Identifying this season's early Giant Killers

Alec Peters' Valparaiso team just missed the NCAA tournament last season while Reggie Upshaw and Middle Tennessee stunned Michigan State in the first round. Getty Images

Welcome back to Giant Killers, our annual metrics-based forecast of NCAA Tournament upsets.

Since 2006, when we started searching for Davids who could take down springtime Goliaths, we have built and refined a statistical model to identify the characteristics shared by springtime Cinderellas, as well as the traits common to heavy favorites who lose at the Big Dance.

And every year we keep launching our analysis earlier in the season, partly because there's more data available, but also because we know why you read Giant Killers. Yes, of course, to do well in your March Madness pools -- and we're proud that last year, six of the eight NCAA Tournament games we projected to have the highest upset chances were actually won by underdogs, and that Syracuse, a 10-seed who made the Final Four, was our highest-rated Killer. But even more, we know you want some advance knowledge about the teams and players who will define the season's key matchups, and who will create its most memorable surprises on the way to crowning a champion. Stick with us, and come tournament time, when you hear about how many Australians play for Saint Mary's or why opponents make special preparations to face VCU's Doug Brooks even though he plays barely 20 minutes a game, you'll smile knowingly.

Our definitions are simple: A Giant is any team that plays an NCAA tournament opponent seeded at least five spots lower in any round. And a Giant Killer is any team that beats a Giant. We have found that successful Killers tend to use high-risk/high-reward strategies. Typically, that means they boost the value of their shots by attempting a lot of 3s and/or create extra possessions by grabbing offensive rebounds or forcing turnovers. Conversely, Giants who win usually choke off underdogs' chances to take risks. For details about how our statistical model works, you can check this out, and we wrote here about how different types of Giants and Killers play different styles.

Fewer than 10 games into this season, mid-major teams are scattered across projected brackets. We can see, for example, that Gonzaga will probably be a Giant, and the best teams from the Atlantic 10 are likely to end up with mid-level seeds. Meanwhile, some preseason conference favorites, like Princeton (Ivy League) and Belmont (Ohio Valley), are off to rocky starts. So in this first edition of Giant Killers for 2016-17, we will look at teams that have excellent underdog characteristics, are on track to earn seeds of 11 or below (making them potential Killers in the first round) and are already outperforming early expectations.