The Science of Winning Brackets

This week everyone is filling out NCAA tournament brackets.

People dream of winning their office pool and securing 12 months of glory and bragging rights over Grace from accounting.

Contrary to popular belief, though, brackets are not lotteries!

There are practices and strategies that will meaningfully increase your chances of making money in tournament pools.

I’m excited to provide my “Robin Hood Guide” to profiting in bracket-based tournament contests.

We’ll start with how pool structure affects strategy, then we’ll talk about the value of zigging when others are zagging with some specific recommendations based on current information.

As always, we’ll close with our NCAA basketball bets of the week.

Let’s get after it.

Understand the Rules of Your Pool

The most important things to understand before starting to pick the winner of individual matchups are: 1) the scoring system of your pool, and 2) the payout structure and number + type of entrants. Let’s get more detailed.

Scoring System

Setting aside more creative structures that reward choosing upsets more than picking favorites, the big question to answer is: how much do points for choosing correctly increase by round?

One extreme case, Yahoo’s Bracket Mayhem contest, doubles the points for picking correctly every round.

With a structure like this one, every round offers you the chance to achieve 32 points, but the weighting of each game increases tremendously over time.

You’ll almost certainly need to choose the champion correctly to win a large pool, barring a crazy upset winner emerging.

The opposite extreme, a structure where each correct pick is worth 1 point regardless of round, means more than half of the total scoring will happen in the Round of 64, and the importance of choosing the champion correctly is miniscule.

When we discuss strategies around deviating from the market, the comparative scoring potential in each round will dictate key points of leverage to separate from the pack.

Payout Structure + Size of the Pool

The more top-heavy a pool’s payouts are structured, and the larger the number of entrants, the more valuable it becomes to build a unique bracket.

To illustrate this point, consider first a family pool with 8 entrants, where each of the top four places win double their money, and the bottom four places lose.

Your entry fee is an even money bet, and so there exists minimal reward for taking risks that differentiate your bracket from the most likely outcome. Finishing in the 60th percentile guarantees you profit, and if you win, it’ll pay for your entry fee only next year.

Now consider the opposite extreme; an 800 person pool where all of the prize money goes to first place.

In this case, your entry fee is a +80000 bet, and the reward for creating an outlier bracket is massive! Finishing in the 60th percentile is worthless, but if you win, it’ll pay your entry fee for 10 lifetimes.

In summation, understanding the payout structure and size of your pool informs how valuable it will be to differentiate from the pack, and the scoring system of your pool informs what types of differentiation will be most impactful.

Skillfully Deviating From the Market

Most bracket pools have dozens, if not hundreds of entrants, paired with payout structures that are weighted toward the 95th percentile or better — the top 2-10 places.

This section will presume you’re trying to win that type of contest, and lay out an approach to maximize your chances of winning money.

We’re playing for a top 5% outcome and indifferent between finishing in the 70th percentile or the 5th percentile, presuming neither one gets paid.

Our approach will be to go against the grain in a strategic way.

Step One: Compare Teams’ Projected Odds with Public Perception

The first step is to identify 1) the likelihood of individual teams progressing in the tournament, and 2) the public perception of how teams will progress.

We will compare the difference between those two factors to create a differentiation score for each team/situation. Let’s use an example to illustrate how this works.

Juice Reel shows me the following odds for Houston and Tennessee to reach the Final Four from the Midwest Region.

Using a tool like this one to convert money line odds to percentage likelihood, we see that Houston projects as 43.5% likely to make the Final Four, and Tennessee projects as 20.4% likely.

That’s one factor in our comparison.

For the second factor, check this page to see the pick distribution the public is making in their actual Yahoo brackets.

Here we see the public choosing Houston and Tennessee to make the Final Four at rates of 46% and 25.9% respectively.

Because the 46% public perception of Houston to advance this far exceeds the betting odds estimate of 43.5%, Houston would have a differentiation score of +2.5%, meaning they are being overvalued that amount by the public.

By comparing those two figures across our brackets, we can find cases in which the public is over-selecting or under-selecting certain teams, and find cases to go against the grain.

Step Two: Picking Against the Grain

Once we have a sense of which teams are being over-selected in each round, we’ll refer back to the pool’s scoring system and find the optimal spots to choose teams with a negative differentiation score — those being under-selected.

This will allow us to create a unique enough bracket to separate from the pack, but to do so in a strategic way.

The easiest place to start is with our champion selection.

This choice will typically have the biggest impact on our ability to win a pool, and shows some clear differentiation trends.

Duke is being picked to win in significantly more brackets than betting odds would imply, so avoiding Duke as a champion is a great way to differentiate.

If you wanted to give yourself a disproportionate chance at winning, choosing the under-selected Auburn, Houston or Texas Tech as your champion would give you an unfair advantage against a large field.

I used this approach to win a 330 entrant pool in 2023, choosing UConn, so hopefully it serves you equally well in zagging your way to a big payoff!

Last week I shared that blindly betting power conference favorites off a bye, in league tournaments, had a 2% ROI over the past five seasons (and was a profitable tactic for me this year).

Our partner, Bankroll U, offers a cool, unique platform to test out such systems and strategies with different risk parameters from a typical sportsbook.

It’s a sportsbook simulator featuring real betting markets and odds, including live bets, paired with progressive gameplay that rewards players with cash bonuses as they advance through its levels. You get the excitement of betting across the board without risking more than your initial investment.

Using March Madness as an example, you could buy a virtual bankroll of $3,000 for $129, bet heavily on NCAA tournament games without risking more than that initial investment, and take home an increasing percentage of profits if you progress through the game’s levels.

Bankroll U is offering an exclusive $5 buy in for a $200 bankroll for all new members during March Madness. So if you want to take some shots this week and protect your downside, give it a try and let me know what you think!

Bet of the Week $$

Last week we lost one unit in total, winning with Duke taking down the ACC tournament title and losing our other long shot conference tournament bets.

Arizona, Wisconsin and Tennessee all made runs to the finals of their tournaments, at +650 or longer odds, but none were able to bring home a trophy.

Even so, since starting the newsletter, bets given out in this section are ahead 14 units, at a positive 22% ROI. We’ll update this regularly.

Based on my research, I am making the following NCAA tournament bets this week:

• Grand Canyon +10.5 on FanDuel @-106 for 1 unit

• Grand Canyon +455 on DraftKings for 0.1 units

• Liberty +7.5 on BetMGM @-118 for 1 unit

• Liberty +260 on ESPNBet for 0.2 units

First round NCAA tournament upsets are happening more frequently than they used to (Nate Silver wrote a great piece on this).

I’m leaning into that trend and playing two teams that I believe have the capacity to win outright against Big Ten teams that proved vulnerable to unexpected upsets during the regular season.

Next week we’ll stay focused on college basketball and find opportunities in the Sweet Sixteen.

Please fill out the survey about today’s newsletter and let me know your thoughts!