The following is part of a series of NCAA tournament posts. You can find the intro to the series here.
Narrative: Late season results are more important than early season.
Teams around the country are going to get “hot” every March. That’s just the nature of the single elimination conference tournament. One team from each conference is going to be left standing for the auto-bid.
Sometimes the biggest problem with looking at a narrative like this is trying to develop a useful definition of “hot”. To start, I looked at efficiency differential in November and December and efficiency differential in March. I then graphed both against number of tournament wins. Only teams seeded five or better were included (2004-2013):
The conclusion from the graphs is not that early-season success is more important than late-season. This is because we aren’t exactly comparing apples to apples. Teams generally only play four or five pre-NCAA tournament games in March. In January and February, however, they usually play somewhere around 12 games.
The actual takeaway here is that only looking at games from a specific time frame is not worth the trade-off of sample size. We learn more and more about a team after every game. The effects of being “hot” going into the tournament aren’t even close to large enough to simply push aside results from the first two months of the season.
The ACC champs are not your typical March momentum team. Unlike UConn in 2011, who appeared to flip a switch and suddenly improve in March, Virginia has steadily improved all season long. I wrote about Virginia’s in-season improvement over a month ago. At the end of January, Virginia ranked just 35th in the country in efficiency differential. Now heading into the tournament they current rank 6th overall.
I think it’s pretty safe to say that Virginia has legitimately improved as the season has gone on. Advanced ratings (like KenPom) do give extra weight to recent games, but are they still undervaluing Virginia because of the slow start?
Take a look at NCAA title contenders (five seed or better) over the last 10 years with similar situations to Virginia:
It’s interesting that no team on the list received a 1-seed. Virginia probably had the worst start to a season for an eventual 1-seed in the last 10 years. So Tony Bennett’s squad is in a little bit of uncharted territory. There’s no doubt Virginia is capable of winning a championship. However, maybe we shouldn’t forget about Virginia’s early season struggles. I should note that the sample size is still very small here and there is a lot of randomness in the NCAA tourney, but teams with steady improvement in the second half of the season haven’t had great tournament success.
Kansas, Alabama, and Georgetown were all unable to avoid first round upsets. UCLA was by far the most successful of the 10 teams, losing in the national championship game.
Ultimately I think there are reasons to believe Virginia is right up there with the popular picks: Florida, Arizona, Louisville, and Michigan State. At the very least, we know they are extremely annoying to play against.