As a reminder, here was my final bracket. I have color-coded it to illustrate how I did. Teams in blue were picked and seeded correctly; teams in brown were one seed line off; teams in red were two or more seed lines off, or not picked correctly at all.
- UConn, Purdue, Houston, Iowa State
- North Carolina, Tennessee, Arizona, Marquette
- Creighton, Baylor, Illinois, Auburn
- Duke, Kansas, Kentucky, Alabama
- Florida, Wisconsin, BYU, Texas Tech
- St. Mary’s, San Diego State, South Carolina, Clemson
- Dayton, Gonzaga, Nevada, Washington State
- Nebraska, Texas, Utah State, Boise State
- Texas A&M, Colorado, Mississippi State, TCU
- New Mexico, Florida Atlantic, Colorado State, Northwestern, Oklahoma, Michigan State
- Drake, Oregon, NC State, Grand Canyon
- James Madison, McNeese, Samford, Duquesne
- Vermont, Yale, College of Charleston, UAB
- Akron, Oakland, Morehead State, Colgate
- Western Kentucky, South Dakota State, Long Beach State, Longwood
- St. Peter’s, Stetson, Grambling, Montana State, Howard, Wagner
Adding it all up, I missed one team entirely, Oklahoma; I missed five other teams by 2 or 3 seed lines; 21 teams were off by one seed line; and 41 teams were perfect.
Is that good? One way to compare is at bracketmatrix.com. Their scoring system gives you three points for accurately predicting a team being in the field; two additional points for every team that is seeded correctly; and one additional point for every team that is seeded plus or minus one.
There are 226 brackets total brackets scored. My score was 345, which was tied for 54th. Not too shabby. Here are scores of some of the better known sites and experts:
- This year’s top score: 355
- Warren Nolan: 349
- Washington Post: 347
- The Barking Crow: 347
- FOX Sports: 345
- MUDVILLE ANALYTICS: 345
- SI.com: 344
- Bart Torvik: 344
- The Athletic: 341
- Jeff Borzello: 338
- Sporting News: 337
- Joe Lunardi/ESPN: 336
- USA Today: 332
- On3.com: 325
- Jerry Palm/CBS: 323
I’m happy with that. The only picks I would like to have back are picking Iowa State over Carolina, which I knew was wrong as I was doing it but couldn’t stop myself, and picking Gonzaga as a 7. My model said Gonzaga was a 6, but I was influenced by external forces to knock them down to a 7.
Now for some speculation about some of the other misses. With FAU, sometimes I get the sense that the committee has its mind made up prior to the conference tournament, and then they can’t be bothered to change it based on what actually happens. FAU lost to Temple, which is a really bad loss, and it doesn’t seem to have hurt them. Same thing with Florida and Kentucky, the committee seems to have ignored the SEC Tournament. The Nevada/Boise State situation (both received much worse seeds than expected) seems to be some kind of conspiracy against the Mountain West. There was speculation that the committee felt that the Mountain West was overrated because most of their Quad 1 wins were within the league. If that’s true, that would call into question the validity of the NET rating, but that’s a discussion for another time.
There is more evidence of the committee ignoring the results of conference tournaments. Look at the St. Peter’s/Longwood situation. Longwood has a much better resume than St. Peter’s. Why did they get a lower seed? Well, a good guess is that the team that was supposed to win the MAAC, Fairfield, did have a better resume than Longwood. You can imagine the committee had Fairfield on the 15 line, and when Fairfield was upset in the conference tournament, they just did not do the work to understand how St. Peter’s resume was different and change the seeds accordingly.
Bracketmatrix has been doing this for a long time and you can look at past results on his site. Lunardi is OK, Palm is below average, although both have had some good years and some bad ones. There are just a handful of prognosticators who have been above average for five consecutive years.
Last year was my first year, and I was well below average. I definitely got better this year. We’ll see next year if my improvement is real and sustainable.
I don’t know if you saw the committee’s 1-68 seed list, but they originally had Gonzaga #21, which translates to a 6 seed. But because BYU doesn’t play on Sundays, they had to be bumped down from a 5 seed to a 6 seed, which moved Gonzaga up to a 5.
As for conspiracies against the Mountain West, I don’t know, but I did learn something new this year. In ESPN’s BPI ratings, Mountain West teams are consistently ranked lower than they are in other metrics. One reason for this may be that the BPI uses altitude as a factor in its ratings. I guess teams that play their home games at a high altitude are harder to beat at home? So, hypothetically, Colorado State would get less credit for a home win over, let’s say the Tar Heels, than UNCW would for the same home win. Seems weird, right?