Battling anecdotal analytics in college football

Every kick has a probability of going wide right. Advanced
analytics will use dozens of variables to calculate that probability
. . . and will tell you how to avoid kicking wide right a second time.
Anecdotal analytics will tell you a silly story based on a single variable.

For too long college football has been plagued by a dependence on anecdotal analytics. The human brain can only process so much information. To try to make the most of its limited hard drive space our brains automatically filter and bundle select bits of data. But the conclusions we reach are also limited, biased by our prejudices and experiences. The more complex the system we are trying to understand the more unreliable our conclusions will be. We simply lack the capacity to manage an event as complex as a college football season, but for generations we depended on this approach for understanding and strategizing in college football.

We now have the ability to break free from anecdotal anlaytics. My brain can only recall a couple hundred plays at any one time, but my computer remembers 2.5 million plays and needs only a few seconds to tell me how many of those were run plays on 3rd and 5 in the 4th quarter. No longer do we need to punt on 4th and 1 at midfield because that's what everyone else does. We don't have to rely on silly mantras like "defense wins championships" or "you have to establish the run". We can quantify performance, attach probabilities to outcomes, and consider all available data and all possible outcomes and not just those that rise up first in our memories. We can correlate defensive performance and win probabilities in high leverage games, or measure the effective point contribution of a running play on subsequent pass plays. We can make football smarter!

But anecdotal analytics will not go gently into that good night. Its steely talons have a firm hold on the brains of coaches, players, and fans (many of whom don't read books or do math). But you can help fight this ignorance and help the search for a cure to anecdotal analytics in college football. Make a donation, follow an ad link, tell others about the site, and know that you are fighting the good fight, and maybe, just maybe, preventing a coach from reducing his team's win probability by kicking that long field goal on 4th and short.

But seriously, if you enjoy the site I appreciate anything you can do to help build the brand or offset the cost of maintaining the site.

Scott Albrecht