Don’t let love for your college alma mater derail your good judgment when filling out a March Madness bracket this week. An unemotional, data-driven approach is the key to a successful showing in your office’s NCAA basketball tournament pool, according to one consulting firm.
Americans will wager an estimated $9.2 billion on more than 70 million brackets this year, according to the American Gaming Association. The average person fills out two brackets and bets $29 per bracket. Nearly everyone allows hidden biases – often as small as the color of a team’s uniforms or an overreliance on a particular expert – to guide their tournament selections.
“People often have these behavioral biases – things we don’t even know are going on with us, that we’re weighing subconsciously, that lead us in the wrong direction,” Timothy Murphy, a research manager at Deloitte Consulting in the Center for Integrated Research, told FOXBusiness.com.
Murphy and his fellow data scientists at Deloitte advocate the use of behavioral analytics. The same data-driven principles that firms use to analyze insurance risk or assess possible investments can be applied to the creation of a successful March Madness bracket. That means removing emotion from the equation and relying on empirical evidence, not hunches or gut feelings, to make educated choices during the NCAA tournament.
Any number of subtle mistakes can ruin a March Madness bracket before the end of the first round, but sports fans tend to fall victim to three common pitfalls when making their selections, Murphy said. One error is the average sports fan’s tendency to place too much confidence on a “hot” team’s recent winning streak, at the expense of an overall look at the team’s quality over the course of a full season.
Known as the “hot hand fallacy,” the theory is based in part on a 1985 study that analyzed whether fans thought a player in the midst of a “hot streak” was more likely to make their next shot. The vast majority of those surveyed believed in “streak shooting,” even though there’s little evidence that recent success in a particular skill-based activity guarantees success in the future.
A winning streak in a conference tournament or near the end of the college basketball season doesn’t necessarily mean a school is headed to the Final Four.
“When you’re making your picks, you shouldn’t look at is as, ‘well, this team won three of their last four games.’ It should be looked at as a full body of literature,” Murphy said. “People that look at these streaks aren’t looking at the bigger picture. They’re creating a story out of random and expected streaks that will come with a large-enough sample size.”
Bracket participants also tend to rely on an individual expert when making their picks, rather than a pool of experts. Even the most respected basketball analysts will get a few games wrong during a tournament as random as March Madness.
When choosing which expert’s advice to follow, fans should look for analysts that are known to use a data-driven approach to their selections, such as famed statistician Nate Silver of ESPN’s “FiveThirtyEight” blog. Better yet, combine the advice of multiple experts.
“Use a pool of qualified experts. You don’t want to use just one,” Murphy said. “They might be hot this year, but they’re almost [always] going to regress to the mean of the general knowledge pool over time.”
Finally, Murphy says it’s important not to let rooting bias affect decision-making. A University of Wisconsin graduate may have a natural instinct to place the Badgers in the Final Four, even if their team has noticeable flaws. The tendency to focus on a team’s positives while ignoring their weaknesses is called the “halo effect.”
Instead of making an emotional choice, would-be bracketologists should come up with a data set that they think best correlates to March Madness success. Murphy suggests ranking teams based on a few stats, such as winning percentage against top-25 teams, and then covering up each tournament team’s name to guarantee an unemotional selection.
“There’s no ‘secret sauce’ to making every pick right,” Murphy said. “We’re just trying to remove our biases and be better at it.”