Margins + Momentum in the Sweet 16
This post is a collaboration with Andy Cox of Crashing The Dance.
The kenpom crowd, by now, is well-versed in matters of teams' full-season efficiency ratings. Ohio State is the nation's No. 1 team in efficiency, due in large to its dominant defense, and Kentucky is No. 2 because of its dominant offense. If the Sweet 16 and Elite Eight were to play out according to big-picture efficiency projections, the Buckeyes and Wildcats would be in the Final Four along with Michigan State and Kansas.
But in forecasting the next rounds of the NCAA tournament, we should also be taking into account how well teams are playing over the past few weeks. Who's surging, who's stagnating, and who may have peaked too soon? Over at Crashing the Dance in 2009, Andy Cox devised a way to gauge teams' quality of play over the course of the season, using a stat he called Net Efficiency Margin, or NEM.
NEM is the difference between a team's efficiency margin in a game and the efficiency margin an "average" Division I team would be expected to have in that same setting. It allows us to track teams' quality of play against a baseline average, and when we map out running, five-game averages of their NEM over the course of a season, it gives us a real, visual sense of their efficiency-based momentum. The following graphs look at each region through the lens of NEM:
South: Kentucky coming down to earth?
Is Kentucky not as much of a juggernaut as it's been built up to be (including in this blog, at the start of the tourney)? This NEM chart casts a bit of doubt on the concept of the Wildcats as a Final Four lock:
Between games 20-26, from about Jan. 21-Feb. 11, UK took off into a NEM stratosphere beyond any of the remaining tourney teams, with an average margin above 50. But over the past few weeks John Calipari's team has come back to earth, to the extent that its NEM is actually in the same ballpark as Indiana's and Baylor's, and below that of fellow title contenders Ohio State and Michigan State. The Wildcats are still the favorite to reach New Orleans, but an upset at the hands of Indiana or Baylor is a realistic possibility.
West: Everybody's Surging
This bodes well for the quality of games in Phoenix: all four teams are on an upward NEM trend.
Louisville really was awful around the 18-19 game mark, but has surged in games 30-37 to the extent that it's the second-best NEM team in the region. Florida's sharp NEM spike could be misleading due to the fact that it feasted on a wounded Virginia team and was witness to Norfolk State's deep, post-Missouri crash in the Round of 32.
East: Big Ten Rules
Ohio State just took over No. 1 in the kenpom.com rankings because it's playing some of its best basketball of the season in March:
The East's other Big Ten team, Wisconsin, went on a similar NEM climb through February and March, and these numbers suggest that the Badgers and Buckeyes will meet in the Elite Eight.
Midwest: Beware of NC State!
Near the end of February, NC State was a middling team that had no business being mentioned as a Sweet 16, much less Final Four, contender. But that streaking red line indicates that the Wolfpack are the sleeper team of the whole tournament. Their most recent NEM numbers are better than Kansas' and as good as North Carolina's.
The Tar Heels' NEM were on the decline before they lost Kendall Marshall, which makes me worry about them even more. Is it possible that the NC State-Kansas game will decide who goes to the Final Four?