Filling out an NCAA tournament bracket with accurate predictions is no easy task. That only got tougher this year, thanks to a significant tweak in the way the selection committee determined the matchups.
In theory, the brackets are designed to give better teams easier paths to a title. The rankings, while subjective, can be used to reward teams that performed the best prior to the tournament, and make things tougher for weaker teams.
The way that was accomplished in the past was with a process known as the S-curve. As Deadspin's Barry Petchesky explains, that design would pit the highest-rated one seed against the lowest-rated two seed, and so on down the bracket. For example, Louisville was named the top overall seed in the tournament, followed by fellow number one seeds Kansas, Indiana, and Gonzaga in that order. Under the S-curve, Louisville would have been placed in a bracket with the weakest of the four number-two seeds. Gonzaga would have matched up with the highest-ranked number-two seed.
Yet in an interview with ESPN, NCAA tournament committee chair Mike Bobinski revealed that the committee didn't use the S-curve at all this year. Instead, the committee gave more consideration to other factors like geographical location, in-season matchups, and conference rivalries.
So how was this year's field affected by the committee ignoring the S-curve?
Louisville looks to have gotten the worst draw. Despite being the top overall seed, the Cardinals got stuck in the same region as Duke, a team that spent part of the season atop the AP rankings, and that, in the committee's own estimation, was the second-best number two seed. Meanwhile Gonzaga, the weakest number one seed, had the luxury of being placed with the weakest number two seed, Ohio State.
Before the committee unveiled this year's bracket pairings, Las Vegas odds pegged Louisville as a 3-1 title favorite. After the bracket came out, those odds dropped to 9-2.
Respected college hoops analyst Ken Pomeroy, writing in Slate, noted that individual seeding isn't really important in determining a team's odds of winning it all. What's far more important, he explains, is the overall strength of a given region. Applying a statistical odds-making tool to last year's field, Pomeroy found that eventual champion Kentucky's title chances could have ranged from a low of 22.8 percent to a high of 37.5 percent depending on which region they were assigned to, with those fluctuations tied directly to the changing competition.
Kentucky, with help from the S-curve, entered the tournament as a 2-1 favorite to win it all.
"A team's chances of winning the tournament will be restricted most by its potential matchups against the great teams in the field," Pomeroy wrote. "Most of the time the great teams will be found in the Final Four — if they're even there at all — and the chance of playing one of them doesn't differ much whether a team is a No. 1 seed or a 3."
After an unpredictable season of upsets in which parity reigned supreme, the tournament seems primed for even more surprising outcomes. The seeding changes will only exacerbate that trend, leaving bracket builders with more uncertainty and more tough choices.