Stats Football Outsiders Rankings

[Cam]

I didn't see a thread for this, so I'm sorry if this has already popped up elsewhere. For those who don't know, Football Outsiders releases rankings of teams offense, defense, and overall performance each year which have seemed to prove pretty reliable. However, they wait until after the 7th week to do so when they have adequate data. The ratings combine two different measurement methods: Fremeau Efficiency Index and S&P rankings. As I understand it, FEI focuses more on the outcome of drives and measures your efficiency in scoring/preventing scoring, while S&P operates on more of a play by play basis. You can read up on their metrics on their site (http://www.footballoutsiders.com/stats/fei).

Here's how we stand so far with our current data:
FEI Overall Rank: 48 (not updated to include last week yet)
FEI Offense Rank: 23
FEI Defense Rank: 42

S&P Overall Rank: 27
S&P Offense Rank: 33
S&P Defense Rank: 35

I think these are some of the better statistical rankings to look at. Our offense favors FEI, for instance, and you can see that reflected both in our rankings and Army's (6th in FEI and 36 in S&P). It's fun to chart the movement of teams over the course of the season and look at how we fair against higher ranked teams (Miami around top 10) and those lower (Tennessee ~65, UNC ~85, Pitt ~100).

 

[ibeattetris]

I was looking at this as well earlier in the week. What is most disappointing in this is our defense ended last season ranked 94 in FEI and 57 in SP. With this kind of improvement I think most of us would have expected to be doing a bit better.

Hopefully we right the ship against Wake tomorrow.

 

[AE 87]

I'm typically a big fan of this site and its advanced stats. However, I think there must be something wrong with their strength of schedule adjustments to rank Army at #6 offense.

 

[ibeattetris]

I'm typically a big fan of this site and its advanced stats. However, I think there must be something wrong with their strength of schedule adjustments to rank Army at #6 offense.

Indeed it seems to be really weird.
Army has played 6 games against tracked teams. The DFEI rank and score are below for their opponents
Buffalo: 74 2.39
Ohio State: 3 0.86
Tulane: 33 1.75
UTEP: 122 3.31
RICE: 121 3.23
Eastern Michigan: 36 1.77
Average DFEI: 2.22

Army's non adjusted score is actually better than their adjusted score (3.4 unadjusted vs 3.35 adjusted) but the difference wasn't much.

I decided to look at the team with the biggest move in adjusted efficiency. Winner is South Florida going from rank 24 unadjusted offense to rank 68 adjust. A swing of 0.74 points (2.91 to 2.17). Their opponents:
Connecticut: 125 3.42
Illinois: 108 2.85
Temple: 40 1.83
East Carolina: 130 4.06
Cincinnati: 120 3.18
Average DFEI: 3.07

Then for fun here is GT: 2.99 unadjusted (19) to 2.82 adjusted (23). Our opponents adjusted ranks:
UTenn: 54 2.03
Pitt: 92 2.65
UNC: 84 2.94
Miami: 27 1.59
Average DFEI: 2.30

Maybe it is still too early in the season for real strengths of schedule to have materialized.

Edit to note: Our rate of turnovers on offense is ranked 123. That has killed us.
Our defense turnover rate ranked 115

 

[smathis30]

FEI usually over emphasized third down percentage, which disproportionately helps tripple option offenses verse other schemes. In my own advanced stats, I keep track of how a team does relative to how every other team has done against said team (eg. compare army in their game vs ohio state relative to how every team has done against ohio state) Army has played above average against every team they played. I have them as the 4th best offense unadjusted and 26th best adjusted. GT is the same, but i had our offensive performance against Jacksonville state as a below average game against an FCS school (by 1 point, but still) Every other game was in the green. I have GT as the 3rd best offense unadjusted and 34th adjusted as our slate of games has been complete ***, my dudes. 101st strongest sos for us vs 110th for army