Much has been made recently about how Pitino has only been to the elite 8 or beyond 3 times in his 10 seasons at Louisville. In his other 7 seasons, he’s been to the NIT twice, lost in the first round three times, and lost in the second round twice. I’m of the opinion that the tournament is a crap shoot. In a single elimination tournament, anything can happen. All a coach can be expected to do is put his team in a position to win; the rest happens.

That said, this is an empirical question. Were Pitino’s three successful teams "better" than his three unsuccessful teams in the regular season?  

* unranked = 26 in AP poll, last two rows are means, tournaments games instead of wins are used to discriminate between NIT teams and first-round losses.

The answer appears to be yes. His three successful teams had lower RPIs, AP ranks, better overall records, and better conference records. (All these results are statistically significant based on independent samples t-t-tests). This suggests that Pitino’s better teams do better in the tournament. If we’re trying to find fault with Pitino, maybe this is where it is. Good Louisville is better than bad Louisville.

This leads to the 2nd question. Should we have seen this coming? On the surface, this year’s 2010-2011 team had  a higher RPI, higher AP ranking, worse overall record, and a worse conference record than Pitino’s good Louisville teams. So, how well should Louisville have done in the tournament? Using a multiple linear regression, I plugged in these 4 stats and used them to predict tournament games. This allowed me to predict how many games UL should have played in this tournament based on the last 9 years of data. Louisville should have played 0.93 games. In other words, we should have expected UL to lose.

I was hoping to find a different result when I ran these data. But stats don’t lie. Good Louisville teams are better than bad Louisville teams, and this team was closer to a bad Louisville team than a good Louisville team.