The Round of 64:
Graveyard for Busted Brackets (2020)

To have winners in any round, you must limit the mistakes in the Round of 64. If a Seed No. 4 wins the national championship and your bracket had them being upset by a No. 13 seed, 630 of the 1920 points have been lost (using the ESPN scoring system.) On the other hand, teams seeded No. 13 or lower have never reached the Final Four.

Let us look at each seed match-up in the Round of 64. All numbers reported are based on the 35 tournaments from 1985 to 2019.

No. 1 vs No. 16 : Yes, it will happen, and it did happen in 2018. Just as the Cubs finally won the World Series (it took 108 years), a No. 16 seed finally won a game; no one knows just when it will happen next. Given that over 40% of the teams in the Final Four are No. 1 seeds, pick all four No. 1 seeds to win.

No. 2 vs No. 15: Yes, this one happens, occasionally. Only eight of the No. 15 seeds have won this game. For many years, the performance of the No. 15 seed looked similar to the No. 16 seed. However, since 2012, four No. 15 seeds won (out of 32 games played). Picking such a winner is tough. Moreover, the penalty for being wrong is enormous, given that over 20% of the No. 2 seeds make the Final Four. Pick all four No. 2 seeds to win.

No. 3 vs No. 14, and No. 4 vs No. 13: These matchup are grouped together because of their similar records. No. 13 seeds are 29-111 and No. 14 seeds are 21-119. Yes, this one happens, and is highly likely to happen. In 30 of the past 35 tournaments, one or more of these seeds pulled a major upset. The penalty for being wrong is stiff, since over 21% of the teams in the Final Four are No. 3 or No. 4 seeds. Moreover, only eight of the No. 13 or No. 14 seeds have ever reached the Sweet Sixteen, and all have lost that match-up. Pick zero of these teams to win if you are constructing only one bracket. If you are constructing multiple brackets, scattering one such upset in a few of the brackets is reasonable.

No. 5 vs No. 12, No. 6 versus No. 11, and No. 7 versus No. 10: These matchups are also grouped together because of their similar records. No. 12 seeds are 50-90, No. 11 seeds are 52-88, No. 10 seeds are 55-85, resulting in a combined record of 157-263, or a 37% winning percentage. Moreover, at least one No. 11 or No. 12 has won a game in every tournament. The combined record of the No. 10, No. 11, and No. 12 seeds in the round of 32 is 66-91, or a 42% winning percentage (better than their performance in the round of 64). They do fade in the Sweet Sixteen, with a record of 17-49, or a 26% winning percentage, with No. 10 seeds and No. 11 seeds getting 16 of these wins. No. 11 seeds have had the most staying power, reaching the Final Four four times. Picking zero, one, or two makes sense, with a slight preference for zero or one if you are constructing only one bracket. If you are constructing multiple brackets, scattering one, two or even three such upsets in a few of the brackets provides some diversity in your bracket pool. There will likely be more than this, but the price in picking too many and getting them wrong is too high.

No. 8 vs No. 9: The true toss up game. Since 1985, the record for the No. 8 seeds is 68-72. Flip a coin, or pick the more interesting mascot. Neither seed stays very long; only 19 of them have defeated the No. 1 seed in the Round of 32, which are fewer wins than the No. 10 seeds (23), the No. 11 seeds (22) or the No. 12 seeds (21) have in the Round of 32.

 

bracketodds.cs.illinois.edu ©  2020 - The Board of Trustees of the University of Illinois.
All Rights Reserved.