A complex estimation algorithm used by commodities traders to forecast prices and the Federal Reserve Bank to estimate GDP growth has a new application: predicting how an athlete’s performance could affect a team’s odds of winning a game. The Quant Edge, a four-year-old venture started by long-time oil trader Todd Gross and backed by New York soccer legend Shep Messing, thinks venture capitalists will want to buy into the idea to the tune of $10 million.
The math is intricate, but the idea is simple enough: Kalman filtering, an algorithm used in machine learning (ML), takes many pieces of information, including some that are seemingly useless, and finds statistically significant connections among pairs of groupings of data points.
The Quant Edge is a sports betting service where bettors can tweak what they think a player’s impact will be on a game and get predictions on how that will affect the over/under and which team likely will win. “We’re using our quantitative analysis in a lot of different ways to bring something to the bettor—we’re helping quantify what they think,” Gross said on a video call.
Gross, who has written chapters in textbooks on commodities trading, came to sports betting after exploring numerous avenues on how to better apply Kalman filtering to financial markets. The sudden opening of the U.S. sports-betting market made him realize the math could be used to predict team and athlete performance—a field not burdened with the regulatory oversight of trading. Kalman filtering isn’t perfect, but it tends to be more accurate than traditional estimation methods because it finds connections a person may never notice—or which would take months of painstaking work to uncover. Use of ML like this is expanding, finding a foothold in everything from figuring out if some consumers are better credit risks than their FICO score suggests to helping NASA identify distant planets.
The Quant Edge allows bettors, who pay as much as $50 a month for access, to tweak the performance of a select group of players—will the quarterback have a poor game, an average game or a great game?—and see how it should affect their bet. The company offers player-impact predictions for NBA games, too.
Gross says people who use the system regularly have more winning bets than losing ones; The Quant Edge’s website touts a 65% success rate “when using player impact tools effectively.” That means don’t bet the house on one game, because ML offers probabilities, not certainties. “We’re fortunate over the past four years in NFL to have really good results,” Gross said.
Recently Gross brought on as investors Shep Messing, best known as goalkeeper for the Pele-led New York Cosmos, and sports marketing expert Rob Striar, whose M Style Marketing has worked extensively with the NHL and who recently bought an indoor soccer team with Ronaldinho. The pair bring a sports expertise that Gross believes his group of hedge fund and commodity trading veterans need. “The guy who originally wrote the model didn’t even know the quarterback threw the football,” Gross said.
The trio now have retained Livingston Securities to raise a $10 million venture capital round that Gross says will go toward developing player impact modelling for global soccer players as well as providing more granular options for NFL and NBA players. That means instead of evaluating the odds if James Harden has a ‘poor,’ ‘average’ or ‘great’ game, bettors could choose numerous values along the spectrum.
The addition of soccer would give The Quant Edge offerings in the three most-bet sports in the U.S. and the most-bet sport worldwide. Regardless of the math wizardry, Gross says his broad patent on the player impact concept gives the business a moat to fend off competitors.
Beyond a more robust subscription service, Gross thinks the expansion and merging of sports betting and content could allow The Quant Edge to create related programming too. “As sports books go for 24-hour news channels, they’re going to need pieces of relevant content to fill all that time … People can quantify what’s going on and bet, and the [bettor] banter back and forth can be a really powerful piece of content that is very timely,” the executive said. “I think we have something very compelling.”