Thursday, March 19, 2009

Pre-Loading Joe Lunardi is a Bad Idea

Make no mistake, I've got no problem with what Joe Lunardi does. Unlike Guys Who Really Played the Game, I can admire the skills it takes to compress the metric tons of stats and psychology that make up the NCAA Selection Committee and come up with reasonable projections.

So I'm with Will of Deadspin, when he says, "In the world of baseball, Nate Silver is lionized for using statistical analysis and past results to predict the future. In the world of college basketball, Joe Lunardi is called a pencil dick for being a desk jockey with the temerity not to include Maryland and Providence...Joe Lunardi has a much better idea of who's going to make the tournament than anyone else you'll see on ESPN's set this week."

Here's my problem--Will wrote that on March 10th, when "Bracketology" is hitting Prime Time, and does get the notice (and ire) of fans and ESPN talking heads.  Bracketology, starring Joe Lunardi, has been premiering earlier and earlier.  I reached my particular breaking point when a co-worker sent me a January 12 projection.  January 12th.  Again, I've got no problem with what Lunardi does, I just think that projections that early are worthless.  And I'd like to prove it.

I don't want to belabor the point, so I'll only deal with the Top 5 seeds, without regard to region.  In fairness to Lunardi, I'll show his most recent projection before the seeds were announced, just to show how good he is, when he bases his projections on actual conference play.  But, and this is telling, I will bold teams that didn't make the NCAA Tournament at all.

Actual #1's:  Pitt, UConn, UNC, Louisville
JL 3/15:  Pitt, Memphis, UNC, Louisville  (3-4)
JL  1/12:  Pitt, UConn, Duke, Wake Forest (2-4)

Actual #2's:  Duke, Memphis, Oklahoma, Michigan St
JL 3/15:  Duke, UConn, Oklahoma, Michigan St  (3-4)
JL 1/12:  Clemson, UNC, Syracuse, Oklahoma  (1-4)

Actual #3's:  Kansas, Missouri, Villanova, Syracuse
JL 3/15:  Wake Forest, Kansas, Villanova, Missouri (2-4)
JL 1/12:  Texas, Louisville, Michigan State, Georgetown (0-4)

Actual #4's:  Wake Forest, Washington, Xavier, Gonzaga
JL 3/15:  Syracuse, Purdue, Washington, Gonzaga (2-4)
JL 1/12:  Xavier, UCLA, Notre Dame, Marquette (1-4)

Actual #5's:  Utah, Purdue, Florida State, Illinois
JL 3/15:  Clemson, Florida State, Xavier, UCLA (1-4)
JL 1/12:  California, Gonzaga, Memphis, Arizona State (0-4)

So you see my point here, yeah?  Predictions that happen before the big tournaments are finished are difficult enough.  Asking for predictions in January, even from someone who really, really knows his shit is a total waste of fucking time. Notre Dame is a real killer here--less than 4 weeks later, I was predicting they'd be playing in the NIT.



No comments: