A buddy of mine from my days at Iowa State, who currently works as a mathematical statistician in D.C., enjoys mixing his passion for statistics with his passion for college football. This hobby has led to the creation of a statistical model, which he describes as “a predictive model for college football scoring margin that relies on validated, underlying assumptions of normality.”
Along those lines, he recently sent me a statistical analysis that compared the performance of the 2009 Cyclone football team to the performance of the 2008 squad. It reinforced what we all believed we knew about Iowa State in year one of the Rhoads era, and did so in an objective manner.
Here is his analysis:
During my free time I dabble with college football stats just for fun, being the stat junkie that I am. I’ve developed a pretty good model for prediction. I update it with new scores each week and recalculate power rankings. Then, at the end of the season, I determine a final ranking. This final ranking has a correlation of 0.80+ for the season’s outcomes, so it’s a good numerical measure of how strong that team was over the course of the year. It works much better than typical stats or rankings that talking heads might rattle off. Of course, the methodology is purely objective.
Here’s how Iowa State fared for 2009:
Power Rating = -2.7
Final Ranking = #71
The Power Rating is the expected score margin against an average team (i.e. a team with a PR = 0, which is close to the 60th ranked team out of 120). So ISU was about a field goal below an average team in 2009. A ranking of 71st doesn’t sound so hot, but then I looked at the previous year.
ISU in 2008:
Power Rating = -11.2
Final Ranking = #98
One word for Gene Chizik… ouch. Expected to lose to average Division I teams by 11 points. That shouldn’t happen to a BCS school. Rhoads’ first team was 8.5 points better PER GAME, which is a huge amount. Note that this doesn’t necessarily mean that their average score margin was 8.5 points better, because it doesn’t work that way. Things like opponent strength and opponents’ strength of schedule factor in to make it more robust.
I’m of the belief that any hint of improvement in the first year of a coaching change is a pretty good thing. For comparison, I’ve rated Rich Rod’s first season at Michigan (2008) as a 17.2 point drop in PR and a 59 spot drop in the rankings. That’s pretty awful, and a reminder of how bad things can go in the first season following a coaching change.
So in summary, and this is restating the obvious to anyone who watched ISU in the past two seasons, but Rhoads really did a heck of a job.
–J. McIllece, Bureau of Labor Statistics