100 Days of ML — Day 6 — Machine Learning Built By The Gridiron or How Can ML Help Save College…



It’s that time of year where I get to disappointingly watch my alma mater (NC State) try to muster some wins against other Tier 2 schools while waiting for some combination of Oklahoma, USC, Ohio State, Clemson, Georgia, and Alabama fight for the right to be controversial.

I first start watching college football in 2004. It was an easy way to make friends for a guy who was painfully awkward around people and who’s social circle included the fame-desperate, but way post-college, comedians in Raleigh at Goodnight’s Comedy Club.

That year’s controversy was the disclusion of the Utah Utes from the (then) BCS bowl games. It took a while, but college football fans finally got their playoff. Much like the expansion of the NCAA tourney field, it didn’t diminish controversy, the controversy multiplied. It was controversy steroids.

Fast forward to last year’s snub of the UCF Golden Knights wherein we got the now-justified national champions in Alabama and the raucous roar of both sides continues to echo louder.

The core issue with college football is that it is an economic, data-driven, and political nightmare. Economically, there used to be major bowl spots for six teams, then eight, then ten. Now we do the weird New Year’s 6 and the playoff thing. Anyway, there’s a lot of teams vying for little room and the FBS schools refuse to have a 16-team playoff to ease the pressure. On the data side, different teams have different schedules. No one really plays each other, so, on the surface, it seemed math had no place in determining who should qualify for these final games. Without a clear-cut methodology, this meant convincing by a lot coaches to fans and TV networks. Nothing sullied the sport of college football more than turning what should be the most important games into a popularity contest.

But where we used to have all these statistics that we didn’t know what to do with, now we have machine learning to help sort through it all.

There are several approaches you can take: a points system based on scoring, yardage, and yard prevention, a ratings draw to keep the networks happy, or Notre Dame always makes it, because no matter what universe we’re in, Notre Dame will never have to play by the same rules as any other college football team ever.

I think the output I’d choose is “game interest level” wherein the match-ups provide the games that are most likely to be close based on each team’s strengths.

There are many applications for ML in the college football space and I hope to hear about them soon.

Jimmy Murray is a Florida based comedian who studied Marketing and Film before finding himself homeless. Resourceful, he taught himself coding, which led to a ton of opportunities in many fields, the most recent of which is coding away his podcast editing. His entrepreneurial skills and love of automation have led to a sheer love of all things related to AI.

#100DaysOfML

#ArtificialIntelligence

#MachineLearning

#DeepLearning

Source: Deep Learning on Medium