There are so many ways ESPN’s influence over the sport of college football depresses me. Here’s another one, straight from MIT’s eighth Sloan Sports Analytics Conference.
The panel, like the other larger productions at Sloan, is being held in a spacious ballroom at the Hynes Convention Center in Boston. Its organizing idea so far has been a particular cogent observation by Oliver, which is that “best” and “most deserving” are two completely different things when it comes to ranking teams for inclusion in a playoff. Addressing both separately, instead of trying to cram them into one metric, is a perfectly sensible thing to do, and for this, ESPN has brought all its considerably weaponry to bear. Oliver and other senior analytics staff have spent the last two years immersed in college football, which has lagged behind other sports in statistical sophistication…
… But we haven’t really talked much about CHAMP and FPI during the football playoff panel. It’s been more focused on how the individual SOS and dominance numbers are good tools for committee members to look at, if they want. It’s enough to make you wonder why ESPN would even bother with the catch-alls. Then, suddenly, Rece Davis, Mark May, and Lou Holtz are bellowing down at the audience in Ballroom A at the Hynes from two giant screens, projected on either side of the room, howling about who the best team in the country is. “Alabama,” says a grinning Holtz. “They’re the best team in the country, they don’t have the best record—that’s the problem.”
… It’s an open secret that the ESPN analytics team generates far more data than it makes public, and certainly more than make it onto TV. “We’re still a TV company first,” many analysts will tell you in private moments, when you ask about stuff that only lives on “dot com.” This means that anything that isn’t generated for a specific story will get dumped into what’s called an NST (notes, stats, trends) pack, and sent out as notes to anchors. If you really like an item, you might phrase it in 140 characters or less, to make it tweetable, though those often go unclaimed as well. You learn whom to pitch to (Kirk Herbstreit is great; Jay Bilas is a sponge) and whom to avoid (maybe stay away from Corso). ..
… That isn’t evil; it’s just good sense. ESPN is not a statistics-generating non-profit put on Earth to further our understanding of sports. But it is the tension at the heart of the entire conference: TV personalities using numbers and concepts with their edges sanded down, a platform (panel, conference, network) that often insists analytics are dichotomous with all other forms of knowledge about sports…
It’s not enough to have various talking heads spinning a narrative your way. ESPN arms them with serious looking metrics that ESPN in its infinite wisdom has concocted to make them sound more authoritative to the listening public. And to, who knows, maybe even a selection committee or two. That’s some seriously pernicious power there.