Here's my sliver of pride in forecasting this year's March Madness tournament: I've got more teams in the Final Four than Bing Predicts.
Of course, that's cold consolation for one of the wackiest NCAA men's basketball tournaments I can remember. Bing and I both saw our brackets busted on second day, when Michigan State shockingly lost. (Thanks Middle Tennessee State.) Perennial favorite Kansas also bowed out last weekend, spelling doom for our brackets, as well as those of countless Jayhawks fans around the country.
So much for poring over analytics and parsing the reports of countless college basketball analysts, former players and buddies. It doesn't matter what your system is or how much computing power you have. The simple truth is bracketology isn't much more than a crapshoot. An informed crapshoot at best.
For a sense of how tough it is to pick the winners, consider this factoid: just 1,140 of the record-setting 13 million brackets submitted to ESPN's Tournament Challenge got the final four right. Yes, just 0.009 percent of the submissions chose North Carolina, Oklahoma, Villanova and Syracuse.
"That's pretty infinitesimal," said ESPN spokesman Kevin Ota.
Of course, Bing has done pretty well overall. The forecasting tool has picked 63 percent of the tourney winners overall, so far. While that's down from 73 percent last year, Bing's 2016 bracket slightly better than President Barack Obama's ESPN bracket, who has picked 60 percent of the winners.
Walter Sun, a top data scientist for Bing Predicts, said it may have to do with parity. Of the 32 games played in the first round, 13 ended in upsets, Microsoft said, adding that 18 of the tournament's 64 games were decided by a slim 5 points or less.
And, Bing, which had Kansas to win the title, still has North Carolina in the Final Four and now gives the Tar Heels a 45 percent chance of winning it all.
"We're still competitive," Sun says. "It's not over yet."