Late Night Thoughts

There's a lot going through my mind with the Selection Show less than 15 hours away. I've spent most of this week thinking about the operation of the tournament and how the operation of the tourney can impact a team's performance in the tourney. Here's what I've ran through my mind on numerous occasions.

 

The Indiana Bubble

The tournament will be played at six locations in the state of Indiana, four of which are in the city of Indianapolis. Here's the breakdown of site selection:

The two things that stick out to me are the S16 sites and the E8/F4 sites. Imagine a team that gets to play its S16 game on the same court that it played both of its R32 and R64 games. The familiarity with the courts (especially with the rims and the depth perception) may be a huge advantage in the S16 against an opponent that played both of its R32 and R64 games on a different court. Likewise, one potentially lucky team could get the opportunity to play five out of six of its games on a court inside Lucas Oil Stadium. This court-familiarity factor is something to pay attention to when the brackets are revealed. (Writer's Note: Notice that I specifically used the phrase "a court inside Lucas Oil Stadium." According to NCAA press releases, L.O.S. will have two courts constructed inside of it, but only one will be used at a time. If anyone finds information on the logistics of this setup (such as who plays on Court A and when and who plays on Court B and when), I would love to see a link.

Another important factor about the Indiana Bubble is time of arrival. According to an interview with Dan Gavitt, eleven teams will arrive in Indianapolis on Sat Mar 13 due to winning their conference's automatic berth. With their guaranteed spot in the tourney, these teams can begin Covid protocols. These eleven teams are Liberty, Morehead State, Winthrop, Loyola-Chicago, Appalachian State, UNC-Greensboro, Cleveland State, Mount St Mary's, Drexel, Oral Roberts, and Gonzaga. This matters to me because it generates consistency and routines sooner than teams winning on Sat Mar 13 or receiving at-large berths on Selection Sunday. This added consistency and routinization is a good catalyst for wins (or for 10/11 of these teams, a Cinderella upset win). I'm not suggesting each of these teams will get a Cinderella upset, but I would bet the highest percentage of R64 upsets will come from this pool of teams.

Last but not least, Indiana-based teams should have a home-field advantage (or at least 2021's version of home-field advantage). I'm looking at you Purdue.

The S-Curve Bracket

We have never seen a true S-Curve bracket and probably will never again (not without another pandemic, God forbid). The NCAA produces a metric called the S-Curve, which is their 1-68 rankings of the sixty-eight tournament teams. The Selection Committee never fills out a bracket according to the S-Curve because it tries to give higher-ranked teams home-field advantage (keeping higher-seeded teams closer to their geographic homes and fan bases). With all games being played in Indiana, the Selection Committee can fill out the bracket exactly according to the S-Curve. It means that the top-overall 1-seed (most likely GONZ) will be placed in the same region as the bottom-ranked 2-seed (8th overall on S-Curve), the top-ranked 3-seed (9th overall on S-Curve), bottom-ranked 4-seed (16th overall seed), and so on and so on, etc. You should see the pattern of 1vs16,8v9.....7v10,2v15 when only looking at the 1- through 4-seeded teams. Why does this matter to me? One of my most valuable tools is the Seed-curve model, which compares the Quality Curve to the Selection Curve for the purpose of exploiting deviations between the two for upset victims and Cinderellas. With the Selection Committee seeding according to the S-Curve (their interpretation of the ranks 1-68), the deviations between it and the Quality Curve (the mathematical interpretation of the ranks 1-68) should be at its widest (in other words, the victims and the Cinderellas should be most obvious).

Randomness

I stated in the opening article this season that there is potential for NCAA tournament games to be forfeited due to Covid protocols. For the operation of the tournament, the NCAA has posted a few guidelines. First, a team must have five players to avoid forfeiture. Second, once the tournament has started (Tues Mar 16 at 6pm), no teams in the field of 68 can be replaced due to failing Covid protocols. This means two things to me:

1. Teams that have endured Covid pauses in their programs are less likely candidates for forfeiture. I do not have a complete list, but BAY, MICH, LOU, VT, and IONA immediately come to mind. It is something to research, not to mention the teams that recently withdrew from conference tournaments due to Covid protocols (KU and UVA).
2. The selection committee has announced the implementation of Replacement Teams. If a team is selected to the tourney, but then is forced to withdraw within the 48-hour window, a replacement team -- if they choose to -- can take their spot in the field. Using the examples of UVA/KU, both are expected to be somewhere in the range of a 3-seed to 5-seed. If they are selected but then forced to withdraw, the first team listed in the First Four Out would be their replacement on their seed-line. In other words, we would have the situation of a 12-seeded team in quality-curve rankings posing as a 4-seeded team in the bracket playing against another 12- to 14-seeded team. If you follow my work, you know this has the makings of an upset alert, especially if the actual 12- to 14-seeded team is under-valued (i.e. - LOYC, WINT, or LIB).

Conclusion

There is a lot of things that go into the Madness of March. For this year and hopefully only this year, the operation of the tournament is one of those factors. Do keep these factors and many more on your mind as you make your picks. As always, thanks for reading my work and I'm looking forward to the unveiling of the bracket in less than 14 hours (yes, it took me an hour to write these thoughts down in a coherent article).

Also, I don't know how busy my job will be this week (but I do have Fri-Mon off on PTO), so I still don't know if it will be quick-hitter articles throughout the week or one-big conglomeration on Wed night/Thurs morning.