Column: Breaking down the Sweet 16

But after I spent so much time putting that thing together, I thought I’d try to find some common characteristics between the teams left standing and maybe try to improve my woeful prognostication abilities (if you had Utah State in the Elite Eight, raise your hand. My hand is raised high and shamefully).

For each of the categories I measured, I calculated the mean and standard deviation of all the teams in the tournament to assign each team a z-score of how many standard deviations above or below the mean they were. If statistics isn’t your bag, the important thing to know is that in a normal distribution, 68 percent of data points are within one standard deviation of the mean and 95 percent of data points are within two standard deviations.

So, if Ohio State has a z-score of 2.00 for offensive efficiency (they actually had a 1.99), that means the team is approximately in the top 2.5 percent of tournament teams in that category. Obviously, not all these teams play equal schedules, and pure averaging isn’t perfect for rate stats, and I would never suggest that anybody but me should think this methodology actually indicates who is going to win.

Again, I picked Utah State in the Elite Eight. That said, I think it still provides a general framework of at least the relative strengths and weaknesses of each team.

By definition, the average z-score is zero in all categories. But when we reduce the original field of 68 to its current field of 16, the remaining teams excel in a few metrics more than others. Most of the z-score averages for remaining teams are above zero, which would make sense given that these teams ought on average to be better than the ones they eliminated, but not that far above. In three categories, however, the average z-score for Sweet 16 teams is 0.63 or higher, meaning that the average team still in the field of 16 was roughly in the top quartile in this category when all the teams were considered. Those three categories? Offensive efficiency, turnover percentage and opponent free throw rate.

While it may seem obvious, this means that in this particular tournament, the most important indicators for success have been making the most of possessions by scoring efficiently and avoiding turnovers and preventing easy free-throw points. This bodes well for teams like Wisconsin, which turns the ball over far less frequently than any team remaining in the field, and especially for overall top seed Ohio State, who ranks at least 1.4 standard deviations above the tournament mean in all three of these categories. The Buckeyes have looked like juggernauts so far in dismantling The University of Texas at San Antonio and George Mason.

The best example in the category of “Team that wasn’t highly seeded but ranks well on these three metrics” is No. 11-seeded Marquette, which has knocked off Xavier and Syracuse (both of which were slightly above average in offensive efficiency and turnover percentage) en route to a Sweet 16 matchup with the No. 2-seeded North Carolina Tar Heels. During the regular season, Marquette struggled on defense in virtually all respects, but its z-scores in the three metrics above are all above 0.50 and their defense has stepped up to support the offense.

Unfortunately, these numbers suggest that the performance of some of the tournament’s darlings has been a fortunate aberration that may come crashing down. No. 10-seed Florida State is by far the worst team remaining in both offensive efficiency and turnover percentage, and has survived not only because of its crushing defense but because of its shockingly efficient three-point shooting. Then again, their next opponents are the Rams of Virginia Commonwealth University, who are one of the worst teams remaining at defending the three. VCU has also been playing much better than its regular season would indicate, shutting down Georgetown and Purdue with its previously porous defense while absolutely raining J’s.

What’s certain is that I have no idea what to expect from the VCU-Florida State regional final. The game pits the best defensive team remaining, Florida State, against the worst defensive team, VCU. During the regular season, Florida State gave up 89 points per 100 possessions while VCU gave up 104. But VCU is one of the top teams at forcing turnovers, while Florida State is the worst at protecting the ball. Florida State turned the ball over on 23 percent of regular season possessions (no other Sweet 16 team had a percentage above 19.1).

In contrast, VCU forced turnovers on 22.5 percent of possessions, second best of the teams remaining and good for a z-score of 0.98. I have literally no idea what will happen in this game, and I am so excited. I’m also trying not to think about the fact that the winner will almost certainly be destroyed by Kansas.

But hey, Kansas could lose to Richmond in a matchup of teams who have tournament z-scores above 1.00 in both three-point percentage and opponents’ three-point percentage.

That’s why they call it March Madness instead of March Predictability. Well, that and the fact that it’s in March.