The Metrics Takeover: How Math is Shaping March Madness and Why the Creators are Uneasy
College basketball’s annual spectacle, March Madness, is increasingly being influenced by sophisticated statistical analysis. What once seemed like a human-driven process of selecting the 68 teams for the NCAA Tournament is now heavily informed by complex metrics. While these data-driven approaches aim to bring more objectivity to the selection process, they’ve also sparked debate and a sense of unease among the very people who created them.
Bart Torvik, a personal injury attorney from Evanston, Illinois, is experiencing this firsthand. His college basketball advanced statistics and rankings website, barttorvik.com, which he started as a hobby in 2014, has become an official component of March Madness. Starting in 2025, Torvik’s "T-rank" will be featured on NCAA Tournament men’s basketball selection committee team sheets, a testament to the website’s growing influence.
"This started as a hobby. It’s still supposed to be," Torvik admits, acknowledging the increased scrutiny and spotlight that comes with official recognition. The website, which once served as a tool for a Wisconsin basketball fan to analyze Big Ten team performance, now generates millions of pageviews, and even garnered a mention from Alabama coach Nate Oats at a Final Four news conference. Torvik’s website, unique in its entirely independent nature, being run off a server bought from Amazon and without corporate ownership, almost crashed on Selection Sunday the past two years. This highlights the demand for the information his site provides.
The incorporation of metrics like Torvik’s into the March Madness conversation is part of a larger trend. Over the past two decades, terms like NET, KenPom, and Torvik have become commonplace in the sport’s vernacular. Fans, coaches, broadcasters, and bracketologists alike rely on these rankings to differentiate between teams ahead of Selection Sunday.
However, this reliance on metrics isn’t without its critics. A virtual roundtable organized by the NCAA brought together representatives from the seven metrics used by the tournament selection committee, including Torvik, Ken Pomeroy (KenPom), Kevin Pauga (KPI), Matt Morris (ESPN), and Alok Pattani (Google).
The participants agreed that the NCAA had improved the selection process by moving away from the outdated Ratings Percentage Index (RPI), developing the NCAA Evaluation Tool (NET), and embracing a variety of ratings systems. However, they also expressed a shared concern: that only some of the seven metrics should be used to pick the teams.
Interestingly, Pomeroy and Torvik themselves suggested their rankings shouldn’t be used for direct selection. Even Pattani, who helped create the NET, didn’t advocate for its sole reliance.
"It’s a little weird I’m on the team sheet," Pomeroy admitted, explaining that his system is more predictive and shouldn’t be used to pick teams based solely on their predictive potential, but on their accomplishments throughout the season.
Despite these reservations, the public often focuses on a team’s NET ranking around Selection Sunday, often losing the nuance of each ratings system.
The seven metrics on NCAA team sheets are categorized as either predictive or results-based. The NET, KenPom, ESPN’s Basketball Power Index (BPI), and Torvik ratings are considered predictive, focusing on a team’s offensive and defensive efficiency. ESPN’s strength of record, the Kevin Pauga Index (KPI), and wins above bubble (WAB) are results-based, evaluating a team’s resume based on the difficulty of its schedule and opponents.
WAB, a new addition to the team sheets, is generating significant interest. Its creator, Seth Burn, believes that if selection committee members just use WAB, it can simplify the process and guide them in making their selections.
While the general principles behind these metrics are public, the exact formulas remain proprietary, even though they are largely similar due to the shared reliance on common input data.
The experts at the roundtable emphasized that results-based metrics should be the primary drivers for team selection. However, it remains unclear whether the selection committee shares this view.
The 12 members of this year’s selection committee, including athletic directors and conference commissioners, either declined to comment or did not respond to inquiries about how metrics would be used in the selection process.
NCAA director of media coordination/statistics David Worlock emphasized that the committee has more information than ever and the NCAA is working to educate them on the meaning and appropriate use of the data. He also suggested the data is currently being used more effectively.
The NCAA maintains that the NET is merely a primary sorting tool, not the ultimate determinant. The selection process still involves human judgment, with committee members casting votes based on factors such as who a team played, where they played, and how they performed.
However, Torvik and others believe that the NET carries significant weight, particularly because of its official affiliation with the NCAA. “Some of the fault lies with the people in charge in that why are these ratings on the team sheets if they’re not being used, and I think the fact is they are used, especially the NET," Torvik said.
The creation of the NET can be traced back to a 2017 meeting in Indianapolis that brought together Pomeroy, Pauga, sports reporters, bracketologists, selection committee members, and NCAA officials. This "summit" aimed to modernize the metrics used by the committee, as KenPom’s efficiency-based ratings had gained popularity among coaches and fans.
The NET was conceived as a way to blend predictive and results-based elements into a single metric. However, its debut was met with skepticism, with some critics calling it the worst rankings they had ever seen.
Despite the criticism, the NET quickly became central to bracketology discussions. It led to the four-quadrant system based on an opponent’s NET rating, which is used to evaluate the quality of wins and losses.
Even within the statistics community, there are varying views on how the NET is used and its potential flaws. Former Syracuse coach Jim Boeheim cautioned against over-reliance on quad one wins, emphasizing that a team’s performance throughout the entire season should be considered.
The NCAA has tweaked the NET formula over the years, including the use of efficiency-adjusted statistics and the removal of winning percentage. The NET does not include preseason data or scoring margin and weighs every game equally, regardless of date.
Despite these modifications, some believe that teams can manipulate the NET ratings by scheduling easier nonconference games to boost their efficiency numbers.
Pomeroy acknowledges the need to move beyond the RPI, but expresses some conflict, as he feels that the statistics community was not sufficiently included in the NET’s development, leading to a formula that is similar to his own.
ESPN’s Matt Morris considers the NET "inferior" because it doesn’t include scoring margin. He believes that if one wanted to use a metric to make money in Vegas, you would not use the NET rating.
ESPN bracketologist Joe Lunardi notes that the selection process is more "metric-dependent" than in the past, reflecting a shift in the composition of the selection committee and increased public awareness of the process.
The rise of available data has transformed bracketology into a year-round cottage industry. The Bracket Matrix, a website that tracks bracketology experts, has seen a significant increase in the number of bracketologists it tracks.
Kevin Pauga believes that the process has become more scripted and predictable due to the abundance of publicly available data.
The emergence of WAB, the new metric approved by the NCAA Tournament selection committee, is seen by some as a potential solution to the ongoing debate about how to use metrics effectively.
WAB measures how many more, or fewer, wins a team has against its schedule versus what a bubble team would expect to have against the same schedule. It is based on the NET and defines a bubble team as one ranked 45th in NET.
Seth Burn, the creator of WAB, is a former accountant who became a successful professional gambler. Pomeroy views WAB as a metric that’s "getting closer to the gold standard of selecting teams" and removing human bias from the equation.
Worlock believes that WAB "levels the playing field a little bit" for teams with fewer quad one opportunities.
Burn hopes that the selection committee will prioritize teams with better WAB ratings, potentially leading to a more objective selection process.
The next steps are uncertain, adding to the sense of unease surrounding the increased reliance on metrics.
"We’re going to know if they pick a team in the tournament ahead of a team from their very own metric that has a better resume," Burn said. "If they do that, that’s going to be the tricky thing for them to explain. If they don’t do it, then we’ve achieved victory. Then WAB has won. That’s what I’m hoping for.”
