Well, that didn't work out so well. For the uninitiated, R squared is a statistical term that outlines the variation in the dependent variable (NCAA Tournament wins) explained by the independent variable (in-conference efficiency margin). Something that is perfectly correlated would have an R squared value of 1. In this instance, a little more than 22% of the variation in NCAA Tournament wins (by power conference teams) is explained by their in-conference efficiency margin. It does show there is a correlation, but its not very strong. Oh well, end of post. Or is it?
How well does in-conference efficiency margin predict a team's eventual seed?
The answer, is much better. Nearly half (49%) the variation in NCAA Tournament seeding among power conference teams is explained by in-conference efficiency margin. With that being the case, we can use the resulting trend line from this relationship to retroactively predict each power conference team's seed and see which teams were over or under seeded. We might expect those teams that were significantly over or under seeded to over or underperform in the NCAA Tournament. Is that true? Let's start by looking at teams that were under seeded by at least four spots.
That's a pretty strong track record. Despite being seeded worse than their first round opponents, these teams posted a cumulative NCAA Tournament record of 9-9. They went a collective 5-4 in the first round, with two teams making the Sweet 16 and one making the Final Four. Also not included in these numbers is that three of these teams needed to win First Four games to even make the main field (Notre Dame, Pitt, and UCLA).
How about the teams that were over seeded by at least four spots.
You might look at the overall record and think these teams were successful in their NCAA Tournament trips. However, note that all seven were top 3 seeds. All won their first round game, but these seven teams collectively went 3-4 in second round games. The two number one seeds that were drastically over seeded both crapped out in the second round and no team from this list made the Final Four.
So who fits the criteria in this year's NCAA Tournament? We'll start with teams that were under seeded by at least four spots.
I did go to an ACC school, but I don't have loser SEC energy and root for the league above all else (I hate Duke and North Carolina), so don't blame rampant homerism for these numbers. Yes, you are reading that chart correctly. Louisville had the in-conference efficiency margin of a one or two seed (+19 per one hundred possessions). The Cardinals are in a tough spot in the 8/9 game, but they could give Auburn some trouble in the second round, especially with the game in Rupp Arena. As for the Tar Heels, in my humble opinion, they did not deserve (achievement) a spot in the tournament, but the metrics (performance) rate them a solid tourney team. I wouldn't be shocked if they upset Ole Miss in the first round.
Now what about the over seeded teams?
Only one team was over seeded by at least four spots and wouldn't you know it, its a team with a recent track record of losing early and a new coach who has never won an NCAA Tournament game.
Before we close, as always, please take a look at my magnum opus from the canceled tournament five years ago (sorry St John's and Michigan State) and thanks for reading. Things look to be trending toward an inevitable Duke/Florida showdown in the championship game, but I think we'll have a damn good time getting there.
Next week, we resume our offseason recaps by checking in on Conference USA.
No comments:
Post a Comment