Dave Sez

Postseason Performance

After filling out pool sheets, the second most popular NCAA Tournament pastime might be comparing conference performance. Everybody keeps track of who has how many teams left. Conferences are ranked by various arbitrary measures - number of bids, number of Sweet Sixteen teams, number of Final Four squads, total wins, winning percentage, etc. Each of those has some merit, but all are critically flawed.

The absolute team counts are handy, but clearly miss a lot of detail. If a conference puts two teams in the Sweet Sixteen, is that always better than the league who gets only one? What if that first league lost 5 teams on the first day, but the other conference put 6 teams into the first weekend?

Total wins is a better look at whole-tournament performance, but it too misses a critical element. If a team loses a game, should it's impact on the tourney be completely ignored? If so, then a conference with 6 bids and 2 wins is no better than a conference with only 1 bid and 2 wins.

Winning percentage is nice over the long run, but for a single tournament, it too lacks. Consider Conference USA this season. It's a weak conference and had an awful year, sending just one team, Memphis, to the tourney. But right now, the Tigers are 2-0. Does that mean that Conference USA is the best league in the country, better even than the SEC with its three Sweet Sixteen teams?

Years ago, I came up with a measure that I liked that I've always called Impact. The Impact of a league is a simple measure - take the leagues tournament winning percentage and multiply it by the number of bids it received. This way, performance is critical, but teams get credit even for taking part. Consider the ACC this year. Georgia Tech was the ACC's last team in the tourney. The Yellow Jackets lost their only game. If you looked at just winning percentage, the ACC would have been considered a better league just by virtue of GT missing out on a bid. Instead, using Impact, the ACC gets credit for the bid, but still pays the penalty for their loss.

In the table below, I show some basic measures, including Impact, for each of the six major leagues. I included the NIT and then a combination of both postseason tournaments as a way to get an even better measure of overall league strength. Note that I don't intend this to be the be-all-end-all of conference comparisons - I don't think you can ignore the regular season - but it's pretty fair. Teams have to perform well enough throughout the year to even get bids and then they need to perform in those postseason tourneys to show how good they really are.

Looking at the table, if you consider just the NCAA Tournament, it's no surprise that the Pac-10 and SEC have the best numbers. The SEC is winning at the highest rate, but the Pac-10 has had a greater Impact, because more of its teams have played.

If you look at all postseason teams, the Big East actually edges out the ACC for most overall Impact. Both leagues are helped immensely by strong NIT showings so far.

Keep in mind that there are still a lot of games left to play. Most of these numbers will come down as the teams start playing games against each other.