In my “Madness Measuring” blog post on December 18, I argued that Pythag efficiency data can help you assess the relative predictability of a tournament. If you compare the Pythag values of the top 20 most efficient teams to their historical counterparts, you’ll find that the years when the elite squads were relatively more efficient correlated to chalky dances, while those years when they were less efficient aligned with upset-laden tourneys.
Take the 2011 and 2007 dances, the maddest and sanest of the 64-team era. 2011 saw the worst top 20 teams in terms of Ken Pomeroy’s Pythag efficiency calculations since the data became available in 2004. Conversely, 2007 featured the highest quality top 20. And what happened? The 2011 dance tied a record for upsets (13) and broke the Madometer unpredictability record (19.8% madness). Meanwhile, the 2007 broke records for yawn-inducing predictability, with just three upsets and a scant 4.1% Madometer reading–both the lowest totals ever.
If you do a line chart of the top 20 from both years, the performance gulf between them is readily apparent. The question is, what if you overlaid this year’s current top 20 Pythag leaders onto the chart? Would the line be closer to the chalkiness of 2007 or the craziness of 2011? The answer is somewhere in the middle:
It’s still too early to assess just where the 2012-13 quality curve will fall, but right now, this year’s best college teams aren’t any better or worse than their historical counterparts. The first through fourth most efficient teams are a little better than the average top four teams from 2004 to 2012. The next five are about average…and then there’s a sizeable drop from the ninth-best team (Michigan) to the tenth (VCU). From ten to 18, the teams are weaker than their average team at their ranking, but not as weak as the teams in 2011. However, after the No. 18 team (Michigan State), the quality falls precipitously. Teams 19 and 20—Oklahoma State and Miami (Fla.)–are worse even than their 2011 counterparts.
Let’s say the location and shape of this curve holds. What kinds of conclusions might we draw—assuming that teams are seeded by their Pythag position (a large assumption)?
- If one seeds are better than average, four seeds are worse than average—and five seeds are historically bad, you might want to pencil all one seeds right into the Elite Eight.
- Depending on how the quality curve extends out for teams ranked 21 to 52, you may want to pick more 5v12 and 4v13 upsets than usual.
- Since the two seeds are as efficient as average two seeds and three seeds are weaker than their seed counterparts, we may be looking at mostly 1v2 match-ups in the Elite Eight.
The deeper we get into conference play, the closer this curve will get to its Selection Sunday shape. At that point, we’ll be able to make a better comparison between the 2013 tourney field and the nine fields that came before it. Periodically throughout this season, I’ll update the curve, give my view on whether it suggests tourney madness or sanity, and draw some conclusions on the basic seed match-ups. Then, when the brackets are set, I’ll compare the top 13 seeds and 52 teams with their historical counterparts. That ought to help you decide which favored, contender and Cinderella seeds to advance—and which to send home early.