I got a lot of email and comments on the mythbusting blog comparing the tourney performance of Power and non-Power conferences. A number of people rightfully took me to task on my claims that there were more Mid-Majors and Smalls in the early years of the dance. The fact is, there were fewer Power conference teams to choose from, so the raw numbers didn’t tell the whole story. Still, I stand by my claim that proclamations of parity between the Powers and Mid-Majors/Smalls are exaggerated. I’m waiting for a non-Power team to cut down the nets before I’ll agree that parity have arrived.
That said, Insider Jbessa had a good suggestion. His notion of parity has less to do with conference affiliation and more to do with the difference in the quality of the teams. It’s an interesting point—so I decided to crunch some numbers along those lines.
The first thing I did was to compare the KenPom efficiency values of six different seed groups over time. I took the seeds two at a time (1-2, 3-4, 5-6, 7-8, 9-10 and 11-12) and tracked their average Pythag value over three three-year periods, since I only have nine years of KenPom data. Here’s what I discovered:
One and two seeds—basically the top eight teams in the country—were slightly less efficient over the last three years than they were between 2004 and 2006. The same goes for three and four seeds, though the gap is slightly wider. The next group of seeds, however, significantly drops off in terms of efficiency. Five and six seeds owned .9126 KenPom ratings between 2004 and 2006. That rating has dropped to .8753—nearly the same efficiency value as seven and eight seeds.
Actually, if you look at the gap between 5-6 and 11-12 seeds, it’s gotten much, much tighter over the years…and that no doubt explains why we’ve has a spate of upset-laden dances recently. From 2004 to 2006, the gap between 5-6 and 11-12 seeds was .0940. Since 2010, it’s been .0513. That’s smaller than the gap between 9-10 and 11-12 seeds during the chalky dances of 2007-09.
Think about it another way. During the decade between 1993 and 2002, the 80 11 and 12 seeds were, 37-80, 5.5 games below the number of wins they should’ve had based on their entire tourney history. Since 2003, the most recent 80 11 and 12 seeds were 46-80, 3.5 games better than expectations. And most of that overachievement has occurred in the last five dances. The 40 11 and 12 seeds since 2008 have won 29 games—7.7 games above expectations.
Okay. So…I think I have to admit that we are seeing a narrowing in the gap between the elite teams and the next level of squads. In particular, teams ranked 17 through 24 are much less efficient, while teams ranked 41 to 48 are more efficient than they’ve been in the past. Generally speaking narrowing of the quality gap (sure, call it “parity”) has manifested itself in more 5v12 and 6v11 upsets.
I’ll keep my eye on this when the brackets get set next Sunday and report back on the relative quality of all the seeds compared to their historical counterparts. It may help you decide how many upsets you want to call in your bracket…and which ones to pick.