Models and Bottles! Bowl Selection Edition

For those of you familiar with this feature from FBT, no explanation is needed. However, if you're just stumbling on to this for the first time, a little background.

I'm a stats and data analyst/researcher by trade, which combined with my unhealthy obsession with college football forces me down the rabbit hole of using statistical and probabilistic methods to analyze the sport from a perspective that makes more sense to me. I'm well within my comfort zone scouring through databases and building statistical models, so it only makes sense for me to bring some of that background to look at college football from a nerd angle (or was it the other way around: college football stats nerdery led me to a career in data analytics? The line is blurry).

Anyway, prior to this college football season, I built a number of computer models, each with a different purpose. There was a model projecting the Gophers' likely win total, another model to predict the probability of victory and another to project the points scored by both teams. If I were to re-simulate the entire season based upon up-to-date data, it'd look like this:

If you simulate the season using the above listed probabilities, the Gophers would reach at least 5 FBS wins (assuming a win over an FCS squad, which of course hasn't always been a good assumption for Minnesota in recent years) to reach a bowl approximately 39% of the time, or 2/5 for those who like fractions.

The mode -or win total that occurred most frequently- was 4 FBS wins. The average over all simulations? 4.2 FBS wins.

As I said yesterday, the Gophers did exactly what they needed to push past the probabilistic tipping point and notch that all important 6th win. They did so by capitalizing on early season momentum and stealing a game against Syracuse, who happened to be struggling through a brutal non-conference schedule. The Orange rebounded in Big East conference play, while the Gophers predictably struggled against some of their tougher Big Ten brethren; the math model at the time was much, much closer between Syracuse and Minnesota.

But today isn't as much about looking back as it is looking ahead, so with that, your projections against the likely slate of foes Minnesota will face in their bowl!

Models

Iowa State

Gopher Win Probability: 7.4%
Score Prediction: Gophers 14.6, ISU 32.2.

Baylor

Gopher Win Probability: 5.2%
Score Prediction: Gophers 19.4, Baylor 39.7.

Bottles

After plugging away at some math models in a rather dry mental exercise, I like to have a little fun and share an Internet libation or two with y'all, typically themed after the opponent, town or situation for that particular week.

Well, since Minnesota won't play for another month, I'd rather us toast to a bowl season that will include our beloved Golden Gophers. Appropriate bubbly: White Star.

Why appropriate? Well, it's Moët & Chandon, bros who also make something called Dom P, which I've heard is pretty good. These guys know their champagne. Still, White Star is fairly overrated, as bubbly goes, much like limping into a bowl game with a 2-6 conference record.

Still, it's important to remember that it's champagne and not some other swill, or a California sparkling wine. Things worth toasting get the real stuff, like a team getting back to bowl eligibility after two miserable seasons. Cheers!