FEI Preseason Projections

by Brian Fremeau

One of the quietest off-seasons in memory erupted suddenly last month in the form of a pair of scandals at Maryland and Ohio State, leaving myself and many other college football fans feeling a bit less enthusiastic about the year ahead. I can say that I'm not looking forward to being reminded of those scandals throughout the season, especially if the Buckeyes make a run at the College Football Playoff, as they are well positioned to do. There's a negative effect such situations have had on my ultimate enjoyment of the sport of late. College football is imperfect, and always has been, but the compounding recent reminders of its imperfections are especially distasteful.

That's a less-than-glowing endorsement of the state of the game in 2018, but let me be clear. I'm certainly looking forward to the games themselves. College football always delivers with its games, often in style, punctuated by unpredictable results and preposterous "have you ever seen anything like that?" moments that ultimately define the season. I'm also looking forward to pulling data from the games, my weekly exercise in better understanding what actually happened by examining possession efficiency in greater detail. My approach to the development and refinement of the FEI ratings and supplemental data found on both this site and my own is designed to strip all of the narrative excitement from the equation. I'm wired in such a way that the data itself renews my interest and love of college football year after year. It's a critical distraction from the narrative noise and nonsense.

I spent a significant amount of time this off-season reviewing and auditing my data set of FBS-versus-FBS game possessions since 2007. Among other reasons for conducting this exercise, I slightly altered my definition of garbage-time data. I actually expanded my definition of garbage time somewhat, to include more end-of-game situations in which the outcome was largely in hand, but there remained a chance for a comeback victory by the losing team. For instance, I previously counted an extended drive that ended the game by a team losing by two scores as "garbage time" because the results of that drive alone as contested would not have changed the game outcome. Those drives were added back into my "non-garbage" possession set under the new definition criteria, because both the team and opponent likely played at least a significant portion of that drive as if the game was not yet in doubt.

Unless otherwise noted, all ratings and supporting data I produce are calculated only after first filtering out garbage possessions, defined as follows:

• An offensive possession of two plays or fewer that runs out the clock to conclude the first half and does not result in a turnover, score, or field goal attempt.
• An offensive possession of two plays or fewer that runs out the clock to conclude the second half with the score tied and does not result in a turnover, score, or field goal attempt.
• A possession in the second half of a game in which eight times the number of the losing team's remaining possessions plus one is less than the losing team's scoring deficit at the start of the possession
• An offensive possession or non-offensive scoring possession by the winning team leading by eight points or fewer at the start of the possession that runs out the clock to conclude the game.

Since 2007, 10.5 percent of all possessions have been classified as garbage possessions according to these criteria. 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.

I updated my national possession efficiency by starting field position data tables and graphics as well. These are representative of my core approach to evaluating team unit performances: that the results of a given offensive drive are impacted in large part by starting field position. Click on the chart image below for the full-size version.

I will use a smoothed-out trendline of the national average net points scored per drive by starting field position data over the entire 11-year span as the benchmark for raw efficiency expectations in my calculations in 2018. That said, I'm keeping a close eye on the year-over-year trendlines in a few key categories. As much as the summary chart above appears to suggest that points per drive by starting field position data is relatively consistent (aside from sample size discrepancies as starting field position approaches the end zone) each year, that's not actually true. Overall offensive efficiency has been trending up since 2007.

Offenses are producing more points per possession while starting further and further from the end zone. The steady climb of the starting field position chart is worth paying very close attention to in 2018, in part due to the new kickoff receiving rules changes that will give teams the option to elect to signal for a fair catch on kickoffs received inside the team's own 25-yard line, and will be awarded with a touchback and starting field position at the 25-yard line as a result.

It's also important to remember that possession efficiency is often interrelated to next-possession efficiency. Offenses are moving the ball better, gaining 2.2 more yards per drive on average in 2017 than ten years earlier. Offenses earned 45.7 percent of available yards in 2017 compared with only 44.0 percent of available yards in 2007. Offensive efficiency is increasing in part due to the turnover rate steadily decreasing, down from 13.8 percent in 2007 to only 11.7 percent in 2017. Turnovers are most often attributed with next-possession short field position for the opponent, and that turnover rate trendline may be one of the most significant factors impacting the starting field position trendline.

These are the data points I'll be tracking and analyzing more closely this season. I'll also be developing an alternative (and more predictive, hopefully) version of the FEI ratings as the season progresses. Let's get it started, let's enjoy the games, and let's dig into the data.

FEI preseason projected ratings

The Fremeau Efficiency Index (FEI) is a college football rating system based on opponent-adjusted possession efficiency. Preseason projections are based on five-year results, recruiting success, and returning offensive and defensive production. Strength of Schedule ratings (SOS) represent the average number of losses an elite team (two standard deviations better than average) would have against the team's regular season schedule. Each team's percent likelihood of finishing the regular season with zero losses (0L), one loss (1L), etc, are calculated from individual game win likelihoods, each a function of the preseason projected FEI ratings and an adjustment for home field advantage.