March Madness is here: A guide to winning your Bracket Pool

A Guide to winning your Bracket Pool

Tired of always losing your family bracket pool to your sister who doesn’t know where Gonzaga is? Or trying to hit it big in your company’s work pool? Welcome to your guide to winning your 2025 NCAA Tournament bracket pool.

Many people think winning a bracket pool is about hours of tape grinding, years of basketball experience, or a whole lot of luck. In reality, putting yourself in a position to win is a numbers game. Especially in larger pools. Through optimization and historical trends, you can build a better bracket that maximizes your odds of finally having bragging rights over your sister.

1. Size matters

The size of your pool is the most important factor in figuring out how to fill out your bracket. The more participants in your pool, the more risks you will have to take in your bracket. A bracket with your brother and sister could go completely chalk and win over 90% of the time. If you’re trying to win ESPN’s Bracket Challenge with over 22 million brackets, chalk isn’t going to cut it. The more brackets in a pool, the less this article will help. Winning ESPN’s pool will require being close to perfect: last year’s winner scored 1810 points, missing just 8 games total and being perfect after the second round. The winer had 11-seed NC State in his Final Four. The bigger the pool, the further you need to take some upsets and high seeds.

If you’re in a smaller pool, the more your margin for error increases, and the more value you can extract from this article. You don’t need to take NC State to the Final Four, because nobody did. You’re not losing points to anyone else. Once your pool hits 10-15 brackets, you’re going to need to start taking more risks to win. Know the likely size of your pool and take risk accordingly.

1. Pick the champion

This one is rather obvious; to win a larger bracket pool, you need to pick the National Champion. Helpful right? Of course, picking the National Champion is more of a science than luck. Using historical trends we can begin eliminating teams as national title contenders. Since 2001, 96% of champions have been top-21 in KenPom’s offensive efficiency (goodbye Michigan State and St. John’s) and 91% of championships were top-31 in KenPom’s defensive efficiency (see ya Alabama and Texas Tech). Every champion since at least 2000 has had a first-round pick in the NBA draft, to play it safe, let’s remove any team that doesn’t have a player any team in Jonathan Givony’s Top-50 on his big board. We’re left with five teams: Auburn, Florida, Duke, Houston and Tennessee. Historical trends suggest one of those five will win it all. It may not seem like rocket science to include four 1-seeds and a 2-seed, but these trends historically included 2023 UConn (4-seed) 2006 Florida (3-seed), and 2003 Syracuse (3-seed). If a low seed could win it all, these trends would pick up on it.

1. Pick the optimal champion

Using that list of five, we can now find the optimal champion pick. When it comes to bracket pools, we need to exercise game theory to find the best possible choice. Let’s look at 2021 for example. That year there were two juggernaut schools that were both overwhelmingly favored to win it all relative to the field: Gonzaga and Baylor. According to betting odds and analytics tools like Torvik’s T-Rank, Baylor was slightly less likely to win the championship than Gonzaga; however, they were the optimal pick that season. Gonzaga was being picked in nearly 39% of bracket pools. In a pool of 100 brackets, even if Gonzaga won the championship, you would have needed to be better than 38 other brackets in the prior rounds. Exceptionally hard to do. Baylor was picked as the champion in 8.4% of brackets. If you picked Baylor, you only had to beat out 8 other brackets. Much more doable. Baylor won, and game theory optimists rejoiced.

ESPN makes their championship choices publicly available, and with millions of brackets, it’s a good barometer for where your bracket pool will pick. We want to avoid picking teams that are being overvalued by the public and find teams that are more likely to win the title than the public suggests. Here is where those five contenders rank in terms of public perception (how many times they’ve been picked as champion in ESPN brackets) and how likely they are to win the title (the average of T-Rank and EvanMiya’s projections)

TEAM	PUBLIC ODDS	TITLE ODDS
Duke	25.0%	22.5%
Florida	21.2%	13.5%
Auburn	10.9%	14.4%
Houston	8.8%	18.6%
Tennessee	4.3%	4.4%

Duke is the most likely team to cut down the nets, but the public recognizes that and is overvaluing Duke. If you’re in a pool with only a few brackets, you’re probably safe picking Duke to win it all, but in pools with 20 or more brackets you’re better off going another direction. Houston is the clear choice by the analytics, and Auburn is also being undervalued.

This method wouldn’t have won you much in 2024 when UConn cut down the nets as the number one overall seed and consensus choice in 25% of brackets, but this method has narrowed down the correct national champion every year from 2015 to 2023.

You can also apply this same logic to your Final Four and Sweet 16 choices to help create a near-game theory optimal bracket.

1. Know your poolmates

The data used in the prior section is all based on the general public. That’s not going to be the case for your bracket pools. You’re picking against an isolated portion of society, and that needs to be considered. If the name of your pool is “Alumnus of Houston University”, more than 9% of that pool is picking Houston to win it all, and you’re likely better off going with Auburn or Duke. Growing up in Spokane, I never pick Gonzaga to win my family bracket pool. Likewise, if you’re in a pool that you know has school ties, you’re likely better off picking against those schools.

1. Work backwards

It’s not the most fun way to fill out a bracket, but it’s the most effective way. Pick your champion first. That’s the most important part of a winning bracket. Picking the champion correctly is worth as much as a perfect first round. Everything should be built around your champion. If Houston is who you want to be your champion, you shouldn’t need to think about a Gonzaga/Houston matchup. You’re picking them to win it all, that requires winning in the second round.

From there, don’t jump back to the first round. Find your Final Four teams and write them in. Then work back to the Sweet 16. Then fill out the blanks in between. A majority of your points will come from being right in the later rounds, your focus should be there to start. No bracket was busted from picking the wrong 8/9 matchup.

1. Historical trends to know

There are a few other historical trends that will help you win on the margins in your bracket:

Avoid the 2-seeds that came from nowhere

Since the tournament expanded in 1985, 25 teams started the year unranked in the AP Poll and got a 2-seed. None of them made the Final Four. In fact, they average just 1.5 wins per tournament. St. John’s and Michigan State both apply this season. Avoid taking the Johnnies or Spartans past the Sweet 16 this year.

Don’t try to find Cinderella

Since 2005, teams who win as a double digit underdog are just 4-20 in the following round. If you’re convinced Troy will knock off Kentucky as an 11.5-point underdog, feel free to pick them. History suggest they won’t be a Sweet 16 team though.

Bad in conference play, good in the dance

There was plenty of discussion about the SEC getting multiple bids from teams with losing conference records. History suggests those teams will find success. Six teams have made the tournament (as an at-large) with a conference record at least four games under .500. Those teams are 5-1 in the Round of 64. This year that is true of Oklahoma and Texas.

Beware power conference tournament champions who didn’t get a 1-seed

In recent years, hot hands from conference tournaments haven’t carried over to the Big Dance. In 2024, Auburn won the SEC tournament and received a 4-seed. They lost in the first round to Yale. In 2023, Arizona won the PAC-12 tournament and lost to Princeton in the first round as a 2-seed. Marquette won the Big East tournament that year and lost in the 2^nd round to Michigan State. In 2022 it was Iowa losing as a 5-seed in the first round. This year’s power conference winners who didn’t receive 1-seeds are St. John’s and Michigan. Caution taking either team on deep runs in your bracket.

Don’t go chalk to the Elite 8

This one is specific to larger pools, where you’re almost guaranteed to share a champion pick. The round that you can separate your bracket tends to be the Elite 8 and that tends to be the round where you need to get your upset picks spot on. Typically, we see 2-3 teams that aren’t top-3 seeds make the Elite 8. Last season it was 11-seed NC State and 6-seed Clemson as well as 4-seeds Duke and Alabama. Analytics are high on Maryland, Kansas and Gonzaga this year relative to their seed lines. Those teams are worth a look in your Elite 8 in deeper pools. (Note: Gonzaga is slated to play Houston in the second round. If your pool allows multiple brackets, make Houston your title choice in one pool and take the Zags to the Elite 8 in the other. Cover your bases. You only need one winning bracket, not two)

There you have it. Congratulations on winning your bracket pool! Be sure to use your winnings on a Bronco Nation News subscription so you can get more expert analysis.

Nathan Carroll is a part-time contributor for Bronco Nation News and host of the TBA Basketball Podcast. Nathan works for Pro Football Focus as part of their data collection team, specializing in charting player data at the FBS level. Previously, he graduated from Boise State’s Honors College with a degree in Quantitative Economics and worked 5 years in Boise State’s athletic department. Follow him on X at Nathan_26_