You see teams; Sheldon Jacobson sees numbers.

Before the 68-team NCAA men’s basketball tournament field was unveiled Sunday, a computer science professor at the University of Illinois had already run roughly 3 million bracket simulations. Where the teams would actually end up in that field? To Jacobson, that didn’t really matter.

“We don’t really care where the teams slot in; we just care about the structure of the bracket,” said Jacobson, who runs the simulations for his website, bracketodds.com. “After the (field is announced), we’ll score those throughout the tournament and report the results. It’s very counterintuitive, but the fact is that the numbers do not lie.”

Hidden in the seeds, Jacobson and many of his fellow statisticians believe, are patterns that unlock the key to successful NCAA Tournament brackets. A combination of identifying the optimal pathways and the right amount of luck, then, could be key to topping your pool this March.

The NCAA men’s tournament began Tuesday night with two First Four games. The First Four women’s games are Wednesday night.

Jim Scott, who chairs the statistics department at Colby College in Waterville, begins by picking a champion from one of the top seed lines.

“There’s going to be upsets, especially in the first couple rounds, but after that, it’s really the top seeds that rise to the top,” Scott said. “Ever since we’ve moved to this larger field, something on the order of 75% of the time, it’s a 1- or a 2-seed that wins it all. So, the first thing I do is I pick the winner, start from there and kind of work backward.”

Jacobson and his team at Illinois began research about 15 years ago into how seedings can predict success in March. Whereas he felt everyone else was focusing on numbers related to basketball — points per game, pace, efficiency ratings and the like — he was more concerned about the numbers in the bracket itself.

Looking at the No. 11 seed gave Jacobson an answer. Teams seeded 11th have made a total of five Final Fours, more than No. 6 seeds (three), No. 7 seeds (three), No. 9 seeds (two) or No. 10 seeds (one). Why? Because 11 seeds have a pathway to a regional title that sets them up particularly well to do so.

“If you’re an 11, your path is much more attractive because you’ll play the 6, which is typically a fringe top-25 team,” Jacobson said. “Now, you might say, ‘Oh, they have to play the 3 and the 2 after that,’ but here’s the thing: There’s going to be upsets along the way. The 2 and the 3 have a good chance of getting knocked off, and that 11 will be getting the benefit of that.”

The trouble, though, is identifying which of those lower seeds are the right ones to pick. No. 11 seeds, after all, are only 58-94 against No. 6 seeds since the tournament expanded to 64 teams in 1984. No. 12 seeds, famous for beating No. 5s, are still only 53-99 in that matchup, meaning it’s more likely the higher seeds will move on.

That’s why statisticians recommend working backward to create your bracket. Jacobson recommends starting with the Final Four or even the Elite Eight before identifying potential Cinderella picks.

Picking some early-round upsets and picking a Cinderella team to make a deep run is also important, according to Scott, the Colby professor. Sure, your bracket might fare better from a points perspective by going with the favorites, but with other people in your competition also likely to do so, your likelihood of getting rewarded for those picks isn’t so high.

“The strategies are kind of like playing the lottery in terms of maximizing your winnings,” Scott said. “If you pick 1-2-3-4-5-6 in the lottery or 7-14-21-28 and multiples of seven and win, you’re sharing that pot with 1,000 other people who also picked that. You want to avoid the straight favorites because, even if you win, you’re going to win a lot less.”

That makes it important, Jacobson said, to look at past brackets, and last year’s total chaos aside, there’s been some reasonable order to the madness lately. The 2022 Final Four saw one No. 1 seed, two No. 2 seeds and a No. 8. In 2021, there were two No. 1s, a No. 2 and a No. 11 — top-heavy Final Fours with a surprise team sprinkled in.

Knowledge of the teams in the field and how that field is made up, of course, always has a place in any bracket, and Jacobson has used some of that in his grand prognostications for this year. His biggest prediction? That the upcoming tournament might not have as many upsets as those of years past.

“Because we just had so many teams that won their conference and stole a bid, you have a lot of teams at 11 or 12 that usually don’t end up there,” Jacobson said. “As a result, I think we’re going to see fewer upsets on the 5 line and potentially on the 6 line as well. We may have a more predictable year than people expect because of the way things panned out there.”

That could be good news for the quest for the perfect bracket, something that’s never been done before — not that people aren’t at least getting closer. In 2017, one bracket was perfect through the first 39 games before being busted. Two years later, a man from Columbus, Ohio, made it 49 games unbeaten before he, too, had his bracket dashed.

Could a potentially chalkier 2024 tournament, then, be the one? Sadly, statisticians don’t think so. The odds, Scott said, could end up being as high as one in 1 trillion when probabilities are assigned to more than 9 quintillion possible bracket combinations.

And one unexpected upset, as San Diego State statistics lecturer Chris O’Byrne noted, is all it takes.

“From a statistical standpoint, I’d be very shocked if it ever happened,” O’Byrne said. “I think you’re more likely to win the lottery, to be honest with you. People do win the lottery, sure, but I’d put the odds of a perfect bracket far outside the odds of winning the PowerBall. There’s just too many possibilities out there.”

« Previous

Softball: Perennial power Biddeford steps into season as a team in transition

Next »

Brayan Bello named Red Sox starter for Opening Day, youngest in 29 years