FanPost

Schmolik's Comments on the NCAA Field and the 2021 Bracket Matrix

Hello, college basketball fans!

So the NCAA Tournament field is set.

Every year before the tournament I post my own bracket, the Schmolik 64, of what I think the NCAA field should be. Then after the actual NCAA field is picked I compare my bracket to the NCAA's and say how they differ and what I disagree with the NCAA Selection Committee about. I feel that people will always complain about this specific team being in the NCAA Tournament and this team being left out but in my opinion to really be able to critique the Selection Committee you have to choose your own tournament field or bracket before the NCAA does. People in the media will talk about this team deserving to be in and this team not deserving to be in. The fact of the matter is the NCAA Tournament each year under its current format requires 68 teams, no less, no more. It's very unlikely for there to be exactly 68 teams that "deserve" to make the NCAA Tournament in a given year. Usually I'll have a team or teams I think deserve to make the field that I can't put in or I'll have not enough teams that deserve to make it and I'll have to find teams I really don't think deserve to make it but I have to choose someone.

In the end, the NCAA and I agreed on 67 of 68 teams. It turns out this is the most the NCAA and I agreed since 2013 when the NCAA and I completely agreed on the entire 68 team field. The only team the NCAA had that I didn't have was Wichita State. In my final Schmolik 64, I didn't even have Wichita State as one of my four "backup" teams. According to WarrenNolan.com, Wichita State was #72 in the NET (these are after Sunday's games, I formed my bracket based on Saturday's NET numbers). They were 2-3 vs Quad 1, 2-1 vs. Quad 2, 6-1 vs. Quad 3, and 4-0 vs. Quad 4. Their Quad 3 loss was their AAC semifinal vs. Cincinnati. When your NET is that low, you have to show you are better than your NET ranking. Michigan State was #71 but they had five Quad 1 wins (including #1 seeds Illinois and Michigan and #2 seed Ohio State, all since February 23) and had no losses vs. Quad 3/4 teams. The only at large team I could find with four or fewer wins vs. Quad 1/2 teams was Utah State and their NET was #39. Mississippi, the team I had in my field instead of Wichita State, was #53 in the NET with three Quad 1 wins and four Quad 2 wins.

In terms of seeding, I got the four #1 seeds exactly. Looking back at past Bracket Matrix records, this is the first time the NCAA and I chose the same four #1 seeds since 2013. My #2 seeds and the NCAA seeds are also the same, the first time that happened since 2013 as well. I posted a Fan Post discussing NCAA Tournament seeds and the "most important" seeds are near the top so if the NCAA gets them right it's more important than if they get the #5 seeds right. Of course when I say the NCAA "gets them right" I am assuming I am right which is of course debatable:)

The highest seed I disagreed with the NCAA with was Oklahoma State. I was considering them as a #2 seed but their loss in the Big 12 Tournament final and Houston's AAC Championship made me switch them at the last minute and drop Oklahoma State to a #3. But the NCAA made the Cowboys a #4 seed! Oklahoma State's NET ranking is very low (#29) but they had 10 Quad 1 wins, second only to Illinois who had 11 wins. Houston (#5 in the NET) meanwhile had just two Quad 1 wins. Oklahoma State beat #3 seed West Virginia two out of three times with both wins in the month of March, one at West Virginia and one in the Big 12 Tournament. This is only the second year the NCAA has used the NET instead of the RPI for the NCAA Tournament (2020 doesn't count since there was no tournament). Another team with a lot of Quad 1 wins but a low NET ranking was Missouri (#47, seven Quad 1 wins). It didn't seem like the NCAA rewarded them or didn't reward them as much as I thought they should have been or had in the past. I had Missouri a #6 seed while the NCAA made them a #9 seed. By contrast, San Diego State had a high NET ranking (#18) but had zero Quad 1 wins. I had them a #8 seed while the NCAA gave them a #6 seed. Conventional logic in the past has been the NCAA considers the NET (or in the past, RPI) more for the Quads than a team's actual NET ranking. But these examples seem to point out that the NCAA might care more about team's NET ranking than their Quad totals. On the other hand, Colgate's NET ranking was practically neglected. It was a given that Colgate wasn't as high as their NET ranking (#9) but the NCAA made them a #14 seed rather than the #12 I gave them.

In addition to comparing my bracket to the NCAA's, I will also compare the NCAA bracket to the Bracket Matrix, a collection of brackets across the internet. As always I like to thank Brian, the site host, for all the work he puts into the Bracket Matrix, giving myself and many other bracketologists a place on the internet.

This season, a total of 201 brackets were included in the Matrix, the first season more than 200 brackets were included. The first season, 2006, only 23 brackets were included.

If the Bracket Matrix was in charge, Wichita State would not have made the field. Mississippi would not have either. Louisville would have. 186 out of 203 brackets or 91.62% of all brackets chose Louisville (I was one of the 15 that didn't include them, I'm not sure if that's good or bad). In the history of the Bracket Matrix, only four teams that were left out of the NCAA Tournament were included in a higher percentage of brackets, 2011 Virginia Tech (97.75%), 2007 Syracuse (96.67%), and 2019 TCU (93.33%), and 2018 USC (92.51%) was pretty close. Among teams selected for the Bracket Matrix, Louisville wasn't chosen by the fewest number of brackets. Syracuse was chosen by 173 brackets, Drake was chosen by 166, and Utah State was chosen by the fewest, 151. Wichita State was the most popular choice among teams that didn't make the Bracket Matrix bracket. 91 of 203 brackets (44.83%) had Wichita State. The NCAA has had way worse NCAA teams % wise than Wichita State. While I didn't agree with Wichita State, a lot of brackets did agree with Wichita State's inclusion. Only 10 brackets had Mississippi. If you consider the Bracket Matrix correct, Mississippi was probably one of my stupidest picks ever.

Just like me, the Bracket Matrix also had the same #1 seeds as the NCAA. The Big Ten Tournament final between Ohio State and Illinois went to overtime and we'll never know what would have happened had Ohio State won in terms of Ohio State's and Illinois's seeds (and possibly Michigan's). With Illinois winning, it was in my opinion the most clear cut #1 seeds I remember in a long time and the Bracket Matrix agreed. The Bracket Matrix calculates an average seed for each team over all of the brackets included. Out of the 203 brackets, Gonzaga's and Baylor's average seed was 1.00 to two decimal places, Illinois's average seed was 1.01, and Michigan's was 1.02! I counted one bracket which had Baylor a #2 seed, three which had Illinois a #2 seed, and four which had Michigan a #2 seed.

The #2 seeds were also crystal clear in the eyes of the Bracket Matrix bracketologists. The average seeds were Alabama (1.97), Ohio State (2.05), Iowa (2.07), and Houston (2.21). I counted five brackets that had Alabama as a #1 seed, four which had Houston as a #1 seed, and one which had Iowa as a #1 seed.

Meanwhile, I said Oklahoma State should have been a #3 seed instead of a #4. The Bracket Matrix agreed with me. Their #3 seeds would have been Texas (3.01), Arkansas (3.03), Oklahoma State (3.04), and Kansas (3.15). West Virginia was without a doubt a #4 seed (3.97) as was Virginia (4.02), Purdue (4.13), and Florida State (4.45). ESPN's Joe Lunardi pointed out Oklahoma State should have been a #3 seed as did USA Today (they also said West Virginia should have been a #4 seed). I had Creighton as a #4 seed instead of #5 but the Matrix agreed with the NCAA that Virginia should have been ahead of Creighton (average seed 5.08).

Some other teams that the NCAA and Bracket Matrix disagreed on:

Florida (Matrix 8.43 or #9 seed, NCAA #7 seed)

LSU (Matrix 6.34 or #6 seed, NCAA #8 seed)

Missouri (Matrix 6.79 or #7 seed, NCAA #9 seed)

Wisconsin (Matrix 7.26 or #7 seed, NCAA #9 seed)

I would say based upon the Bracket Matrix the NCAA did do a good job with this year's bracket and considering the unique situations of the season with COVID-19 interruptions and incomplete data that is something to praise them for.

Coming later this week, another annual tradition along with the Schmolik 64, Schmolik Bracket Analysis!

User-created!