NCAA Division 1A Football

Copyright 2004-2005, Paul Kislanko

Often people who are not familiar with the nature and limitations of statistical methods tend to expect too much of the rating system. Ratings provide merely a comparison of performances, no more and no less. The measurement of the performance of an individual is always made relative to the performance of his competitors and both the performance of the player and of his opponents are subject to much the same random fluctuations. The measurement of the rating of an individual might well be compared with the measurement of the position of a cork bobbing up and down on the surface of agitated water with a yard stick tied to a rope and which is swaying in the wind.
Arpad Elo in Chess Life, 1962
In theory, there is no difference between theory and practice; In practice, there is.
Chuck Reid
Commentary
5 Aug 2005 Schedule Strength Including Future Opponents
Describes a method for characterizing a schedule based upon opponents rankings by any measure.
11 Aug 2005 Distance Matrix for 2005
Looks beyond opponents to OOs, OOOs, and OOOOs.
24 Aug 2005 The post-2005 Bowl Mess
Are there "too many bowls" or are we just not managing them right? 28 Sep Update
Dec 30 Update.
18 Sep 2005 College Football Divisional Alignments
Are there really just 4 divisions, or are there five? 29 Sep Update. 5 Dec 2005 Updated Scoring Summary
29 Nov Interdivisional scoring summary.
6 Oct 2005 The Twiddling Continues
It turns out the Harris folks have a feature to fill the knee-jerk reaction requirement for the 2004-05 offseason.
12 Oct 2005 Transparency in the BCS Polls
They don't have to disclose individual ballots to get the process right.
3 Nov 2005 A Simpler, Better Formula?
This is the first of two articles suggesting a simple way to avoid the most serious problems with the BCS formula.
9 Nov 2005 Counting Rankings
The conclusion of last week's suggestion, with a sample from before and after the games of November 5th.
10 Nov 2005 Is there a Home Field Advantage?
The scoring stats suggest that for the most part, there's not.
15 Nov 2005 Why they play the games
A look at the teams and games that are the hardest to explain in the context of the 2005 season.
21 Nov 2005 Automatically filling out a ballot
This is actually about how to turn the ISOV ranking into a list ordered by the probability that a higher-ranked team will beat a lower-ranked team. The ISOV order does not, by itself, have this property.
27 Nov 2005 Normalizing Scoring Statistics
A good offense can make a good defense look bad, and conversely. A bad defense can make a bad offense look good, and so on. But just as you can find an SOS related to winning percentage, you can find one related to scoring, and use that to normalize team statistics.
5 Dec 2005 BCS v non-BCS scheduling myths
Repeating a common misconception often enough only makes it more common, it doesn't make it true. Facts do not support some of the claims made about scheduling.
References
Standings
Includes links to consolidated conference and team Schedules & Results Where a ranking is indicated for a team, it is the better of the AP and Coaches' Poll rankings. Games are highlighed if the sum of the team rankings is 50 or less (top 25s, top 10 vs top 40, etc).
Interconference Records
The individual conference pairings are listed with the average relative strength of the opponents.
Scoring Stats
Scoring offense normalized by SOS(opponents' scoring defense) and scoring defense normalized by SOS(opponents' scoring offense). In the report mPF is measured Points For, mPA, measured Points Allowed, and PF/PA are the normalized values. mPF and PA will not match the NCAA stats, because the data only includes D1A opponents.
Common Opponents
Lists all team pairs that have a common opponent, indexed by team name. Indicates number of wins by each member of the team-pair against the common opponents, and links to schedules for each team.
College Football Rankings comparison
An invaluable resource compiled by Kenneth Massey.
Summary by majority consensus All rankings; By number of ranks
Ratings
Polls Current
Current Final 
Week 14 TWP(ISOV)
Week 13 TWP(ISOV)
Week 12 TWP(ISOV)
Week 11
Week 10
Week 9
Week 8
Week 7
Week 6
Week 5
Week 4
Week 3The field is not really connected enough for the full PA- algorithms to be applied. For week 3 what is being reported as PA(rating) is really the ASOS(rating). This will become the value of a win over the team when the full version is available.

For more information, see About the Ratings.

2004 Home Page