by Brian Fremeau
The Georgia Bulldogs came into their game against South Carolina last Saturday as one of the top contenders for the College Football Playoff. In last week's FEI ratings, which were still based in small part on preseason projections, Georgia ranked fourth overall and was one of only four teams that boasted a top-10 offense and defense. They also ranked fourth overall in likelihood to run the table (34%) to an undefeated regular season record. And they were playing at home against the Gamecocks, a team coming into the game with only one FBS win under its belt in 2019.
South Carolina defeated Georgia in overtime by a final score of 20-17. The Bulldogs outgained the Gamecocks in drive yards (measured from starting field position to end field position) by 179 yards in total and by an average of 9.1 yards per drive. Georgia's offense scored two touchdowns, South Carolina's scored only one. Georgia punted only three times, South Carolina punted seven times. But Georgia turned the ball over four times to South Carolina's zero turnovers, including a pick-six that put the Gamecocks up 17-10 in the second quarter, and those turnovers played a huge role in sealing Georgia's fate.
Was this the biggest upset of the year to date? In terms of betting lines, it wasn't even the biggest upset of the weekend. Bowling Green defeated Toledo by a final score of 20-7 as a 26.5-point underdog. The Gamecocks were 20.5-point dogs against Georgia, officially the third-biggest upset of the year in terms of betting lines. There were five upset victories in FBS vs. FBS games in 2018 among teams that entered the game as an underdog of 20 or more points. Since 2007, there have been only 59 such upsets in 1,595 opportunities. Favorites by 20 or more points have had a 96.3% win rate over the last 12 seasons. Teams favored by between 20 and 25 points have had a 94.2% win rate in the same span.
FEI projections heading into the game had Georgia projected to win by 29.4 points, with a 94.5% win likelihood. FEI also projected Toledo to defeat Bowling Green, but not by as large a margin -- 27.5 points and 92.7% win likelihood. As far as this projection model is concerned, South Carolina's victory was the biggest upset of the year to date.
How big of an upset was it retroactively? The answer to this question will evolve over the course of the season. If South Carolina puts together a strong second half of the season, or if Georgia trips up again once or twice down the stretch, we may look back on this game as surprising in the moment, but not extraordinary in terms of the relative strengths of the two teams. On the other hand, if South Carolina takes a precipitous downturn over the remainder of the year, and Georgia steamrolls everyone else it its path, we may look back on this past weekend's result as even more inexplicable than our preconceptions about both teams going into the game.
If we use this week's FEI ratings, updated to include the results of all FBS vs. FBS games played to date and devoid of any preseason projection influence, we can retroactively calculate the likelihood of every game result played thus far. For this analysis, I'm only considering the outcome win or lose, not the margin of the game result. South Carolina is now ranked 35th overall in this week's FEI ratings and Georgia has slipped to seventh overall. Instead of a 94.5% win likelihood, FEI would now project an 80.0% win likelihood for the Bulldogs -- still a significant upset, but not the most retroactively surprising result of the year based on current FEI ratings.
Below are the top 10 biggest retroactive surprise results of the season to date, based on current FEI ratings and the actual game winner (not the margin of victory or defeat):
- 8/30: Nevada 34, Purdue 31 (10.9% likelihood)
- 10/12: Bowling Green 20, Toledo 7 (12.0%)
- 8/31: Wyoming 37, Missouri 31 (12.9%)
- 9/14: BYU 30, USC 27 (16.3%)
- 9/21: Pittsburgh 35, Central Florida 34 (18.5%)
- 9/28: Massachusetts 37, Akron 29 (19.8%)
- 10/12: South Carolina 20, Georgia 17 (20.0%)
- 10/5: Stanford 23, Washington 13 (21.9%)
- 9/21: UCLA 67, Washington State 63 (25.4%)
- 10/5: Virginia Tech 42, Miami 35 (25.8%)
Whether or not the Georgia game will be remembered at season's end as the biggest shocker of 2019 remains to be seen. It's impact on Georgia's playoff prospects appears to be very significant, however. The Bulldogs are by no means out of the playoff chase, with remaining games against strong opponents in Florida, Missouri, and Auburn, and the potential for an SEC Championship Game berth and a showdown with whichever juggernaut emerges from the SEC West. But with top opponents comes a much more precarious situation and no more room for error. Prior to the game against South Carolina, Georgia had a 79% chance of finishing the regular season with zero or one loss. Coming out of the game, the Bulldogs project to have only an 18% chance of finishing the regular season at 11-1. No team to date has been seeded into the College Football Playoff field with two losses, so Georgia can't afford to stumble at any point the rest of the way if it wants to claim one of those bids.
There's a lot of football to be played in the second half of the 2019 season, including a number of regular-season matchups between current playoff contenders. Georgia's stumble was a shocker in its own right, and a bit of a shock to a system that had been reliably impenetrable for top teams this year. Might it foreshadow other stunning results down the stretch?
2019 FEI Ratings (through Week 7)
The Fremeau Efficiency Index (FEI) is a college football rating system based on opponent-adjusted possession efficiency, representing the per possession scoring advantage a team would be expected to have on a neutral field against an average opponent. Unadjusted possession efficiency (PE) is calculated as a function of offensive, defensive, and special teams game splits. Schedule strength is represented by each team's average per possession opponent adjustment (OA). Opponent-adjusted offense ratings (OFEI), opponent-adjusted defense ratings (DFEI), and opponent-adjusted special teams ratings (SFEI) are calculated in a similar manner as overall FEI ratings. Team records against all FBS opponents (W-L) and 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.