homegame

Bracket Crepe Hanging

The most insightful comment on handicapping the NCAA Basketball Tournament was delivered a few days ago by Florida coach Billy Donovan.

"There's a reason," Donahue said, "why some guy put up a billion dollars for anyone who could predict all the games."

First, maximum kudos for Donovan for either not knowing or pretending not to know who Warren Buffett is. Secondary kudos for dropping a wet blanket on the annual frenzy in which innumerable sports fans forecast the championship of a sport that unless Nielsen Co. is a big fat liar, relatively few of them ever watch.

Inspired by Donovan's comments, I've spent the past few days scouring the how-to bracket guides and actual predictions offered by mass media outlets and their journalist employees on the tournament, and I've found them all to be most illuminating. Alas, the light they shine isn't on college basketball, it's on the people and institutions making the predictions.

It was uncanny how the analysis offered and forecasting techniques discussed were the exact same approach used by the media outlets where I saw them to discuss, analyze and forecast EVERY topic they cover. These articles weren't predictions, they were expressions of a methodological worldview.

The tournament articles on Nate Silver's new 538 Website were full of math, some with work shown, and percentages. The article in today's Wall Street Journal assumed the only reason anyone who make out a bracket was to gamble, and was therefore straightforward value-based investment analysis. Esquire's Website's guide to bracket-filling was accompanied by photos of a mostly undressed starlet. And so on.

Best of all was ESPN's. Every single analyst on their post Selection show picked Michigan State to win it all, on the "best performance in a conference tournament by the team of one of the supercoaches who we slobber on and goes on our shows a lot" theory. It would've caused a real crisis of conscience for the gang if Duke had beaten Virginia in the ACC tournament. Nice of Coach K to spare his guys that anxiety.

In fairness to the sycophants of Bristol, math, value analysis and the starlet theory all like Michigan State, too. It is the general consensus of the predicting community that the Selection Committee must've been asleep since the Super Bowl and thus made prohibitive favorites Michigan State and Louisville number four seeds in the grossest miscarriage of justice since Dred Scott.

It's the general opinion that seeding was erratic, to put it kindly. The consensus is so strong in this regard that as part of its buy low strategy, the Journal recommended picking Wichita State, the most disrespected one seed since maybe Rutgers in 1976, to make the Final Four, as no one else in your pool will.

The consensus could be right. Louisville and Michigan State are splendid teams.  I, however, have come to distrust the power conference tournaments as bracket guides. Their results are not reliable, because for many of their participants, the conference tourney is nothing more than an historic and prestigious set of controlled scrimmages. That's not much of a preview as to how a team will react when it's one and done for keeps.

I am also leery of predictions made on the grounds the Selection Committee didn't know what it was doing. Like all committees in the history of mankind, it's hardly infallible. But it is always composed of men and women who know ever so much more about what's happened in a college basketball season than just about everybody else, Digger Phelps, Nate Silver and me most definitely included. They have reasons for their seedings, and those reasons should be weighted, or at least guessed at, before turning in an office pool entry.

Working at home as I do, I got no office pool. What I got today is a Final Four guess, based on observation, some years of experience, and not inconsiderable guesswork. That methodology won't get me a fancy Website, but it's proven about as sound as any other method over the Marchs of yesteryear.

I got Arizona, Kansas, Villanova and (sigh) Louisville. I'm pretty sure all four won't be wrong.

posted by Michael Gee @ 2:42 PM