FPI rankings out for 2017

[lamont]

http://www.espn.com/blog/statsinfo/post/_/id/129505/ohio-state-is-no-1-in-preseason-fpi-1-0

Bama #2
Auburn #5
LSU #6

Seems to agree with my assessment

Georgia #13
Florida #15
Tenn #20
A&M #27
State #33
Kentucky #34
UPig #36
Mississippi #38

Really close to the way I see the SEC in 2017. Bama, Auburn, and LSU the cream of the West. The rest in a dogfight for 4th. Just tough for us playing at UPig and at A&M

[dawgday166]

http://www.espn.com/blog/statsinfo/post/_/id/129505/ohio-state-is-no-1-in-preseason-fpi-1-0

Bama #2
Auburn #5
LSU #6

Seems to agree with my assessment

Georgia #13
Florida #15
Tenn #20
A&M #27
State #33
Kentucky #34
UPig #36
Mississippi #38

Really close to the way I see the SEC in 2017. Bama, Auburn, and LSU the cream of the West. The rest in a dogfight for 4th. Just tough for us playing at UPig and at A&M

Does anyone have any data as to how accurate they are each season, from preseason to final bell. My perception is they usually miss a decent amount on some teams, besides Bama that is. But hell, anyone can pick Bama to be #1 or #2 and get it right.

Not saying we aren't down the scale about where they got us, but not sure if rest of SEC is as high as they have them either.

[Todd4State]

Does anyone have any data as to how accurate they are each season, from preseason to final bell. My perception is they usually miss a decent amount on some teams, besides Bama that is. But hell, anyone can pick Bama to be #1 or #2 and get it right.

Not saying we aren't down the scale about where they got us, but not sure if rest of SEC is as high as they have them either.

I think FPI changes throughout the season as results are tallied if I'm not mistaken. I don't think it's meant to be a metric that predicts the winners from game one on and not change.

[DeviousDawg]

Here is a look at the 2017 FPI rankings for SEC teams compared to the average composite recruiting rankings from 2013-2016 for SEC teams:



Team
Avg. Recruiting Ranking(2013-2016)

Team
2017 FPI Ranking


Alabama
1

Alabama
2


LSU
4.25

Auburn
5


Auburn
8

LSU
6


Georgia
8.25

Georgia
13


Texas A&M
10.75

Florida
15


Ole Miss
11

Tennessee
20


Florida
11.25

Texas A&M
27


Tennessee
12.25

South Carolina
32


South Carolina
20.75

Mississippi State
33


Arkansas
24.5

Kentucky
34


Mississippi State
26.5

Arkansas
36


Kentucky
32

Ole Miss
38


Mizzou
37.5

Mizzou
41


Vanderbilt
42.75

Vanderbilt
47



On average, excluding Ole Miss, the other 13 SEC teams FPI ranking is 5.96 spots from their average recruiting ranking. Ole Miss's FPI is 27 spots below their average recruiting ranking.

In conclusion, your recruiting ranking is generally who you are. Unless you are Ole Miss.

[blacklistedbully]

aTm pretty bad too, but UNM the worst by far.

[archdog]

Cause UM bought theirs. Haha.

[Prediction? Pain.]

Does anyone have any data as to how accurate they are each season, from preseason to final bell. My perception is they usually miss a decent amount on some teams, besides Bama that is. But hell, anyone can pick Bama to be #1 or #2 and get it right.

Not saying we aren't down the scale about where they got us, but not sure if rest of SEC is as high as they have them either.

Here's ESPN's 2016 preseason FPI rankings:

http://www.espn.com/college-football/statistics/teamratings/_/year/2016/key/20160826040000

Like lots of other preseason advanced-stat rankings, this one had some hits and misses nationally. In the SEC West, they predicted A&M right, but missed on all the others by varying degrees.

Order of FPI 2016 preseason rankings for SEC West teams, with predicted national FPI rank in parentheses:

1. LSU (2)
2. Bama (6)
3. U. Miss. (7)
4. Texas A&M (14)
5. Auburn (17)
6. Arkansas (25)
7. MSU. (30)

2016 final standings:

1. Bama (8 - 0)
2. Auburn (5-3)
3. LSU (5 -3)
4. Texas A&M (4 - 4)
5. Arkansas (3 - 5)
6. MSU (3 - 5)
7. U. Miss. (2 - 6)

So they got A&M right, were off on Bama, Arkansas, and MSU by 1 spot each, missed LSU by 2 spots, missed Auburn by 3 spots, and U. Miss. by 4 spots. As for final national rankings, those were probably even a bit more off across the board, other than with Bama and maybe, depending on the poll, Auburn. LSU, U. Miss., A&M, and Arkansas all finished way lower than predicted in terms of national ranks, and we didn't get any votes from the AP or Coaches to put us close to #30 nationally. Just glancing at the SEC East preseason FPI rankings, they look off by similar margins.

But again, the accuracy of these things is variable regardless of the system because of all the different factors they weigh (recruiting, returning experience, schedule, offensive and defensive efficiency stats from the previous year, etc.) Here's another advanced-stat preseason poll from 2016 for comparison:

http://www.sbnation.com/college-football/2016/2/8/10922288/2016-college-football-rankings-projections

Here's how 2016's S&P+ preseason rankings shook out for the SEC West shook out, again with national ranks in parentheses:

1. Bama (1)
2. LSU (2)
3. U. Miss. (7)
4. Arkansas (17)
5. MSU (21)
6. Auburn (24)
7. Texas A&M (25)

So other than Bama, this poll also whiffed on everyone else by 1, 3, or 4 spots.

[Prediction? Pain.]

I think FPI changes throughout the season as results are tallied if I'm not mistaken. I don't think it's meant to be a metric that predicts the winners from game one on and not change.

Yeah, it apparently changes every day, per ESPN. But they do tout it as predictive:


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 10,000 simulations of the rest of the season using FPI, results to date, and the remaining schedule. Ratings and projections update daily.

I didn't realize this till looking at the FPI a minute ago, but it actually includes predicted total wins and losses, too.

[dawgday166]

I pay a little attention to the FPI, but only casually. Until next year's preseason ranking came out, since 2014 they've always had OM too high IMO. They seemed to have some kind of a secret crush on Freeze or something like that.

And ... they've missed on some others too, but they usually get Bama about right.

[Beaver]

http://www.espn.com/blog/statsinfo/post/_/id/129505/ohio-state-is-no-1-in-preseason-fpi-1-0Kentucky #34

Looking at Kentucky's schedule, they could very easily be 5-1 when they come to Starkville. And they'll be coming off a bye...

[smootness]

http://www.espn.com/blog/statsinfo/post/_/id/129505/ohio-state-is-no-1-in-preseason-fpi-1-0

Bama #2
Auburn #5
LSU #6

Seems to agree with my assessment

Georgia #13
Florida #15
Tenn #20
A&M #27
State #33
Kentucky #34
UPig #36
Mississippi #38

Really close to the way I see the SEC in 2017. Bama, Auburn, and LSU the cream of the West. The rest in a dogfight for 4th. Just tough for us playing at UPig and at A&M

So now you're attempting to preemptively prove yourself correct based on guesses by other people?

[HSVDawg]

Here's ESPN's 2016 preseason FPI rankings:

http://www.espn.com/college-football/statistics/teamratings/_/year/2016/key/20160826040000

Like lots of other preseason advanced-stat rankings, this one had some hits and misses nationally. In the SEC West, they predicted A&M right, but missed on all the others by varying degrees.

Order of FPI 2016 preseason rankings for SEC West teams, with predicted national FPI rank in parentheses:

1. LSU (2)
2. Bama (6)
3. U. Miss. (7)
4. Texas A&M (14)
5. Auburn (17)
6. Arkansas (25)
7. MSU. (30)

2016 final standings:

1. Bama (8 - 0)
2. Auburn (5-3)
3. LSU (5 -3)
4. Texas A&M (4 - 4)
5. Arkansas (3 - 5)
6. MSU (3 - 5)
7. U. Miss. (2 - 6)

So they got A&M right, were off on Bama, Arkansas, and MSU by 1 spot each, missed LSU by 2 spots, missed Auburn by 3 spots, and U. Miss. by 4 spots. As for final national rankings, those were probably even a bit more off across the board, other than with Bama and maybe, depending on the poll, Auburn. LSU, U. Miss., A&M, and Arkansas all finished way lower than predicted in terms of national ranks, and we didn't get any votes from the AP or Coaches to put us close to #30 nationally. Just glancing at the SEC East preseason FPI rankings, they look off by similar margins.

But again, the accuracy of these things is variable regardless of the system because of all the different factors they weigh (recruiting, returning experience, schedule, offensive and defensive efficiency stats from the previous year, etc.) Here's another advanced-stat preseason poll from 2016 for comparison:

http://www.sbnation.com/college-football/2016/2/8/10922288/2016-college-football-rankings-projections

Here's how 2016's S&P+ preseason rankings shook out for the SEC West shook out, again with national ranks in parentheses:

1. Bama (1)
2. LSU (2)
3. U. Miss. (7)
4. Arkansas (17)
5. MSU (21)
6. Auburn (24)
7. Texas A&M (25)

So other than Bama, this poll also whiffed on everyone else by 1, 3, or 4 spots.

Good info. Basically any asshole on here can throw out a preseason prediction with no advanced data whatsoever like the FPI uses, and still probably achieve the same degree of accuracy.

[lamont]

So now you're attempting to preemptively prove yourself correct based on guesses by other people?


No- showing that the FPI agrees with my preseason assessment

[smootness]

No- showing that the FPI agrees with my preseason assessment

So, what I said.

[THE Bruce Dickinson]

No- showing that the FPI agrees with my preseason assessment

In week 9 of 2016 the FPI had a 3-5 Ole Miss team ranked as the #14 team in the country.

The FPI is garbage. Your opinion being the same as the FPI would not be something I would brag about.

[lamont]

That's the fun We'll see how it plays out

[dawgday166]

In week 9 of 2016 the FPI had a 3-5 Ole Miss team ranked as the #14 team in the country.

The FPI is garbage. Your opinion being the same as the FPI would not be something I would brag about.

This is why I don't put a lot of faith in it either. Whatever algorithms they are using or whatever data they are feeding those algorithms certainly appears to me to be somehow manipulated at times IMO. OM has been their "darling". OM went on from there to lose to Vandy and Us.

Also, other SEC teams seem to stay up there sometimes (AU & LSU) when they should be further down IMO.

[Jack Lambert]

Here is a look at the 2017 FPI rankings for SEC teams compared to the average composite recruiting rankings from 2013-2016 for SEC teams:



Team
Avg. Recruiting Ranking(2013-2016)

Team
2017 FPI Ranking


Alabama
1

Alabama
2


LSU
4.25

Auburn
5


Auburn
8

LSU
6


Georgia
8.25

Georgia
13


Texas A&M
10.75

Florida
15


Ole Miss
11

Tennessee
20


Florida
11.25

Texas A&M
27


Tennessee
12.25

South Carolina
32


South Carolina
20.75

Mississippi State
33


Arkansas
24.5

Kentucky
34


Mississippi State
26.5

Arkansas
36


Kentucky
32

Ole Miss
38


Mizzou
37.5

Mizzou
41


Vanderbilt
42.75

Vanderbilt
47



On average, excluding Ole Miss, the other 13 SEC teams FPI ranking is 5.96 spots from their average recruiting ranking. Ole Miss's FPI is 27 spots below their average recruiting ranking.

In conclusion, your recruiting ranking is generally who you are. Unless you are Ole Miss.

Damn Ole Miss has really fallen.

[MetEdDawg]

This is why I don't put a lot of faith in it either. Whatever algorithms they are using or whatever data they are feeding those algorithms certainly appears to me to be somehow manipulated at times IMO. OM has been their "darling". OM went on from there to lose to Vandy and Us.

Also, other SEC teams seem to stay up there sometimes (AU & LSU) when they should be further down IMO.

There's a definite name and recruiting bias. All these websites and services make money when big name teams and highly ranked recruiting teams. Unfortunately, we aren't a sexy name because we aren't a big name and don't finish high in recruiting. So they don't pander to us as much as they do other programs.

[Todd4State]

There's a definite name and recruiting bias. All these websites and services make money when big name teams and highly ranked recruiting teams. Unfortunately, we aren't a sexy name because we aren't a big name and don't finish high in recruiting. So they don't pander to us as much as they do other programs.

They also make money because they "are right". Even though my Mom could tell you that Alabama has the number one class right now without knowing anyone that is actually in it. Alas when Bama wins the NC in five years they will write a retrospective about how right their experts were the whole time.