Fremeau Efficiency Ratings

College football power ratings and analysis

FEI Week 9 Ratings and Garbage Time

by Brian Fremeau

I've been thinking about garbage time lately, not because I'm considering another revision to my definition of garbage time, but because I'm curious to know if there is any insight to be gleaned from examining the relationship between non-garbage time and garbage time in more detail.

Readers may recall that I updated my criteria that distinguishes non-garbage possessions from garbage possessions this offseason. The changes weren't dramatic, but they did result in counting more drives as non-garbage than I had previously. Based on the new criteria applied to every FBS game from 2007 to 2017, 10.5 percent of all possessions have been classified as garbage possessions. FBS games have averaged 23.8 non-garbage possessions per game (11.9 non-garbage possessions and 11.2 non-garbage offensive drives per team per game) in that span. So far in 2018, 10.7 percent of all possessions have been classified as garbage possessions, and FBS games have averaged 23.5 non-garbage possessions per game.

A couple of recent comments by readers and Twitter followers had me pick at the garbage time data a bit more. Last week in the FEI Week 8 column comments, one reader asked about discrepancies between FEI ratings and Bill Connelly's S&P+ ratings, and I summarized what I judged to be the primary differences between our respective systems as follows:

  • Bill and I calculate garbage time differently. I'm not certain, but it is possible that Bill is measuring Alabama defensive performance in the second half of its games differently than I am. Anecdotally, the Crimson Tide have been exceptional until later in games, and I wonder if S&P is picking up on that drop-off in a way that FEI is not.
  • Play success and drive success can be two very different things. A team may run four consecutive plays of 8-yard gains and then throw an interception. S&P might judge that to be four successful plays out of five, and FEI might treat that result similarly to a team that threw a pick on the second or third play of that sequence.
  • We have our own ways of calculating opponent adjustments. There are ways to emphasize single-game performances versus season-long performances, even with the same underlying data, but with different results.

The first point is an important one. Which raw data is included and which raw data is not could potentially have a significant impact on the output of success rates and opponent adjustments. I posited that Bill may be counting more data than I do for Alabama, for instance, but the opposite could also be true. Bill also changed his garbage time criteria this offseason, and though there may be some synchronization of our raw data sets as a result of his changes and mine, there are certainly some differences. Bill indicates that play data after a team reaches a 37-point lead in the second quarter is discounted by S&P+. My definition counts every first-half possession (except for end-of-half clock kills) as non-garbage. So when Alabama took a 40-point lead against Arkansas State in Week 2 into halftime, and expanded that lead to 50-7 late in the third quarter, I'm counting all of those possessions that led to the 43-point lead as non-garbage, whereas Bill stopped counting play data in the second quarter.

To be honest, though, I'm only speculating on some of these differences. For as much as we have collaborated over the years, neither one of us has looked under the hood of each other's formulas. Bill and I probably need to collaborate on a deeper FEI vs. S&P+ at some point and pick at each other's stuff a bit more than we have to date, in order to both answer these kinds of questions more concretely, but also because to probably learn a bit more to improve our own systems as a result of that kind of evaluation.

One of my Twitter followers also inquired about garbage time last week, but was interested more in the relative speed at which teams overcome their opponents and get to garbage time. This may seem like a relatively straightforward question, calculating the number of non-garbage possessions per game by team, but it doesn't get at what this person was after.

Fewest Non-Garbage Possessions Per Game
Team W-L NGP/G
Army 6-2 8.9
Kansas State 2-5 9.4
Coastal Carolina 4-3 9.7
Washington 5-3 9.8
Connecticut 0-7 9.9
Georgia State 1-6 9.9
Louisiana Lafayette 3-4 9.9
Florida International 5-2 9.9
Mississippi State 4-3 9.9
Alabama 8-0 10.0

Alabama is on this list, and is the only team on this list that has consistently overwhelmed its opponents and pushed games into garbage time with ruthless efficiency. Army has some of that overwhelming efficiency going for it as well, but the Knights do so by controlling the ball and limiting overall game possessions more often than not. Most of the rest of the teams on this list may have some ball control elements, but some are here simply because they are overcome by opponents crushing them and getting to garbage time, not the other way around.

If we approach the question differently, we may get closer to what we're after. Instead of calculating the number of non-garbage possessions per game, let's instead calculate the number of non-garbage possessions as a percentage of the total possessions per game.

Smallest Percentage of Possessions
Per Game that are Non-Garbage
Team W-L NGP%
Alabama 8-0 .745
Fresno State 6-1 .789
Central Florida 6-0 .810
Connecticut 0-7 .812
Georgia State 1-6 .821
Utah State 6-1 .824
Oklahoma 7-1 .825
Maryland 5-3 .829
Michigan 7-1 .837
Louisiana Lafayette 3-4 .841

This list still includes teams that are on the wrong end of acceleration toward garbage time, but some of the best teams at pushing games into garbage time certainly rise toward the top. These are dominant teams in non-garbage time. Alabama, Fresno State, Central Florida, Utah State, Oklahoma, and Michigan all rank among the top six teams in the nation in Possession Success Rate, a measure of each team's ability to get scores on offense and stops on defense. Those two things in combination lead to dominating victories, and usher in garbage time sooner as a result.

Maybe there's more to explore with regard to garbage time than these two questions, and as always, I'm open to feedback in the comments here or on Twitter.

FEI Week 9 Ratings

The Fremeau Efficiency Index (FEI) is a college football rating system based on opponent-adjusted possession efficiency. Adjusted Possession Advantage (APA) ratings represent the per-possession scoring advantage a team would be expected to have on a neutral field against an average opponent, calculated as a function of current FEI overall, offense, defense, and special teams ratings.

Strength of Schedule ratings (PSOS) represent the average number of losses an elite team (two standard deviations better than average) would have against the team's regular season schedule to date. Offensive FEI (OFEI) is scoring value generated per drive adjusted for starting field position and opponent defenses faced. Defensive FEI (DFEI) is scoring value generated per opponent drive adjusted for starting field position and opponent offenses faced. Special Teams FEI (SFEI) is scoring value generated per possession by a team's non-offensive and non-defensive units adjusted for opponent special teams units faced. The team's record to date against opponents ranked in the FEI top 10 (v10), top 20 (v20), top 30 (v30), top 40 (v40), and top 50 (v50) are also provided.

Ratings and supporting data are calculated from the results of non-garbage possessions in FBS vs. FBS games.

Click here for ratings for all 130 FBS teams.

Rk Team Rec FEI APA Rk PSOS Rk OFEI Rk DFEI Rk SFEI Rk v10 v20 v30 v40 v50
1 Clemson 7-0 .277 2.40 2 .46 84 2.93 17 .7 1 .052 41 0-0 0-0 1-0 2-0 4-0
2 Georgia 6-1 .270 2.29 3 1.30 7 3.54 4 1.5 16 .216 8 0-1 1-1 2-1 3-1 3-1
3 LSU 6-1 .269 1.84 6 1.29 8 2.52 43 1.4 11 .346 2 1-0 2-1 4-1 4-1 4-1
4 Alabama 8-0 .252 2.82 1 .26 114 4.10 2 .8 2 .002 61 0-0 0-0 1-0 1-0 2-0
5 Oklahoma 7-1 .240 2.18 4 .67 59 4.22 1 2.1 59 .111 23 0-0 1-0 2-0 2-1 2-1
6 Michigan 7-1 .216 1.86 5 1.00 24 2.83 19 1.1 3 .231 5 1-0 1-1 2-1 2-1 3-1
7 Kentucky 6-1 .208 1.40 13 .91 32 2.12 74 1.3 8 .223 6 0-0 2-0 3-0 4-0 4-1
8 Michigan State 5-3 .198 1.36 16 1.11 13 2.22 67 1.4 10 .153 17 0-1 2-1 2-2 3-2 4-3
9 Iowa 5-2 .183 1.39 15 .72 52 2.29 64 1.2 5 .251 3 0-0 1-1 1-1 1-1 2-2
10 Washington 5-3 .180 1.42 12 .64 61 2.71 26 1.6 20 .048 44 0-0 1-0 1-1 1-1 2-1
11 Florida 5-2 .179 1.29 20 1.53 5 2.47 47 1.6 23 .243 4 1-2 2-2 2-2 2-2 2-2
12 Notre Dame 8-0 .179 1.49 10 .42 92 2.71 27 1.4 14 -.046 75 1-0 1-0 1-0 1-0 2-0
13 Penn State 6-2 .175 1.40 14 .78 43 2.54 41 1.4 13 .157 15 1-1 1-2 1-2 1-2 2-2
14 Ohio State 7-1 .172 1.55 7 .74 49 3.53 5 2.1 54 .096 29 0-0 1-1 1-1 1-1 2-1
15 Iowa State 4-3 .170 1.32 17 1.06 15 2.30 63 1.3 7 .103 26 0-2 0-2 2-2 2-2 3-2
Rk Team Rec FEI APA Rk PSOS Rk OFEI Rk DFEI Rk SFEI Rk v10 v20 v30 v40 v50
16 Mississippi State 4-3 .161 1.49 9 1.39 6 2.72 25 1.2 6 .085 35 0-2 0-3 1-3 1-3 2-3
17 Washington State 6-1 .160 1.30 18 .38 98 3.59 3 2.5 91 -.029 67 0-0 1-0 1-0 1-0 2-0
18 Purdue 4-4 .157 1.21 23 .91 30 2.77 23 1.8 36 .102 28 0-1 1-1 1-3 1-3 2-3
19 Utah 5-2 .148 1.18 24 .68 56 2.68 28 1.7 29 .159 14 0-1 0-2 0-2 0-2 1-2
20 Fresno State 6-1 .147 1.50 8 .17 123 3.00 14 1.4 12 -.080 85 0-0 0-0 0-0 0-0 0-0
21 North Carolina State 4-2 .146 1.24 22 .96 26 3.25 10 2.2 64 .086 34 0-1 0-1 0-1 1-2 2-2
22 Duke 4-3 .140 .88 29 .56 72 2.37 57 1.9 42 -.042 74 0-0 0-0 2-0 2-1 3-1
23 West Virginia 5-1 .138 1.30 19 .54 73 2.94 16 1.6 26 .018 51 0-0 0-1 1-1 1-1 1-1
24 Missouri 3-4 .136 1.26 21 1.66 3 2.82 20 1.6 22 .223 7 0-3 1-3 1-3 1-4 1-4
25 Central Florida 6-0 .134 1.44 11 .06 130 3.39 6 1.8 32 -.195 113 0-0 0-0 0-0 0-0 0-0
26 Auburn 4-3 .130 .92 28 1.03 20 2.07 77 1.5 15 .121 20 1-1 1-2 1-2 1-2 1-2
27 Army 6-2 .127 1.15 25 1.00 23 3.28 9 2.2 66 -.030 68 0-1 0-1 0-2 1-2 1-2
28 Miami 4-3 .121 .93 27 .87 33 2.07 79 1.4 9 .027 48 0-1 0-1 0-1 0-2 0-3
29 Northwestern 5-3 .117 .77 36 1.13 11 2.03 83 1.6 24 .107 24 1-1 2-1 2-2 2-2 3-2
30 Texas Tech 4-3 .116 .84 32 .70 55 2.58 37 2.0 48 .114 22 0-0 0-1 0-2 1-2 2-2

Comments

2 comments, Last at 01 Nov 2018, 5:10pm

1 Re: FEI Week 9 Ratings and Garbage Time

by Chappy // Nov 01, 2018 - 12:16pm

This article was helpful. (I asked the question about Alabama last week). My first reaction was that you had the difference between your systems flipped since S&P apparently discounts garbage time more--Alabama's D is much lower ranked in S&P then your system. However, your later analysis makes me think maybe Alabama has such a large amount of garbage time, even though discounted, there is just such a large volume of plays/possessions happening that the second string or garbage time defense is disproportionately bringing down their ranking.

I suppose an alternative interpretation is that S&P has the "correct" discounting of garbage time and your system is giving more credit than is due for plays/possessions made in garbage time.

Login or Register to post comments

2 Re: FEI Week 9 Ratings and Garbage Time

by Brian Fremeau // Nov 01, 2018 - 5:10pm

It is difficult to tell exactly what factors are in play on this, but I agree that there may be multiple things happening and a simple explanation of one system counting things different than the other may be too simple an answer. Another factor may be that my counting more possessions in a blowout deeper into the game may, in fact, be dragging down Alabama's opponent's ratings more than Bill's method drags down their opponents. If so, the issue may not be due to Alabama's raw data but the raw data of its opponents making them look worse in my system than in Bill's; which would then have an impact on the opponent-adjustment part of the equation.

Login or Register to post comments