A look inside the numbers to try to narrow NCAA bracket
At this time of year, everyone’s in search of a system—some way of simplifying the brackets and making it a little easier to pick one champion out of all those contenders.
You could go with the most experienced coaches, or perhaps the teams playing the best defense. Or you could fill out the whole bracket based on team mascots .
Fortunately, thanks to the proliferation of advanced stats, there is another way. Basketball junkies are familiar with Ken Pomeroy’s work at kenpom.com , and there’s similar data available at barttorvik.com . In fact, the latter site is especially helpful because of the filters it offers for archived stats. For each year since 2008, it’s easy to go back and see what type of profile the eventual champion had before the NCAA Tournament.
A quick look at the past 11 national champions reveals some interesting information. It’s possible for a team to win the title while being deficient in some areas. For example, Villanova’s 2016 title team was just 210th in offensive rebounding percentage heading into the tournament.
However, there are a handful of important stats in which each champion since 2008 was ranked solidly in the top half of Division I. Let’s take a look at those—and see if they can help us narrow down this field of 68.
The main measure of overall team strength at barttorvik.com is called T-Rank. Each of the past 11 national champs was in the top 23 in T-Rank, based on their performance prior to the NCAA Tournament. The 2014 Connecticut team, which won the title as a No. 7 seed, is the one that was No. 23. Plenty of this season’s longshots can look to those Huskies for inspiration.
So let’s start our exercise by listing the current top 23 in T-Rank—our initial list of national title contenders: Virginia, Gonzaga, Duke, Michigan, North Carolina, Michigan State, Texas Tech, Kentucky, Tennessee, Purdue, Virginia Tech, Houston, Auburn, Wisconsin, Louisville, Florida State, Kansas State, Iowa State, LSU, Kansas, Mississippi State, Clemson and Maryland.
Cross Clemson off the list for failing to make the tournament.
Stat No. 2: Adjusted Offensive Efficiency
This is a measure of how many points a team scores per 100 possessions—adjusted to account for opponent strength and venue. The national champs in our sample were all ranked in the top 55 in adjusted offensive efficiency. (Yup, 2014 UConn was the team at No. 55.)
A couple teams in our 2019 group don’t make the cut on this one: Say goodbye to Wisconsin (70th) and Kansas State (87th).
Stat No. 3: Adjusted Defensive Efficiency
The same stat as the last one, but for defense. Each national champ was in the top 40, with 2011 UConn (40th) and 2009 North Carolina (38th) pressing their luck the most.
So we eliminate anyone not in the top 40 this season: That means you, Auburn (44th), Iowa State (63rd), LSU (64th) and Mississippi State (54th).
Stat No. 4: Defensive eFG%
Effective field goal percentage (eFG is a measure of field goal percentage in which 3-pointers are given extra weight. You want to hold your opponents to a poor eFG%, obviously. Each of the last 11 national champions entered the tournament in the top 102 in defensive eFG%. Duke in 2015 was the worst in this regard. The Blue Devils were in a three-way tie for 100th in defensive eFG% before the NCAAs.
Purdue (117th) drops off our contenders list this season for being just a bit substandard in this area .
Stat No. 5: TO%
Now things are really going to get interesting. This stat—turnover percentage—measures how often a team turns the ball over, while accounting for the team’s total number of possessions. The shakiest recent champion in this regard was UConn in 2014, which was one of 10 teams tied for 116th in the category before the tournament began. So we’ll say any prospective champion this year should be in the top 125 in this stat.
That eliminates: Michigan State (189th), Texas Tech (151st), Kentucky (172nd), Florida State (222nd), Kansas (191st) and Maryland (258th).
Duke (121st) and Louisville (124th) barely make it past the threshold, but they remain safe for now .
Stat No. 6: Defensive Free Throw Rate
This is a measure of how often a team puts its opponents on the line. The calculation is: (Free Throw Attempts Allowed)/(Field Goal Attempts Allowed). It’s certainly nice to keep opponents off the free throw line, but in 2013, Louisville was able to win the tournament after entering as the 121st-best team in the country in that regard. So anyone in the top 121 can stay on the list.
So long, Tennessee (223rd), Houston (284th) and Louisville (141st).
So, if you’ve been keeping track, you’ll see that we’re down to the following six teams: Virginia, Gonzaga, Duke, Michigan, North Carolina and—if you’re looking for a longshot—Virginia Tech. Could someone else win the title? Of course. You never know when another outlier like that 2014 UConn team might come along. But the six teams in this group appear to have few red flags, statistically.
The question now is how to pick one champion from that group of six. At a certain point, even math has its limits.
Time to start researching their mascots.