Picking March winners: An inexact science

I begin to realize that things are getting out of hand when I raise my voice at my housemates.

"Get outta here! If Samford beats Syracuse, I'll transfer there. 'Cuse is almost all seniors."

And while I feel ready to stake my life savings, or various parts of my anatomy, that the Orangemen will have no problem with the Bulldogs, in the next breath I argue vehemently that Winthrop - a Big South Conference heavyweight - will knock off heavily-favored Oklahoma in the first round.

From selection Sunday until the games commence on Thursday, my die-hard housemates and I have nightly meetings to discuss these pressing issues while we predict the winners of the NCAA Tournament pools we enter. The gatherings last a few hours at a time and take place late at night, when most sane people have chosen to sleep rather than decide if Butler's current 15-game winning streak gives them any sort of momentum against Florida.

We do the same thing every year, breaking down the matchups until we know more about a team's strengths and weaknesses than they do. I stare at some of the games until I'm cross-eyed, thinking I'll find the key to the game if I just look hard enough at the statistics.

"Aha! Pepe Sanchez's assist-to-turnover ratio is .045 higher than Jason Williams'. And you gotta be strong at the point guard position to win it all in today's game, so Temple it is. . ."

I'd like to think my hard work pays off, but every year there is incontrovertible evidence to the contrary.

Any veteran of NCAA pools knows what I'm talking about. While you watch your carefully chosen Final Four pick get upset in the first game of the first day, some office pool lightweight gets 30 of the 32 first-round games correct, including foreseeing the first-ever defeat of a one seed by a 16 seed.

The worst part is that these people often make their picks based on a cute gimmick that has no business determining the outcome of a college basketball game, yet still defies all laws of probability with its deadeye accuracy. Like who would win in a fight between the mascots. I've actually known people who use that as a basis for picking. And they've done better than me.

Why, just two years ago, I was in a pool in which the second-place finisher picked all of his winners by selecting the team with the least number of letters in its name. Needless to say, he was the only entrant to pick upstart Utah in the championship game (although his champion, TCU, didn't make it out of the first round).

It's rather humiliating. But I'm to blame for my occasional lack of success as well, because I never learn from the mistakes I make in tourneys past. Bracket in hand and eyes squinting with some misguided analysis, I say the same things every year, setting the same traps for myself.

Despite the individual bracket makeup that varies year by year, my picks always have haunting similarities:

n The Big Ten always goes undefeated in first round games, with an unlikely handful advancing to the Sweet Sixteen. My pro-Big Ten bias has, at times, convinced me that an all-Big Ten Final Four is not out of the question. It's a little out of control.

n UCLA never bows out before the Sweet Sixteen. For some reason, I always feel that I am in on a huge secret with this constantly athletic and flashy Pac-10 team. Nobody ever sees them play because they are on the West Coast - myself included, as I discover when they underachieve and lose in the first round like last year.

n Arizona somehow overcomes its uncanny affinity for first-round upsets and makes it at least two rounds further than anyone else in the pool has them going. I rode my hometown boys all the way in 1997, and I take it personally when anyone brings up the 50 straight first-round exits they suffered before that. I pick them out of principle.

And so on. I can study statistics all I want, but my picks always end up influenced by personal biases I have for or against teams.

But that's what makes it fun. With my luck, the year I give in and pick UCLA to lose in the first round is the year the Bruins will play the way I always think they will.

Besides, UCLA has less letters than most teams, so they're a lock to reach the Final Four.

- Andy Latack finished seventh out of over 300,000 people in espn.com's Tournament Challenge his freshman year. He's always looked for a way to get that fact in print. E-mail him at

latack@umich.edu.