Inside the development of SEQNZR's new betting product

BetSEQNZR has attracted significant interest from sportsbooks seeking to license or acquire the technology.

While making the rounds and visiting clubs at MLB’s Winter Meetings last year, the team at SEQNZR, which makes tools to optimize lineups and make in-game decisions, heard the same advice from several baseball executives: “You guys should go into betting.”

The legacy SEQNZR product is used daily by the Cincinnati Reds and more than 100 college baseball and softball teams, but after repurposing the core simulation technology for sports betting, the offshoot product, BetSEQNZR, has attracted significant interest from sportsbooks seeking to license or acquire the technology.

BetSEQNZR is also one of five finalists in the Global Gaming Expo’s Innovation Showcase who will present in Las Vegas on Wednesday.

“We had to convince ourselves that it was worthwhile for us to go into it because we could either spend more time on the in-game side doing something that is proven and that we already had a product, or we could completely start from scratch and enter an entirely new industry that's super competitive,” SEQNZR CEO Brian McAfee said. “You've got a lot of big players in it that do really well for themselves.

“Within four months,” he added, “we had a product that was capable of pricing pregame lines, in-game lines, in-game parlays, and it’s measurably 1-to-3%, almost 4%, better than every single US sportsbook.”

Those favorable metrics are derived using a trading bot that SEQNZR built, which uses The Odds API to capture every sportsbook’s lines and compare them to the in-house probabilities using the Root-Mean-Square Error (RMSE) technique.

BetSEQNZR’s lead data scientist for building player projection models, Matthew de Marte, said that, based on conversations they’ve had around the industry, most competing suppliers use complicated machine learning models that don’t utilize as many batter and pitcher variables as BetSEQNZR. Their approach is more reliant on subject-matter expertise and advanced data sources.

McAfee played some minor league ball, and de Marte worked a few years for the Los Angeles Angels as a quantitative analyst before joining the R&D department of the KBO’s Lotte Giants. He is also, along with pro pitcher Dan Straily, a co-founder of Vaulted Baseball that has worked with more than 60 MLB pitchers in providing detailed scouting reports.

“We’re approaching it from the major league front office perspective, and we’re trying to pin down, ‘What is this player's true talent level?’” de Marte said, “and using that and putting it into a fashion where we can use it for the simulation. We’re, essentially, then letting the simulation figure out the odds for us.”

“We're also taking into account unique data sources,” he added. “Statcast data really isn't utilized a lot in the betting space to create odds. We feel like that's kind of our advantage.”

Furthermore, de Marte continued, when projecting a pitcher’s strikeout rate, the BetSEQNZR model draws on granular data to assess not just his recent success but the quality of the individual pitches he’s been throwing — a potential harbinger for changes in either direction.

“It's able to pick up on potentially underlying trends a bit faster than if you're only looking at observed information,” he said.

Many sportsbooks currently draw odds from multiple suppliers, McAfee noted, which leads to inefficiencies in pricing. The hope of BetSEQNZR is to be that one-stop-shop for all MLB odds. The company said it will be fully ready for deployment in time for the 2024 season.

“You really have to price it with a simulation,” he said. “If you use correlation math, it just gets too challenging to be confident in the true odds that you're offering. And so you end up with these massive holds on parlays.”

The meetings McAfee will hold at the Global Gaming Expo may well lead to a decision on whether to sell the IP behind BetSEQNZR or opt to license it. Either way, he said the plan is to split that part of the business off from SEQNZR, which has previously integrated or collaborated with 6-4-3 Charts, PlaySight, BaseballCloud, Driveline Baseball and The Alliance Fastpitch. (McAfee and de Marte both have day jobs as data scientists at Nike, a notable coincidence because they met before de Marte started his role and he applied independently.)

“SEQNZR is still going to live on, and we have some exciting conversations going on there,” McAfee said. “But it's so small in comparison to the betting world.”