March Madness is always rough on favorites. In your typical men’s college basketball season, the statistical favorite goes into the NCAA tourney with only about a 20 percent to 25 percent probability of winning the championship. In other words, even the “best” team is three or four times more likely to lose early than it is to win it all. We can be vaguely confident that one of the top handful of teams will emerge from March’s wreckage unscathed, but that’s about it. Over the past seven seasons, the eventual champ was (on average) the tourney’s seventh- or eighth-most-likely winner beforehand, according to simulations using Ken Pomeroy’s ratings.
And this year has the potential for a lot more anarchy than usual. Michigan State is the consensus top team according to a variety of predictive metrics, but that’s not unanimous — there are some solid indicators pointing to Villanova, North Carolina and surging Kansas as well.14 And if the Spartans truly are the best, they’re one of the weakest top-ranked squads in recent memory. Since Pomeroy began tracking team ratings in the 2001-02 season, only one No. 1 team — the 2005-06 Duke Blue Devils — owned a lower pre-tournament power rating than Michigan State has now. Plus, Sparty might not even earn a No. 1 seed; only 11 of the 77 bracket prognosticators aggregated by Bracket Matrix see the team headlining a region in the NCAA Tournament.
This isn’t meant to pick on the Spartans (sorry, Nate!), but it does illustrate that if even the nation’s strongest team has a number of factors limiting its championship odds, it’s going to be an unusually wide-open year for the tournament.
To see exactly how much this year figures to depart from the norm, I plugged Pomeroy’s ratings into ESPN’s latest Bracketology projection, calculating each team’s odds of winning the tournament if Selection Sunday goes as Joe Lunardi predicts. Virginia checked in first with a 12 percent probability of winning — significantly lower than the 20 percent to 25 percent chance held by the typical pre-tourney favorite — followed by Villanova at 11.7 percent and Michigan State at 10.4 percent. I then re-simulated the bracket after assigning each team the typical strength connoted by its Pomeroy ranking — i.e., how much better would the top teams’ odds be if they were as strong as in an ordinary season?
If Michigan State were as strong as the average top-rated team, it would win this year’s tournament a shade less than 17 percent of the time, instead of little more than 10 percent. The same goes for Villanova, a likely No. 1 seed currently ranked second by Pomeroy; instead of being expected to win nearly 18 percent of the time, the Wildcats’ low rating limits them to odds below 12 percent. In fact, the collective tournament win probability for Pomeroy’s top 3 teams is about 14 percentage points lower than it would be in a typical season, simply because 2016 has so much parity — the best teams aren’t as good as top teams normally are.
Because the NCAA Tournament is a zero-sum game, the rest of the field gains from this lack of top-heaviness. Although pretty much every team ranked No. 4 to No. 25 by Pomeroy benefits some (teams ranked below that don’t really have much of a chance either way), the biggest beneficiaries are positioned in the back half of the top 10 and into the teens:
All of this means more teams will have more of a chance this year than in just about any other season in memory. Get ready for one Mad March.