[fullmetal]

#1 Alabama
#2 Oregon
#3 Ohio State
#4 TCU
#5 Oklahoma
#6 Florida
#7 Boise State
#8 LSU
#9 Virginia Tech
#10 Southern California
#11 Auburn
#12 Stanford
#20 Utah
#23 Oregon State
#27 California
#31 Arizona
#34 Arizona State
#52 Washington
#53 UCLA
#72 Colorado
#94 Washington State
#142 Harvard

Ratings linked here.

[fullmetal]

#1 Alabama
#2 Oregon
#3 Ohio State
#4 TCU
#5 Oklahoma
#6 Florida
#7 Boise State
#8 LSU
#9 Virginia Tech
#10 Southern California
#11 Auburn
#12 Stanford
#20 Utah
#23 Oregon State
#27 California
#31 Arizona
#34 Arizona State
#52 Washington
#53 UCLA
#72 Colorado
#94 Washington State
#142 Harvard

Ratings linked here.

[JeffInCorvallis]

Before any games have been played does this mean anything?  I mean, I know the preseason coaches' poll is not worth anything, either, but how does Sagarin's computer assign any numbers to any team with no data?  5-6 weeks into the season he gets much more interesting.

[JeffInCorvallis]

Before any games have been played does this mean anything?  I mean, I know the preseason coaches' poll is not worth anything, either, but how does Sagarin's computer assign any numbers to any team with no data?  5-6 weeks into the season he gets much more interesting.

[terry]

I think Auburn is returning 6 starters -- the fewest in the nation. How does that translate to a # 11 ranking?

[terry]

I think Auburn is returning 6 starters -- the fewest in the nation. How does that translate to a # 11 ranking?

[Griffins78]

how does Sagarin's computer assign any numbers to any team with no data?

That is a VERY good question. Doesn't make any sense. If there is historical data (from previous years) built into the model to give it an initial ranking then it undermines the credibility of the rankings when there is better data based on games played this year. If it incorporates data on returning players then it also undermines the credibility of the rankings when there is better data based on games played.

[Griffins78]

how does Sagarin's computer assign any numbers to any team with no data?

That is a VERY good question. Doesn't make any sense. If there is historical data (from previous years) built into the model to give it an initial ranking then it undermines the credibility of the rankings when there is better data based on games played this year. If it incorporates data on returning players then it also undermines the credibility of the rankings when there is better data based on games played.