Saturday, March 26, 2011

Cinderellas and the Little Ten

The belle of the ball in the NCAA tournament has to be VCU, who have made it to the Elite Eight after needing to participate in a play-in game. By any measure, expanding the field to 68 teams has to be counted as a success. But what of other "lessons" people seem to be garnering from seeing the outcomes?

Illinois is part of the Big Ten and we did our part in the tournament by winning our first round game as a 9 seed by beating UNLV. We even gave Kansas a run for their money in the next game. But there are no Big Ten teams in the Elite Eight this year, the only "power conference" not to secure a slot in that esteemed grouping. The SEC leads the way with 2 teams (Florida and Kentucky). The ACC has North Carolina, the Big Twelve Kansas, the Big East Connecticut, and the Pac Ten Arizona. Along with VCU the other Cinderella team is Butler. (How can you be a Cinderella if you made it to the championship game the year before?) . Perhaps not having the school name be a state is sufficient for putting on the glass slipper.

What I want to ask here is whether we can infer from this that the Big Ten was overrated or that the best conference in fact was the SEC, using the number of teams in the Elite Eight as the criterion for the ranking. And on that, here are a few points. First, teams that are younger (mainly freshmen and sophomores as the starters) are apt to get better over the season as the players learn their roles and adjust to each other. However, all inter-league play happens early. So the results there, with a few exceptions, tend to be biased against the younger teams and favor the more experienced clubs. If a young but talented team comes on late that shouldn't be a surprise. But it often is because this learning over time isn't in the rating systems. North Carolina seems to be in this situation.

Second, teams that get to be number 1 in the country in the polls start to play defensively, to protect a title that they really haven't earned but to which they've been anointed. To the extent that such a team also has young players, the rating can block the learning. I'd put Ohio State in this category. It didn't seem to me they improved that much from mid season on. And if Diebler is missing from the outside they are not that amazing a club.

Third, if you just look at wins and losses, that would seem to leave a lot to chance, allowing a better team to be upset by a lesser team. But if you look at a basketball game as a series of possessions, with the winning team the one that wins the most possessions, you come to a different sort of conclusion. In my back of the envelope calculation, a possession is 20 seconds on average so a game without overtime has 120 possessions. When there is a decisive margin, it is pretty good evidence that the better team one. On that VCU was simply better than Georgetown and Purdue, but about even with Florida State. Since that game went to overtime, one can certainly imagine the final score going the other way and likewise for Kentucky against Ohio State. On the other hand, if you look at the graph of the game flow on Butler versus Wisconsin, you see that Butler had a commanding lead in the middle of the second half, and then Wisconsin "made the score more respectable" at the end. In this case the point spread at the 30 minute mark is probably a better indicator, meaning Butler was clearly the better team. On that metric, Illinois did better against Kansas. That game was close until about 5 minutes left.

Four of the Big Ten's losses in the tournament were to Elite Eight teams and only one of those was close (Kentucky over Ohio State). Transitivity has a way of failing in this sort of analysis (if A is better than B and B is better than C then A is better than C). But it is impossible to make inferences about quality of teams without assuming it. Given the evidence, it is hard not to conclude that the Big Ten was overrated, certainly at the top of league.

There is then the question not of inferring quality at the top, but rather quality at the middle, measuring conference quality by the number of teams it gets into the tournament or perhaps the number of teams normalized by size of conference. On that one maybe 10 years ago I'd have said that there are two many automatic qualifiers from mediocre conferences. Now my conclusion is the big name schools are displacing some decent and talented lesser name schools. The Big Ten got 7 this year. That was probably 2 too many. I'll let Charles Barkley be the one who whines about the Big East.

For the rest of this tournament, I will root for Cinderella. It would be great for College Basketball if Butler won it all. And it would be wonderful to see VCU upset Kansas. If you are not a fan of one of the other Elite Eight teams, why else watch?

No comments: