The 2020 NCAA Division I Men’s Basketball Tournament, as most of us know by now, was cancelled. This is extremely unfortunate because it has become a national past time where people of all ages and experience level get a sheet of paper in front of them and try to decide who will win and who will lose in a 64 team national tournament.
And while we won’t get to taste the madness this year, it is never too early to start preparing for next year’s tournament!
There are always plenty of tips available for filling your brackets and a lot of “experts” pop up at this time of the year. But with 64 teams in the mix and very little time to prepare, it’s extremely hard to be right. In fact, the numbers are shocking! (more on that later)
In this article, The ScoreMetrics Lab will look at how one can approach this challenge in an analytical fashion.
March Madness bracket scoring
Let’s begin by reminding ourselves of the scoring rules for March Madness brackets pools.
Most often, players get awarded points in pools for each correctly chosen matchup and the amount of points awarded increases for each round of the tournament.
A common scoring system for correct picks looks like this:
First Round: 1 point
Second Round: 2 points
Sweet 16 : 4 points
Elite 8: 8 points
Final Four : 16 points
Championship: 32 points
Obviously predicting the teams that go far in the tournament is important, since that’s where the biggest points are available.
Now, there are various estimates out there on the chances of someone picking a perfect bracket. Some mathematicians say it’s one in 2.4 trillion, while others go for 9.2 quintillion. We’ll just conclude that it’s impossible.
So, you’re not going to get all your picks right this year. But what really matters for picking a good bracket?
Let’s try tackling this challenge with a ScoreMetrics oriented approach and see what happens if we remove emotions from the equation and attack the bracket with a system.
A simple March Madness bracket system that beats most users
So far, we’ve established that it’s really hard to hit your picks at a high rate. And unless you are an expert in college hoops, it will probably take more time than you have available to develop a sound analytical approach to making picks for each matchup.
The question then becomes, does an easy strategy exist that beats the average player?
Turns out it does. In an article on the NCAA’s website, Daniel Wilco covers one in detail. And this one’s extremely simple: blindly picking the better-seeded team in each matchup has resulted in far better than average results in the NCAA’s own Bracket Challenge Game throughout the years.
Here’s what the NCAA’s data looks like for this seed-based system from the past nine tournaments in comparison to the performance of average users:
SEED-BASED BRACKET SCORE
AVERAGE USER SCORE
Most years, this system beats the average user by a country mile (or, by exactly 20.4 points on average, if we want to get a bit more analytical here).
Here’s the breakdown of the number of correct picks in those brackets, also from the NCAA’s site:
TOTAL POSSIBLE PER YEAR
The 2012 and 2013 scores are very high here since the number one overall seeds (Kentucky and Louisville) won the tournament in those years. You’ll notice that there were only two other years in which there were less correct first round picks than in 2012 and 2013 – a good indicator for the importance of having teams in your bracket that end up going far in the tournament.
Having correct picks in the final rounds makes all the difference, so going for the early-round upsets can be fatal to your bracket. The number one seeds are historically twice as likely to make it to the Final Four compared to any other teams. So, if you do go for an underdog, you better have solid reasoning behind the decision (more on this below).
Getting advanced with your March Madness bracket
Was the system above too simple for your liking? Or maybe the people in your bracket consist of college hoop experts and analytics geniuses?
While picking the better-seeded teams will most likely rank you in the upper echelon of players in your March Madness bracket pool, it likely won’t land you first spot against skilled opponents.
Here are a couple of ideas on getting more advanced with your strategy, if you’re willing to put some time and effort into your March Madness brackets in 2020:
1. Utilize data and existing predictive models
Many smart and experienced people are working on predicting results for the tournament, and a lot of the data they come up with is publicly available. Taking advantage of this makes a ton of sense, instead of just trusting your intuition or relying on tips from your buddies.
For example, the FiveThirtyEight publish a probabilistic March Madness forecast based on a model that combines data from sources that have historically been strong indicators of success in the tournament. Using this type of model as a base for your analysis makes things easier and ensures that you’re working with something factual.
Other sources for quality predictive data? Take a look at the futures odds bookmakers are giving to teams in the tournament. They’re pretty good at what they’re doing. But don’t forget that public money has an effect on the odds!
2. Find the right spots for contrarian picks
This is kind of a continuation on that first point we just made. Besides understanding which teams have the best odds of making it far in the tournament, it’s very advantageous to know who your opponents are picking to find the undervalued teams that have a good chance of playing in the later rounds.
For example, ESPN have a “who picked whom” feature for their March Madness bracket challenge, which allows you to see what percentage of users picked each team for each round of the tournament.
A simple way to find undervalued teams with that sort of data would be to compare the percentage chance a team’s being given in one of the forecast models to the percentage of users having picked that team.
To illustrate the point, let’s say that the FiveThirtyEight model gives team X a 15% chance to win the tournament, but only 2.5% of users have picked them to win it all. This would mean that team X are undervalued by 12.5% in this case.
Hopefully this article has been helpful in demonstrating some ways in which players can approach March Madness bracket pools in an analytical fashion.
The ScoreMetrics Lab does not recommend this as a proper investment, but it wishes you good luck for the tournament if you choose to play!
And before we go, don’t forget to have a look at our head trader John Todora’s new book – “Zero Correlation Investing – The Score Metrics Secret”, which is currently on sale for a limited time. Get yours now and learn how to build profitable sports betting systems!
The ScoreMetrics Lab is the engine that runs the Sports Trading System operation, consisting of a team of researchers and writers who are constantly testing and retesting algorithms. They work hand in hand with our Head Trader and Creator of ScoreMetrics, John Todora to help find new breakthroughs and develop new systems.