As a mathematician and the head of Centre College’s sports analytics initiative, Jeff Heath brings a number cruncher’s approach to correctly predicting NCAA Tournament games.
Heath, 41, has created a system that deploys various metrics to project how the NCAA tourney bracket will play out. Yet even a college professor of math and data science has learned a hard lesson about predicting March Madness.
“Randomness always wins,” Heath says. “We will come up with our best rating, we will be really excited, and in comes Saint Peter’s and screws up everything.”
While there are no guarantees, history shows there are some factors that, if evaluated properly, might gain you an edge in your NCAA Tournament pool.
Here are 10 tips that could help you fill out a winning bracket:
1. What to make of conference tournaments . Of the last 10 NCAA champions, only two — Villanova in 2018 and Kansas last season — won their league tourneys.
Since the NCAA Tournament bracket expanded to at least 64 teams in 1985, only 16 out of 37 subsequent NCAA champs won their conference tournaments.
However, history says that conference tourney performance can help you figure out who NOT to choose to win it all. Since 1985, no team that lost in its first conference tourney game has gone on to win the NCAA title.
2. The “upset line” has moved . The traditional “5-12” upset line in the NCAA Tournament bracket still yields its share of surprises. Over the past six NCAA tourneys, the No. 12 seed has beaten the No. 5 nine times in 24 games. In the past three renewals of March Madness, the No. 12s have split even with the No. 5s, going 6-6.
However, the most likely spot for first-round upsets in recent NCAA tourneys has been the “6-11” line. Since 2016, No. 11 seeds have a winning record vs. No. 6s at 14-10. At least one 11 seed has beaten a No. 6 in each of the past 10 NCAA Tournaments.
3. There’s no percentage in picking upsets of top-three seeds . Though a No. 15 seed has beaten a No. 2 seed in each of the past two NCAA tourneys (looking at you, Kentucky), there’s not much historical percentage in picking upsets of top-three seeds in the first round.
Since 2016, No. 1 seeds are 23-1 (looking at you, Virginia) vs. No. 16s; No. 2s are 21-3 vs. No. 15s; and No. 3s are 22-2 vs. No. 14s.
4. What makes a team vulnerable to an early exit . If you trust the numbers geeks, teams that are the most susceptible to NCAA tourney upsets are those with: a higher than average turnover rate; poor three-point shooting percentage; and/or those that average taking fewer free throws a game than their opponents.
5. No. 1 seeds tend to make the second week . Since 2016, 19 of the 24 No. 1 seeds have made it to the round of 16.
Over the same time frame, 14 No. 2 seeds, 14 No. 3 seeds and 13 No. 4 seeds have survived and advanced to the second week.
6. A challenging year for those who value “program momentum.” Since 2016, the teams with the most victories in the NCAA tourney are:
1. Villanova 20; 2. Gonzaga 19; 2. North Carolina 19; 4. Kansas 18; 5. Michigan 14; 6. Duke 13; 7. Oregon 11; 7. Texas Tech 11; 9. Baylor 10; 9. Houston 10; 9. Virginia 10; 12. Kentucky 9; 12. Purdue 9; 12. UCLA 9.
What makes the “riding the hot program” school of tourney projecting tricky in 2023 is that three of the top five teams on the above list — Villanova, North Carolina and Michigan — and perhaps as many as five of the top 10 are not expected to make the field of 68.
7. No. 1 seeds tend to cut down the nets . A No. 1 seed has won the NCAA Tournament in eight of the past 10 renewals of March Madness, including the last five in a row.
The conventional wisdom in the current season, however, is that the level of separation between the projected No. 1 seeds and the rest of college hoops is not as great as in most years.
If you buy that, this might be the year to pick someone other than a No. 1 to win it all.
8. Picking the national champion . Thirteen of the 20 NCAA champs since 2002 have ranked in the top 20 in both adjusted offensive and defensive efficiency in the Pomeroy Ratings.
Last year’s winner, Kansas, finished No. 6 in offensive efficiency and No. 17 in defensive efficiency — by the end of the NCAA tourney.
However, on Selection Sunday, 2022, KU was not one of the five teams in both top 20s. In fact, none of that five — Gonzaga (No. 1 offensive, No. 7 defensive); Arizona (No. 5, No. 20); Houston (No. 10, No. 11); Baylor (No. 9, No. 14); and UCLA (No. 15, No. 12) — even made the Final Four.
This year, through games of Thursday, there were only four teams in the top 20 in both offensive and defensive efficiency at kenpom.com — Houston (11, 4); Alabama (19, 3); Connecticut (10, 15); and Texas (18, 11). UCLA (21 and 2) was close.
9. Believe in blue . Since 2004, every team but two that has won it all has had blue among its primary school colors.
The exceptions are Baylor (green and gold), which won in 2021, and Louisville (red, black and white), which won in 2013 — but subsequently vacated the NCAA title due to rules violations.
Among this season’s Pomeroy Ratings “dual championship qualifiers,” only UConn (navy, grey and white) incorporates blue into its uniform colors.
The other three, Houston (red, silver, white), Alabama (crimson and white) and Texas (burnt orange and white) eschew the blue.
10. System picks . Heath, the Centre College math professor and sports metrics enthusiast, incorporates strength of schedule, head-to-head outcomes, point differentials and where games are played into his team-rating system.
According to Heath’s system, the top five teams in men’s college hoops for 2022-23 are: Houston, Alabama, Tennessee, UCLA and UConn.
Of course, two of those five teams, Tennessee (Zakai Zeigler) and UCLA (Jaylen Clark) have recently lost key players to injury.
Good luck surmounting the “randomness” as you try to navigate “Bracket-ville.”
