10/27/08 Gridiron Power Index (GPI), James Madison Holds at No. 1

[CSN-info]

http://www.championshipsubdivisionnews.com/skins/andreas_01/img/GPI.JPG

The Gridiron Power Index (GPI), the index ranking for the NCAA Division I FCS and a top indicator of at-large playoff selection continues with James Madison in the top spot again this week. No other team has been ranked No. 1 this year in the GPI.

The Colonial Athletic Association, the largest league in the FCS has seven teams in the top 25; the Southern Conference has five; the Big Sky and Missouri Valley Football Conferences placed three each; the Great West Football and the Southland Conferences have two each; and the Big South Conference plus the Ivy and Patriot Leagues have one each.

10/27/2008 GPI Top 25
1. James Madison (1.00)
2. Montana (4.13)
3. Richmond (4.88)
4. Appalachian St (5.13)
5. Wofford (5.50)
6. Cal Poly (5.88)
7. Villanova (6.13)
8. Weber St (6.75)
9. New Hampshire (9.38)
10. Northern Iowa (11.63)
11. Elon (12.00)
12. Cent Arkansas (12.25)
13. William & Mary (12.63)
14. Massachusetts (13.38)
15. W Illinois (14.88)
16. S Illinois (15.38)
17. N Arizona (17.25)
18. Furman (18.63)
19. Maine (19.13)
20. Lafayette (21.25)
21. Harvard (22.75)
22. Liberty (23.13)
23T. UC Davis (23.25)
23T. Ga Southern (23.75)
25. McNeese St (24.88)

Full GPI Detail
http://www.collegesportingnews.com/stats/writer/GPI/20081027gpi.html

Conference Ranking:
Rank, League, Total Average
1. Colonial Athletic Association (22.16)
2. Southern Conference (29.29)
3. Big Sky Conference (30.45)
4. Great West Football Conference (31.40)
5. Southland Conference (33.57)
6. Missouri Valley Football Conference (36.78)
7. Ivy League (49.02)
8. Big South Conference (49.09)
9. Patriot League (49.61)
10. Ohio Valley Conference (50.92)
11. Mid-Eastern Athletic Conference (61.33)
12. Northeast Conference (66.50)
13. Southwestern Athletic Conference (69.98)
14. Pioneer Football League (75.43)
15. Independents (79.28)

[AZGrizFan]

Conference Ranking:
Rank, League, Total Average
1. Colonial Athletic Association (22.16)
2. Southern Conference (29.29)
3. Big Sky Conference (30.45)
4. Great West Football Conference (31.40)
5. Southland Conference (33.57)
6. Missouri Valley Football Conference (36.78)

Look out, Big Bad SoCon....somebody's sneakin' up on you!!! ;) ;) ;)

[Grizzaholic]

Conference Ranking:
Rank, League, Total Average
1. Colonial Athletic Association (22.16)
2. Southern Conference (29.29)
3. Big Sky Conference (30.45)
4. Great West Football Conference (31.40)
5. Southland Conference (33.57)
6. Missouri Valley Football Conference (36.78)

Look out, Big Bad SoCon....somebody's sneakin' up on you!!! ;) ;) ;)

And look who is down there.xwhistlex

[bluehenbillk]

xazzx

[GSUhooligan]

xazzx

Do you live vicariously through your conference when your football team sucks?

[bluehenbillk]

My post has zero to do with the CAA or UD.

[dbackjon]

Weber at number 8 - getting there - much better than the AGS poll!

[dbackjon]

And the computers average Weber at #3 - computers>>people!

[ToTheLeft]

How about the Big South moving up. :)

[URMite]

And the computers average Weber at #3 - computers>>people!

For some strange reason I like the combination of the two of them better!:p

[89Hen]

I never get tired of picking on the computers...

Sagarin
19. Samford (wins over West Georgia, Faulkner, WCU, El Cid)

Keeper
8. UMass (no wins over ranked teams)

Self
18. UC Davis (wins PSU, UNC, SUU, Iona, UND)

Laz
7. Elon (1-2 vs ranked teams)

[boonegoon]

How is Weber below Montana?

[danefan]

How is Weber below Montana?

Head-to-head is not a factor that computers can analyze.....

[KAUMASS]

I never get tired of picking on the computers...

Sagarin
19. Samford (wins over West Georgia, Faulkner, WCU, El Cid)

Keeper
8. UMass (no wins over ranked teams)

Self
18. UC Davis (wins PSU, UNC, SUU, Iona, UND)

Laz
7. Elon (1-2 vs ranked teams)


89Hen, here is some more fun for you-check out the average of 32 polls at this site-College football rankings-
http://www.mratings.com/cf/compare1aa.htm

For some more fun, click on "Nolan" on that site and check out his schedule of strength rankings for FCS this year. Good stuff.

[wideright82]

89Hen, here is some more fun for you-check out the average of 32 polls at this site-College football rankings-
http://www.mratings.com/cf/compare1aa.htm

I hope you are not using this give credibility to the GPI, and nolan is laughable. He is the hedge fund of the polls.

[ccd494]

Interestingly, the computers know that Maine is still in the Union. Poll voters apparently do not.

[dbackjon]

Interestingly, the computers know that Maine is still in the Union. Poll voters apparently do not.

Another reason that computers OVERALL are better. Yeah, you can cherry pick certain things that are off, but they show less bias than the voters, and don't care where/when games are played

[URMite]

Another reason that computers OVERALL are better. Yeah, you can cherry pick certain things that are off, but they show less bias than the voters, and don't care where/when games are played

You don't think there should be some bias in favor of away or recent games? Maybe not as much (or little) as is given, but none?

[dbackjon]

You don't think there should be some bias in favor of away or recent games? Maybe not as much (or little) as is given, but none?

I should clarify - I was referring to out west, etc. Not home or away, which of course there is, and should be a bias for.

[appfan2008]

I am well aware of all of the factors that play into why certain conferences dont play in the playoffs or dont get auto bids... ivy with their reasons, swac title game so on and so on big south, great west, nec, pfl, have their reasons but is anyone else think that it is sort of wrong that our playoffs dont incorporate the best conferences??? some of the conferences that have autobids are ranked significantly lower than ones with out:(

[KAUMASS]

I hope you are not using this give credibility to the GPI, and nolan is laughable. He is the hedge fund of the polls.


It's a different perspective and fun to look at, but no, I don't take it seriously.
I take the GPI top 25 seriously..26 to 50 can be a bit out of whack due to the fact that most are given 26 points a piece for the polls. SN, AGS and Coaches Poll do not vote after 25..If you have a #32 place team in the GPI, but SN or AGS has voted them #22, it can skew the numbers after averaging the 5 computer rankings. (7 total computer rankings, top and bottom number thrown out)
I think they should give each team 26 or just average the 5 computer rankings for teams above #26. Just my xtwocentsx. If someone can explain how it works better, I'm all ears.

[WileECoyote06]

I am well aware of all of the factors that play into why certain conferences dont play in the playoffs or dont get auto bids... ivy with their reasons, swac title game so on and so on big south, great west, nec, pfl, have their reasons but is anyone else think that it is sort of wrong that our playoffs dont incorporate the best conferences??? some of the conferences that have autobids are ranked significantly lower than ones with out:(

But some conferences may be top-heavy and bottom-loaded.

[wideright82]

It's a different perspective and fun to look at, but no, I don't take it seriously.
I take the GPI top 25 seriously..26 to 50 can be a bit out of whack due to the fact that most are given 26 points a piece for the polls. SN, AGS and Coaches Poll do not vote after 25..If you have a #32 place team in the GPI, but SN or AGS has voted them #22, it can skew the numbers after averaging the 5 computer rankings. (7 total computer rankings, top and bottom number thrown out)
I think they should give each team 26 or just average the 5 computer rankings for teams above #26. Just my xtwocentsx. If someone can explain how it works better, I'm all ears.

Actually, Kau, you explained how it works perfectly, and that is my BIGGEST problem with the GPI. The way it is calculated does not actually take into account any statistical relevance. I think rather than bitching about this like i have been doing i should take some initiative to create one that I think is better. I do find the other one interesting though as there actually is some statistical relevence to it, and probably is much "better" (statistically) than the GPI is. Not sure why it isn't used more though.

[KAUMASS]

Actually, Kau, you explained how it works perfectly, and that is my BIGGEST problem with the GPI. The way it is calculated does not actually take into account any statistical relevance. I think rather than bitching about this like i have been doing i should take some initiative to create one that I think is better. I do find the other one interesting though as there actually is some statistical relevence to it, and probably is much "better" (statistically) than the GPI is. Not sure why it isn't used more though.

Let me know when you come up it...xthumbsupx

[Syntax Error]

... out of whack due to the fact that most are given 26 points a piece for the polls...That was advised by Kenneth Massey. It is statistically correct.

[Syntax Error]

89Hen, here is some more fun for you-check out the average of 32 polls at this site-College football rankings-
http://www.mratings.com/cf/compare1aa.htm
For some more fun, click on "Nolan" on that site and check out his schedule of strength rankings for FCS this year. Good stuff.Massey calls that the dirty laundry list of all ratings and polls. The GPI has been #1 there since it began.

[KAUMASS]

That was advised by Kenneth Massey. It is statistically correct.


Hey Syntax,

The only question I would have is when a team like S. Carolina state is ranked 34th this week in the GPI. Instead of 26 apiece for the polls, they were ranked 23,22,20. If they had 26 apiece, they would have had a total score of 33 that would put them at #39 in the GPI.

Does Massey have an explanation or why this is advised? I can see if everybody over 25 had 26. Are they trying to reward a team like SCS and filter them up like a bonus multiplier for getting some love in the AGS, SN or Coaches poll?

Thanks...

[Syntax Error]

Hey Syntax,

The only question I would have is when a team like S. Carolina state is ranked 34th this week in the GPI. Instead of 26 apiece for the polls, they were ranked 23,22,20. If they had 26 apiece, they would have had a total score of 33 that would put them at #39 in the GPI.

Does Massey have an explanation or why this is advised? I can see if everybody over 25 had 26. Are they trying to reward a team like SCS and filter them up like a bonus multiplier for getting some love in the AGS, SN or Coaches poll?

Thanks...xconfusedx The GPI is a mix of different kinds of computer ratings and human polls. Every team not in the top 25 of the polls gets the same polls rating (26).

[Dukie95]

If a team is in the 26+ category, then the human polls have no affect on your score. They are only given credit for what the computer polls say.

An arbitrary number has to be assigned for those columns. If you give them all a zero, they'll be at the top, if you give them something higher, like, say 50, there will be too big a gap in the rankings.

Think of the GPI as a central calculation all the computer polls together, but the top 25 are further refined (and cross-checked) by calculating in the human polls. When it comes down to it, how much are we really concerned about the 3 or 4 points spread that contributes once you get above 26?

[KAUMASS]

xconfusedx The GPI is a mix of different kinds of computer ratings and human polls. Every team not in the top 25 of the polls gets the same polls rating (26).

Take a look at the GPI this week. South Carolina State is ranked 34th in the GPI. They received a 23,22, and 20 in the AGS, SN and coaches poll.
They were not given a 26 for all 3 polls, thus improving their GPI numbers by 5 spots. If they were given a 26 for the AGS, SN and coaches poll, they would be at #39 in the GPI this week.

Any thoughts? Thanksxthumbsupx

[KAUMASS]

If a team is in the 26+ category, then the human polls have no affect on your score. They are only given credit for what the computer polls say.

An arbitrary number has to be assigned for those columns. If you give them all a zero, they'll be at the top, if you give them something higher, like, say 50, there will be too big a gap in the rankings.

Think of the GPI as a central calculation all the computer polls together, but the top 25 are further refined (and cross-checked) by calculating in the human polls. When it comes down to it, how much are we really concerned about the 3 or 4 points spread that contributes once you get above 26?

Not to concerned, but just curios why everyone is not given a 26. It could bump up a team by 5 or 6 spots in the GPI.

[wideright82]

That was advised by Kenneth Massey. It is statistically correct.

You're statistically correctxcoffeex

[uofmman1122]

You're statistically correctxcoffeexhttp://i158.photobucket.com/albums/t109/Marty_Grant1122/InsultMaster.jpg

BUUUUUUURRRRRRRRRNNN!!

xlolxxlolxxlolx

[89Hen]

but they show less bias than the voters, and don't care where/when games are played
So does throwing darts at a board. xnonono2x

[89Hen]

The GPI is a mix of different kinds of computer ratings and human polls.
Only for 30 or so teams. After those, it is 100% computer ratings. xpeacex

[danefan]

Digging up a thread, I know, but does anyone know how come the Wolfe Rankings aren't used in the GPI?

Or is it and I just missed it?

http://prwolfe.bol.ucla.edu/cfootball/ratings.htm

[Syntax Error]

Digging up a thread, I know, but does anyone know how come the Wolfe Rankings aren't used in the GPI?

Or is it and I just missed it?

http://prwolfe.bol.ucla.edu/cfootball/ratings.htmWolfe doesn't start his ratings until much later I think.

[danefan]

Wolfe doesn't start his ratings until much later I think.

The reason I was wondering was because I was reading the FCS Handbook which requires the use of Wolfe for the modified GPI ranking. I figured it would have already been in the GPI.

I also like Wolfe because they rank every division, including NAIA, which gives a better chunk of data for FCS teams who play across a lot of divisions.

[Syntax Error]

The reason I was wondering was because I was reading the FCS Handbook which requires the use of Wolfe for the modified GPI ranking. I figured it would have already been in the GPI.
I also like Wolfe because they rank every division, including NAIA, which gives a better chunk of data for FCS teams who play across a lot of divisions.I think Sagarin is the only one in the GPI which doesn't do that.