as the #1 system in fairness (assuming the ranking of ranking systems is fair )

http://rsport.netorn.ru/cf/rankrank.htm

You have to like a ranking system called Mathematical Monkey Rankings for NCAA Division I-A Football.

I did notice that the AP and USA Today polls are not in the fairness ranking. Or did I miss them?

Kizarvexis

[Velcro]

they probably had toooo poor a rating... sports writers suck at voting.

[Sitting_BULL]

It's funny, all the non biased polls (computer not human) put Texas ahead of USC.

[Kizarvexis]

It's funny, all the non biased polls (computer not human) put Texas ahead of USC.

Computer polls still have a bias because of the GIGO principle. Garbage In, Garbage Out describes the condition that the information that comes out of a program is only as good as the information that you put in. The computer polls use formulas to process what ever stats the poll maker selects to rate the teams. Those stats and the weight he gives them are the bias in the computer poll.

Kizarvexis

[Sitting_BULL]

Computer polls still have a bias because of the GIGO principle. Garbage In, Garbage Out describes the condition that the information that comes out of a program is only as good as the information that you put in. The computer polls use formulas to process what ever stats the poll maker selects to rate the teams. Those stats and the weight he gives them are the bias in the computer poll.

Kizarvexis

uh...hehu...uh.....hehi.........you said poll.

[Velcro]

But at least the GIGO factor is consistent, week to week... and the fact that a bunch of them are used to compile the computer ranking portion of the BCS probably averages out some of the error.  

No matter what they do with the human polls, history comes into play...and not even just recent history.

[USFMikeB]

Computer polls still have a bias because of the GIGO principle. Garbage In, Garbage Out describes the condition that the information that comes out of a program is only as good as the information that you put in. The computer polls use formulas to process what ever stats the poll maker selects to rate the teams. Those stats and the weight he gives them are the bias in the computer poll.

Kizarvexis

Absolutely correct.  However, at least the bias in a computer poll is discernible because the assumptions can be reviewed.  Additionally, the assumptions are applied consistently, rather than whimsically.

Human polls are intrinsically flawed, and should NOT be used for any level of reliability.  Why are they intrinsically flawed?  Because the voters do not have a consistent level of empirical evidence on all 119 teams in I-A football.  The coaches only watch their own games, and highlight tapes of their opponents, and often rely on assistants to select their rankings for them.  The media really only watch the big games, and have an intrinsic bias to create sensationalism.

I prefer Sagarin, as he plainly discusses what all teams are gauged on.