As most readers on this site know, the Fremeau Efficiency Index is one of the best metrics for college football. Luckily, we have expanded FEI rankings for the last three seasons available.
Success for a football team requires many different factors, most of which have been documented on this website. Lately, the focus has been on an improvement of the defense. As Bud and others have pointed out, even if we have a giant leap forward in our defensive efficiency, the brutality of our schedule will likely hold us back.
What I wanted to do was look at the relationship between strength of schedule, offensive and defensive efficiencies, and field position. To do this, I looked at the historic FEI rankings available. Unfortunately, these FEI rankings are end of season rankings and are not the week 14 rankings that would only include the 12 regular season games. Since expanded week 14 rankings are not available, end of the season ones will have to suffice.
I ran a multiple regression to predict the season win totals using four dependent variables: SOS rank, offensive efficiency rank, defensive efficiency rank, and field position rank.
Here are the results of the regression analysis.
We end up with a pretty good R square value with an average error of around 1.5 games.
A couple of things to take away here.
1) The impact your offense and defense efficiencies are very similar, but in the end, the offense is the single largest contributor to your season wins total.
2) Strength of Schedule is very important, but not the end all be all for improving seasons win total. This is because strength of schedule is offset by a disproportional amount due to offense and defense efficiencies.
Florida State should win 7.66 games next season assuming a couple of things. If we assume our offense remains a top 10 offense, our field position regresses back to the three year average (rank 25), and our defense improves to the 50th percentile rank. Everyone who is predicting 9 and 10 win seasons are simply overestimating FSU’s defensive turnaround and underestimating the difficulty of this schedule.
But wait, there is one little caveat. This is against FBS competition. Florida State only plays 11 regular season FBS schools this upcoming season. The 7.66 win total does not include the game against Samford. If we assume that FSU wins this game 99.9% of the time, FSU’s actual predicted wins becomes 8.65.The average number of FBS teams any team plays in a year in this data set is 11.94.
What do you think?