2009 Gridiron Power Index (GPI) Debuts With Richmond as No. 1

[CSN-info]

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

2009 Gridiron Power Index (GPI) Debuts With Richmond as No. 1
College Sporting News

UPDATED!!!!!!

Read More ... (http://www.championshipsubdivisionnews.com/index.php/2009/10/06/2009-gridiron-power-index-gpi-debuts-wit-1?blog=5)

[TexasTerror]

GPI confirms that the CAA is the best league. The Great West at second? Does it help they have fewer teams and far fewer chances at a weak link? The SWAC, despite their OOC gains - still the worst full scholarship league.

Conference Ranking:
Rank, League, Total Average
1. Colonial Athletic Association (22.25)
2. Great West Conference (25.20)
3. Missouri Valley Football Conference (29.54)
4. Big Sky Conference (29.98)
5. Southern Conference (30.53)
6. Southland Conference (34.28)
7. Ohio Valley Conference (43.27)
8. Big South Conference (44.16)
9. Ivy League (45.08)
10. Patriot League (46.75)
11. Mid-Eastern Athletic Conference (49.96)
12. Northeast Conference (57.68)
13. Southwestern Athletic Conference (57.76)
14. Pioneer Football League (67.80)
15. Independents (71.85)

[89Hen]

And so begins another year of picking apart the computers.

Self
7. Northern Arizona


Wait, I really don't need to go any further, do I? xlolx

[PantherRob82]

And so begins another year of picking apart the computers.

Self
7. Northern Arizona


Wait, I really don't need to go any further, do I? xlolx

I agree computer polls are worthless, but what are your arguements on the total GPI?

[catbob]

And so begins another year of picking apart the computers.

Self
7. Northern Arizona


Wait, I really don't need to go any further, do I? xlolx

Are you saying you have NAU at #7? The GPI shows New Hampshire.

[CSN-info]

Are you saying you have NAU at #7? The GPI shows New Hampshire.

The GPI averages the computer ratings, eliminating the top and bottom rating. NAU for example took #4 Montana to overtime.

[catbob]

Ah I gotcha, makes sense now.

[Go...gate]

Colgate in there but not Holy Cross? xconfusedxxconfusedxxconfusedx

[CSN-info]

Computer ratings get better as more games are played.

A modified GPI is used by the Division I Football Championship selection committee. How well did it indicate at-large team selection in 2008?

The committee chose these eight teams:
Villanova
Richmond
Montana
New Hampshire
Wofford
Cal Poly
Northern Iowa
Maine

This is what the modified GPI indicated:
2. Villanova (2.60)
4. Richmond (4.00)
5. Montana (4.80)
6. New Hampshire (6.60)
7. Wofford (6.80)
8. Cal Poly (8.20)
10. William & Mary (10.60)
11. Northern Iowa (11.00)

The committee selected: 12. Maine (12.00) over 10. William & Mary (10.60).

The regular GPI indicated:
3. Montana (4.50)
4. Villanova (4.63)
5. Richmond (4.75)
6. Cal Poly (5.25)
7. Wofford (7.75)
8. Northern Iowa (8.13)
9. New Hampshire (8.25)
12. William & Mary (13.00)

Again the committee selected: 14. Maine (15.00) over 12. William & Mary (13.00).

So in both cases the two versions of the GPI indicated all but one at-large selection.

[89Hen]

I agree computer polls are worthless, but what are your arguements on the total GPI?
garbage in > garbage out

[89Hen]

Are you saying you have NAU at #7? The GPI shows New Hampshire.
"Self" is one of the computers.

[89Hen]

Computer ratings get better as more games are played.
Like last year when Sauceda and Self had a 6-5 EWU at 14 and 15 respectively? xlolx

[panther25]

The GPI averages the computer ratings, eliminating the top and bottom rating. NAU for example took #4 Montana to overtime.



It should probably look at eliminating statistical outliers instead of just the top and bottom ranking... you know.... to be worth the server space it takes up.

[ToTheLeft]

Where are all the MEAC fans now, chirping about how the Patriot and Big South are such weaker conferences and weaker schedules... I guess the facts will send a message in this case. xpeacex

[GannonFan]

So in both cases the two versions of the GPI indicated all but one at-large selection.

And so did 99% of the knowledgeable posters on this site. All this work, all these mechanations, and the GPI can't pick the at large teams any better than Joe "Average" Poster can. And there have been several years where the GPI, due to the many computer "oddities" that 89Hen likes to point out (oddities that even exist at the end of the regular season, since let's be honest, computer programs don't do a terribly good job of ranking teams with such a small and such a non-interlaced sample size) has not predicted the 8 at large teams any better than a typical poster on this site.

Gotta hand it to the GPI - it's been mangling the ranking process year after year. xnonox

[KAUMASS]

Why do we go through this year after year?....If you don't like the channel, don't turn it on. The GPI is a hybrid computer-human poll. It gives a pretty good predicter of who will make the playoff field towards the end of the year. It's not perfect, but I don't think any system is perfect. There are a large number of posters here that could hand pick the 8 at large teams at the end of the year, but it doesn't work that way. No matter how you cut it or what system is devised or used by the committee, there are always going to be a few teams left out that are deserving. Heck, when the playoff field goes to 20, there will still be teams ranked anywhere from 18-30th that are bitchin' they didn't get in.

[panther25]

Why do we go through this year after year?....If you don't like the channel, don't turn it on. The GPI is a hybrid computer-human poll. It gives a pretty good predicter of who will make the playoff field towards the end of the year. It's not perfect, but I don't think any system is perfect. There are a large number of posters here that could hand pick the 8 at large teams at the end of the year, but it doesn't work that way. No matter how you cut it or what system is devised or used by the committee, there are always going to be a few teams left out that are deserving. Heck, when the playoff field goes to 20, there will still be teams ranked anywhere from 18-30th that are bitchin' they didn't get in.

So if someone doesn't like an article about who the best, say, running back in the nation is, we who disagree should sit back? No. People discuss it. I for one would like to see a combo poll of computers and humans (although pointless) that at the very least does something that follows the basic sampling rules of statistics, but since there is no place for that in here.... i will goxcoffeex

[bluehenbillk]

Oh yea! The GPI is back, and it even has a sponsor this year, I see where all computer polls are being monitored by the TD Bank online banking system, that should help with quality control.

Always is a confidence spike to come out in the annual debut to admit that computer rankings, especially in the early part of the year, are out of whack.

Why is it every year when this comedy gets published do I get an image in my head of Samir from Office Space taking a baseball bat to the copier or fax or whatever it was. Someone more clever than me can post that pic in this thread, it'd be appropriate.

[KAUMASS]

So if someone doesn't like an article about who the best, say, running back in the nation is, we who disagree should sit back? No. People discuss it. I for one would like to see a combo poll of computers and humans (although pointless) that at the very least does something that follows the basic sampling rules of statistics, but since there is no place for that in here.... i will goxcoffeex


That's what the GPI is..

Panther25, I see you are a newbie to the board...Welcome. The GPI has been discussed here to death in previous years...

[KAUMASS]

Where is Syntax Error?

[GannonFan]

And so begins another year of picking apart the computers.

Self
7. Northern Arizona


Wait, I really don't need to go any further, do I? xlolx

Aw man, don't knock the Self ranking - heck, he's got Delaware at #10 - yup, a full 11 spots better than an undefeated New Hampshire team. You go Self!!!! xlolxxlolxxlolx

[panther25]

That's what the GPI is..

Panther25, I see you are a newbie to the board...Welcome. The GPI has been discussed here to death in previous years...

Thanks, but I joined well into the GPI talk so I'm familiar with the banter! well actually that is my point. The GPI doesn't follow those rules. Nowhere in statistics does it say to knock off the top and bottom sample number for validity. Not to mention, how are the polls/computer rankings used. No doubt it isn't all of the available ones in the nation, so where does the random sample and sample size come in to play to make sure it is an adequate representation of that population?


Jus sayin. xcoffeex

[smallcollegefbfan]

Where is Syntax Error?

Believe he stopped posting a while back when he got banned over something. Just a guess though. You might want to ask him. I think SE is also CSN-Info though but I am not sure. No way to prove it either way so I am not accusing him but just guessing that he is going into CSN-Info mode and staying away from his other screen name. Although, there could be someone else who is CSN-Info so I am not really sure.

[CSN-info]

Thanks, but I joined well into the GPI talk so I'm familiar with the banter! well actually that is my point. The GPI doesn't follow those rules. Nowhere in statistics does it say to knock off the top and bottom sample number for validity. Not to mention, how are the polls/computer rankings used. No doubt it isn't all of the available ones in the nation, so where does the random sample and sample size come in to play to make sure it is an adequate representation of that population?

The GPI follows all statistical rules as set by the GPI advisor, Kenneth Massey. If you have questions about computer ratings and how they are used in the GPI then please direct those to him at http://www.masseyratings.com

Keep the disinformation out of the discussion.

[danefan]

The GPI follows all statistical rules as set by the GPI advisor, Kenneth Massey. If you have questions about computer ratings and how they are used in the GPI then please direct those to him at http://www.masseyratings.com

Keep the disinformation out of the discussion.

Disinformation? Seems like the guy was pointing out that the GPI's statistical rules aren't consistent with accepted statistical standards?

How is that disinformation?

Is Massey now the end-all-be-all of statistics?

[panther25]

The GPI follows all statistical rules as set by the GPI advisor, Kenneth Massey. If you have questions about computer ratings and how they are used in the GPI then please direct those to him at http://www.masseyratings.com

Keep the disinformation out of the discussion.

what is disinformation. And how does one person set statistical rules? don't you think that should be, i don't know, universal.




Panter25's Rules of Mathematics:

1. 2+2=39
2. 1 is an even number
3. 5 is equal to the letter 'Q' if inside an absolute value bracket
4. the number 9 is really an exponent in any equation



disinformation... you silly goose xlolxxlolx

[danefan]

Man - somebody is touchy to criticism. Posts are beind deleted already? To a guy who has only been here for 50 posts? After all these years, you'd think Ralph would get over the fact that people disagree with the premise of the GPI.

Its nothing personal. It just doesn't add up.

Ohhhhh......I know you like the pun!

[GannonFan]

The GPI follows all statistical rules as set by the GPI advisor, Kenneth Massey. If you have questions about computer ratings and how they are used in the GPI then please direct those to him at http://www.masseyratings.com

Keep the disinformation out of the discussion.


Disinformation? Seems like the guy was pointing out that the GPI's statistical rules aren't consistent with accepted statistical standards?

How is that disinformation?

Is Massey now the end-all-be-all of statistics?

I agree - come on, there's plenty of arbitrary statistical rules to follow in such a system as rating 120 or so teams that play 11 games total. And regardless of what you follow, it's not necessarily any better than a different interpretation. Thus the problem with small sample sizes and the lack of integrated schedules. Even Massey on his own web site talks about the inherent difficulty in dealing with teams that are themselves variables. There's no end-all be-all statistical model to do this and everything else we have is open to appropriate criticism. xthumbsupx

[OhioHen]

Where is Syntax Error?

MIA as a poster since June 30.

Any guesses as to what current screen name (if any) has replaced SE?

[danefan]

And if CSN-Info is going to be the spokesman for the GPI, CSN-Info needs to take the criticism that comes along with it.

Kenneth Massey has his own ratings system. If we were discussing his ratings we'd talk to him. We're not. We're discussing the GPI, which is a product put forth by CSN.

[89Hen]

The GPI follows all statistical rules as set by the GPI advisor, Kenneth Massey. If you have questions about computer ratings and how they are used in the GPI then please direct those to him at http://www.masseyratings.com

Keep the disinformation out of the discussion.
xlolx I've conversed via e-mail with Ken Massey. He is very arrogant and heaven forbid you criticize computer modeling.

[Lehigh Football Nation]

Colgate in there but not Holy Cross? xconfusedxxconfusedxxconfusedx

What's real interesting is that the PL is ranked above the MEAC even though their two top teams are in the Top 25 GPI but the PL's are not - and that's with Lehigh and Georgetown dragging down the GPI of the conference with zero wins. That means the bottom of the MEAC - incredibly - must be worse than the bottom of the PL by quite a significant margin.

Another interesting point is how the NEC and SWAC are almost neck-in-neck at the bottom of the GPI. While the Big South is well above the PL. Maybe Liberty or Gardner-Webb needs to be more seriously talked about when it comes to at-larges...

[panther25]

xlolx I've conversed via e-mail with Ken Massey. He is very arrogant and heaven forbid you criticize computer modeling.

Heaven forbid is right!!!!



What is funny is he created a computer model out of other computer models/polls thus creating another computer model different from the computer model that he already made using a formula for some other computer model..... which is then made to seem as if there is statistical relevence to it........


er........ I mean... the GPI is the best thing in the world. xthumbsupx

[DSUrocks07]

Where are all the MEAC fans now, chirping about how the Patriot and Big South are such weaker conferences and weaker schedules... I guess the facts will send a message in this case. xpeacex

xlolx

[GannonFan]

What's real interesting is that the PL is ranked above the MEAC even though their two top teams are in the Top 25 GPI but the PL's are not - and that's with Lehigh and Georgetown dragging down the GPI of the conference with zero wins. That means the bottom of the MEAC - incredibly - must be worse than the bottom of the PL by quite a significant margin.

Another interesting point is how the NEC and SWAC are almost neck-in-neck at the bottom of the GPI. While the Big South is well above the PL. Maybe Liberty or Gardner-Webb needs to be more seriously talked about when it comes to at-larges...

Well, that would be the case if the NEC, SWAC, or PL were real threats to capture at large bids. There'll be plenty of teams from conferences rated much higher than any of those that will be more likely to secure at larges. xthumbsupx

[Lehigh Football Nation]

*sigh* Not sure if I've mentioned this before or not, but no computer ratings system is perfect. They are based on formulas, and football game outcomes are based on a whole lot of different variables that have nothing to do with formulas. Nobody who has a formula in this game feels any differently about this truism.

However,... when you aggregate different formulas you can get a more nuanced picture, and the more data points you get the better the picture is. Better yet, mix formulas with human polls (such as Sports Network, and AGS) in order to get human opinion factored in as well. The result? The GPI. No other FCS poll factors in both statistical formulas and moderated human polls in this hybrid way - making the most "balanced" of any of the polls out there.

[DSUrocks07]

What's real interesting is that the PL is ranked above the MEAC even though their two top teams are in the Top 25 GPI but the PL's are not - and that's with Lehigh and Georgetown dragging down the GPI of the conference with zero wins. That means the bottom of the MEAC - incredibly - must be worse than the bottom of the PL by quite a significant margin.

Another interesting point is how the NEC and SWAC are almost neck-in-neck at the bottom of the GPI. While the Big South is well above the PL. Maybe Liberty or Gardner-Webb needs to be more seriously talked about when it comes to at-larges...

Well....one COULD say that Bethune-Cookman and Howard are dragging down the MEAC's GPI as well. With Howard having only wins over WSSU and Georgetown (both winless), and Bethune-Cookman with zero wins as well (Including a bad loss to D-II Shaw)

...but I'll let you continue with your campaign for a BSC at large bid xcoolx

[89Hen]

*sigh* Not sure if I've mentioned this before or not, but no computer ratings system is perfect....

Better yet, mix formulas with human polls...
One, you will never get a computer guy to admit that and two, it's been shown that the GPI is actually less accurate than the AGS Poll on it's own. xpeacex

[smallcollegefbfan]

One, you will never get a computer guy to admit that and two, it's been shown that the GPI is actually less accurate than the AGS Poll on it's own. xpeacex

The more information we can get the better. While I don't fully agree with what the GPI or the polls are saying right now I know that things will play themselves out, like they always do.

I thought the GPI was pretty accurate at the end of the playoffs after seeing everyone play. While fairly accurate there are human elements to playoff selection that we take into account when guessing on the field of 16 that the computers don't because computers use stats and numbers only and don't take human error into account.

[FCS_pwns_FBS]

Great West the second best conference in the FCS? xlolx

The only reason the Great West went up is that
1)Cal Poly beat SDSU with SDSU's top QB out...
2)UC Davis played a reasonably close game with a way overrated Boise State team
and the GPI of these two is 40% the average GPI of that conference.


Just once I'd like one of these geeks to release their methods/forumlas for computing GPI. I bet you I could pick them apart.

[Grizzaholic]

After reading this thread off and on today, and noticing posts vanish while others remain, I have but one question.

This GPI thing, and yes it is a THING, is not all that reasonable nor is it very good. Yet every year when it comes out and gets beaten to death AGS loses members. Why you ask? Because they disagree with the GPI and then get banned for disagreeing.

Come on AGS BOD, if people disagree with it, let it be. PLEASE?!

[Grizzaholic]

One, you will never get a computer guy to admit that and two, it's been shown that the GPI is actually less accurate than the AGS Poll on it's own. xpeacex

xthumbsupx

[WMTribe90]

I'm one that finds the GPI as useful tool. Is the GPI perfect, no, but nothing is. It can point out glaring biases or ommission in human polls, which can lack some objectivity of be short-sighted (rely to much or recent results as opposed to an entire season). For instance, I agree with the GPI's ranking of JSU and SFA. For some reason the human polls ahve been slow to move them up. Probably becuase they haven't historically been strong programs.

[Grizzaholic]

I've seriously about had it with the censorship on this board, and I'm one that finds the GPI as useful tool. Is the GPI perfect, no, but nothing is. It can point out glaring biases or ommission in human polls, which can lack some objectivity of be short-sighted (rely to much or recent results as opposed to an entire season). For instance, I agree with the GPI's ranking of JSU. For some reason the human polls ahve been slow tomove them up.

xthumbsupx

[UAalum72]

The more information we can get the better.
But this isn't more 'information', it's multiple interpretations of the same facts. That may be 'better' only if you assume they cancel each others' weaknesses; not, if you think they also cancel each others' strengths and result in a regression to a mean of computer and human rankings. That mean MUST be worse than whichever you think is a 'better' method.

Not only that, multiple human polls are only better if the biases are randomly distributed. In a relatively small world like FCS coaches and writers, who presumably have some contact with each other, biases may reinforce themselves rather than be reduced.

[MplsBison]

Screw any type of rankings and selection committees.

The top 2 teams in each of the 8 AQ bids go to the 16 team playoff, and that's that.

[danefan]

But this isn't more 'information', it's multiple interpretations of the same facts. That may be 'better' only if you assume they cancel each others' weaknesses; not, if you think they also cancel each others' strengths and result in a regression to a mean of computer and human rankings. That mean MUST be worse than whichever you think is a 'better' method.

Not only that, multiple human polls are only better if the biases are randomly distributed. In a relatively small world like FCS coaches and writers, who presumably have some contact with each other, biases may reinforce themselves rather than be reduced.

Especially when, as in this case, there is overlap in voters.

[R.A.]

What's real interesting is that the PL is ranked above the MEAC even though their two top teams are in the Top 25 GPI but the PL's are not - and that's with Lehigh and Georgetown dragging down the GPI of the conference with zero wins. That means the bottom of the MEAC - incredibly - must be worse than the bottom of the PL by quite a significant margin.



So we know that this criterion is flawed, because the bottom of the MEAC, Howard, defeated the bottom of the Patriot League, Georgetown on Georgetown's campus.

So if the bottom of the Patriot League according to the GPI is better than the bottom of the MEAC... why are they losing to us then??

The system is obviously flawed.

[R.A.]

Where are all the MEAC fans now, chirping about how the Patriot and Big South are such weaker conferences and weaker schedules... I guess the facts will send a message in this case. xpeacex

Holla at us at the end of the season.

[JackFan]

And so begins another year of CHERRY-picking apart the computers.

Self
7. Northern Arizona


Wait, I really don't need to go any further, do I? xlolx

I dont know what you have against the jacks, we were a play here and there away from beating Montana. Any time you talk big sky you exclude us I think the only way you will be happy is if your hens are ranked 1. keep your east coast bias to yourself.

[UAalum72]

Screw any type of rankings and selection committees.

The top 2 teams in each of the 8 AQ bids go to the 16 team playoff, and that's that.
Too late this year, but I'll back that next year after the Big South and NEC get autobids (sarcasm)

[MplsBison]

Too late this year, but I'll back that next year after the Big South and NEC get autobids (sarcasm)

That's fine.

Give AQ to the PFL and SWAC too (if they agree to no more than one sub-DI game a year).


Top 24 teams in FCS = top 2 from each AQ conference. Win your conference or shut up.

[89Hen]

I thought the GPI was pretty accurate at the end of the playoffs after seeing everyone play.
EVERYONE NEEDS TO READ THIS

You need to look at the end of the regular season GPI and compare it. Waiting until after the playoffs is not valid because the computers will assign the teams that get further a higher SOS value and therefore is really a backwards looking system. Human polls will do that too.

Here are the teams with some the largest deltas from AGS to ARC (average rank computer) for the end of the regular season last year. You tell me which you think is a more accurate assessment...

UNI = 4 AGS, 11 ARC (semifinalist, lost by a point)
Maine = 18 AGS, 12 ARC (first round losers, last team in field)
UMass = 25 AGS, 13 ARC (7-5)
SCSt = 13 AGS, 23 ARC (gave App all they could handle)
EWU = NR AGS, 17 ARC (6-5)
JSU = 21 AGS, 35 ARC (8-2 vs I-AA)

Honestly, who do you think is more accurate, the AGS or the sum of the computers?

[89Hen]

I dont know what you have against the jacks, we were a play here and there away from beating Montana.
xconfusedx Sorry you feel slandered, but I don't think ANYONE in their right mind would say NAU is a #7 team. Your #7 was the most obvious standout as a joke. Right behind that was Delaware at #10 in that same computer model if it makes you feel any better. xpeacex

[89Hen]

Screw any type of rankings and selection committees.

The top 2 teams in each of the 8 AQ bids go to the 16 team playoff, and that's that.
xlolx xlolx xlolx xlolx xlolx xcoffeex

[JackFan]

xconfusedx Sorry you feel slandered, but I don't think ANYONE in their right mind would say NAU is a #7 team. Your #7 was the most obvious standout as a joke. Right behind that was Delaware at #10 in that same computer model if it makes you feel any better. xpeacex


Not really upset that you feel we are not # 7 material, theres no way I would put us there either. The thing that im gonna defend is the fact that we are not a joke. Just ask the guys from Missoula how much of a joke our guys are. We are dealing with a lot of injuries on the defensive side of the ball and if not for that I would be a lot more confident in calling our guys a definite 20-25 ranked team which would of course be more appropriate. The funny thing is these things really don’t matter as our system sorts it self out rather then the BCS situation. 89HEN im sure your guys are really good too, as you guys always are I wish you guys the best of luck.

[89Hen]

The thing that im gonna defend is the fact that we are not a joke.
At #7 you are... the same way Delaware or Furman or Colgate or any other 20-40 team would be. The computer is simply wrong on that one.

[OL FU]

At #7 you are... the same way Delaware or Furman or Colgate or any other 20-40 team would be. The computer is simply wrong on that one.

Leave us out of this conversation:p:D

I think Sagarin had us a number 10 last weekxeyebrowx

[89Hen]

Leave us out of this conversation:p:D
If you're left out of this conversation, you'd be conversationless. :p

[OL FU]

If you're left out of this conversation, you'd be conversationless. :p

I guess that's better than being "speechless":)

[darell1976]

I will pay attention to this in 2012.

[GannonFan]

I will pay attention to this in 2012.

Trust me, it won't be any better or any more useful then - certainly won't be worth the wait. xlolxxthumbsupx

[coover]

Computer ratings get better as more games are played.

A modified GPI is used by the Division I Football Championship selection committee. How well did it indicate at-large team selection in 2008?

The committee chose these eight teams:
Villanova
Richmond
Montana
New Hampshire
Wofford
Cal Poly
Northern Iowa
Maine

This is what the modified GPI indicated:
2. Villanova (2.60)
4. Richmond (4.00)
5. Montana (4.80)
6. New Hampshire (6.60)
7. Wofford (6.80)
8. Cal Poly (8.20)
10. William & Mary (10.60)
11. Northern Iowa (11.00)

The committee selected: 12. Maine (12.00) over 10. William & Mary (10.60).

The regular GPI indicated:
3. Montana (4.50)
4. Villanova (4.63)
5. Richmond (4.75)
6. Cal Poly (5.25)
7. Wofford (7.75)
8. Northern Iowa (8.13)
9. New Hampshire (8.25)
12. William & Mary (13.00)

Again the committee selected: 14. Maine (15.00) over 12. William & Mary (13.00).

So in both cases the two versions of the GPI indicated all but one at-large selection.

Interesting. Let's see how it would work this year supposing ...

1. Cal Poly loses a close game this week to favored Montana in Missoula and

2. Cal Poly wins out the rest of the year (tough, but quite possible - I know some Weber State fans who might argue against that).

In that case, Poly would win the Great West and probably end with a fairly high (should I say low?) GPI, but their record would be 8 - 3, with one win over a DII team, and two losses to FBS teams, giving them only 7 DI wins. Will the commitee give them an at-large berth?

Is it their fault that they find it difficult to find DI games out of conference? They only scheduled the DII team because they could not find a DI team willing to play them.

[kalm]

Like last year when Sauceda and Self had a 6-5 EWU at 14 and 15 respectively? xlolx

Eastern had one bad loss (PSU). It's other losses were to #2 Montana, 6-5 Sac St., Colorado, and Texas Tech.

Eastern had wins against MSU, Weber, and NAU.

I'm not sure where Eastern ended up in the human polls, but the strength of schedule had to be fairly high up and we finished on a three game win streak.

Objectively looking at it, 14th isn't really that far off.

[SumItUp]

In that case, Poly would win the Great West and probably end with a fairly high (should I say low?) GPI, but their record would be 8 - 3, with one win over a DII team, and two losses to FBS teams, giving them only 7 DI wins. Will the commitee give them an at-large berth?

Is it their fault that they find it difficult to find DI games out of conference? They only scheduled the DII team because they could not find a DI team willing to play them.

Can Cal Poly get an at-large birth with a 7-3 D1 record? I think it's possible, but not likely given the likelihood of a number of bubble teams with 8 D1 wins.

Is it their fault they find it difficult to find D1 games out of conference? You won't find much sympathy from most on this board.

I also enjoy playing the "What if" game, but things will be sorted out as the games are played this year. It's a guarantee that 2-3 teams will be on the outside looking in and feeling slighted at the end of the season.

[JohnStOnge]

I continue to think that those who say the better "computer" rankings are worthless haven't objectively evaluated their accuracy in predicting outcomes. I think Hen 89 has compared their accuracy to some kind of thing where people pick on some kind of Delaware website but what he's talking about is people picking selected games and comparing that to the accuracy rate of models predicting all games.

[ASU_Fanatic]

App State at 15, wtf!!!??? We better than all the other teams in front of us....I hope.

[BDKJMU]

I agree computer polls are worthless, but what are your arguements on the total GPI?

Human polls are worthless too. SIDs and Sportswriters voting who don't pay attention to games outside their conferences. Look at some of the incredulous receiving votes categories. Human, computer, they're all equally worthless this early in the season. The only thing they do now is for people to waste time on message boards like this arguing about them. GPI shouldn't have come out this early. I find it comical how people get all bent out of shape about them, human and computer.

Ignore them all now and check back the 1st week in Nov. They'll be less worthless then when everyone has 8-9 games under their belt vs the 4-5 they do now.

[slycat]

Where is Syntax Error?

Hes here. Just under a different user name. You can tell by the writing style. Plus who else defends the GPI with so much passion?

[slycat]

App State at 15, wtf!!!??? We better than all the other teams in front of us....I hope.

You're not better then McNeesexlolxxlolxxlolxxlolxxlolxxlolxxlolxxlolx