Week 7 Coaches Poll

  • You are viewing Orangepower as a Guest. To start new threads, reply to posts, or participate in polls or contests - you must register. Registration is free and easy. Click Here to register.

Cimarron

It's not dying I'm talking about, it's living.
Jun 28, 2007
53,954
18,069
1,743
#61
https://www.msn.com/en-us/sports/nc...its-new-top-25-rankings/ar-AAPCWAp?li=BBnbfcL

ESPN really doesn't think highly of us. This explains the nearly 80-20 FPI in favor of Texas yesterday and the same for Iowa State this coming week.

For B12: Texas at 12, ISU at 9, Baylor ranked.
Rest of the rankings are a joke too: Clemson, Florida, Nebraska.
All I have to say... what the crap are they on?
Is there such a thing as a computer model? All that means is we can put any bias we want into a model, use excel to do some simple addition and call it a computer model to give it credibility. It’s meaningless and no more accurate than the person who programs the model. Unless they tell us the model it’s meaningless. I shouldn’t say meaningless, but rather there is little to gauge it’s accuracy.
 
Jun 16, 2008
672
81
1,578
Maryland
#62
Is there such a thing as a computer model? All that means is we can put any bias we want into a model, use excel to do some simple addition and call it a computer model to give it credibility. It’s meaningless and no more accurate than the person who programs the model. Unless they tell us the model it’s meaningless.
There is a saying that every model is wrong some models are useful. Found this description on the ESPN page that lists their FPI rankings. They say they run a bunch of simulations (which means a model of some sort) but they don't say what they are using in any real detail.

https://www.espn.com/blog/statsinfo/post/_/id/122612/an-inside-look-at-college-fpi

"The Football Power Index (FPI) is a measure of team strength that is meant to be the best predictor of a team's performance going forward for the rest of the season. FPI represents how many points above or below average a team is. Projected results are based on 20,000 simulations of the rest of the season using FPI, results to date, and the remaining schedule. Ratings and projections update daily. FPI data from seasons prior to 2019 may not be complete."
 
Last edited:

OctaviusRexLTH

Deputy
A/V Subscriber
Aug 16, 2019
1,644
646
243
30
Alabama
#63
Found this description on the ESPN page that lists their FPI rankings.
"The Football Power Index (FPI) is a measure of team strength that is meant to be the best predictor of a team's performance going forward for the rest of the season. FPI represents how many points above or below average a team is. Projected results are based on 20,000 simulations of the rest of the season using FPI, results to date, and the remaining schedule. Ratings and projections update daily. FPI data from seasons prior to 2019 may not be complete."
FPI just can’t let go of our rocky start.
 

Cimarron

It's not dying I'm talking about, it's living.
Jun 28, 2007
53,954
18,069
1,743
#72
There is a saying that every model is wrong some models are useful. Found this description on the ESPN page that lists their FPI rankings. They say they run a bunch of simulations (which means a model of some sort) but they don't say what they are using in any real detail.

https://www.espn.com/blog/statsinfo/post/_/id/122612/an-inside-look-at-college-fpi

"The Football Power Index (FPI) is a measure of team strength that is meant to be the best predictor of a team's performance going forward for the rest of the season. FPI represents how many points above or below average a team is. Projected results are based on 20,000 simulations of the rest of the season using FPI, results to date, and the remaining schedule. Ratings and projections update daily. FPI data from seasons prior to 2019 may not be complete."
They didn't really share anything.

We take into consideration points scored, points allowed, home field advantage, blah, blah, blah...

Every model has a degree of error, Y = a + bX +e

There is a lot of noise in these models that can't be accounted for. Sick/hurt players, which ones are sick/hurt, the effect of the backup, and the list goes on. The trick in prediction models is determining which of those many variables are significant.

Here is a simple model for the Big 12. My prediction is that the higher ranked team from 2020 will win the game in 2021. So far that has been correct 82% of the time.
 
Jul 11, 2007
376
214
1,593
#73
I guess what the ESPN FPI thing says is something like this. You could be Miss St, lose to bama, florida, georgia, LSU, Ole Miss and Kentucky all by a margin of 1 point. Even though you are 0-6 because you only have fluff left you are #1 in our FPI poll.

Prediction of how you should do moving forward only and this tells us very little.

This isn't too far off base when you look where they have Nebraska.
 

Cimarron

It's not dying I'm talking about, it's living.
Jun 28, 2007
53,954
18,069
1,743
#74
I guess what the ESPN FPI thing says is something like this. You could be Miss St, lose to bama, florida, georgia, LSU, Ole Miss and Kentucky all by a margin of 1 point. Even though you are 0-6 because you only have fluff left you are #1 in our FPI poll.

Prediction of how you should do moving forward only and this tells us very little.

This isn't too far off base when you look where they have Nebraska.
I suspect the ESPN prediction model is designed to support their own bias.
 

wrenhal

Federal Marshal
Aug 11, 2011
12,273
4,552
1,743
#77
Nov 18, 2010
3,011
1,193
1,743
#78
And oh by the way, Alabama lost to an unranked team a week ago and they're already back in the 'blueblood invitational'.

College football needs a real playoff.
 
Dec 11, 2011
1,359
1,183
1,743
38
#79
And oh by the way, Alabama lost to an unranked team a week ago and they're already back in the 'blueblood invitational'.

College football needs a real playoff.
Yeah, they never left!
I'll believe they're out of it when dirt is shoveled onto their corpse.
Maybe not even then....