Interpreting the First Round Lines

A few early thoughts on the first round lines, more in a macro sense than focusing on any particular game:

When the bracket was first announced and I was looking at projected lines for each game according to my ratings, I started noticing that in the middle of the bracket, the lines looked to be pretty close to pick'ems, even when there was a notable seed difference between the two teams (say, 6-11). Sure enough, my ratings don't have much of a difference between the average six seed and the average 11:

(X-axis is seed, y-axis is my rating of the average team on that seed line.)

I actually have the average six as 0.4 points worse than the average #11, and the graph is very flat between those two. Really, there is a steep dropoff from the 1s to the 2s, then another between the 2s and the 3s, but the average 3/4 is only a couple points better than the average 11.

I wondered if this would come though in the lines as well, and it indeed did. The following graph shows the average spread for each seed; the average for #1s will probably go up a bit after the play-in game when the Duke line is released.

There's more of a dropoff from 3-6 here, but that's mostly because the average #14 is a lot worse than the average #11. For the most part, we see the same thing; there are really about seven elite teams, and then 35 or so that are roughly equivalent.

I wondered the other day if my ratings were too high on the ACC, and if that may be contributing to my unusual stance on Duke (best team ever). There is unfortunately not a great way to determine this -- particularly since Duke's first round game is not yet lined -- but here's the difference between the lines my ratings projected and the current Pinnacle numbers, averaged across each conference.

The sample size is unacceptably low on these, and they'd definitely need to be regressed. So don't take this to mean that I'm 1.6 points too high on the ACC, but it's likely that I'm overrating the conference at least some. The big misses here were Florida St. (had them -3.9 vs. Gonzaga, it's -1), and Georgia Tech (had -1.4 against Oklahoma St., it's +1.5), but I was about a point optimistic on Wake, Clemson, and Maryland as well. We'll have to see how these five lines move, and what the Duke spread is released at on Wednesday.

Underrating the SEC was not something I had predicted my numbers are doing, but it appears that may be happening as well. I sold Kentucky, Tennessee, and Vanderbilt short by 2-3 points each in their first round matchups. This isn't really that surprising -- KenPom has the SEC as the fifth best conference, but far behind the other four, which isn't an opinion that you'll find too many places. So that's something to watch for as well.