William Hill and Stats Perform have entered into a first-of-its kind data partnership that could have major impact on player prop bets and their probability over the next decade.
Under the partnership, the sports technology group will utilize artificial intelligence and machine learning techniques to enable William Hill to create unique player prop wagers.
The deal primarily involves the U.S. market. During the first phase of the rollout, Stats Perform will supply probability lines for William Hill’s NBA and NFL props. If successful, the partnership could be expanded to several other sports.
“For me, this is absolutely on the sweet spot of what Stats Perform will be in the future to the industry,” Stats Perform Chief Commercial Officer Andrew Ashenden told Sports Handle.
The probability lines are determined through Stats VQ, a platform launched in April that uses proprietary algorithms and A.I. to come up with more accurate prop bets. The launch came a week after Europe-based Perform Group announced a proposed merger with Stats LLC that sought to combine two of the world’s leading sports data providers. The merger was finalized in July.
“For years, sportsbooks in the UK and Las Vegas have been using player props and predictions created through traditional data,” said Dr. Helen Sun, chief technology officer at Stats. “Given the advancements in AI and machine learning, it’s time we harness the technology to improve overall accuracy.”
The data sets used by Stats VQ tracks every single player movement from all 22 players on the field for every single play of a game. From a backward-looking perspective, the platform can assess how a player has fared against the Broncos in press-man coverage, Cover-2, Cover-3, and other schemes in the past. More critically, the technology will ultimately try to predict how he will perform against Broncos corners he’ll be facing for the first time.
The data sets are too complex for humans to interpret, Ashenden said. But with machine learning-generated player predictions, Stats VQ can analyze vast amounts of data at rapid speeds in order to derive probabilities for what could happen next in a game.
“The applications for it are endless,” Ashenden said. “I think it’s the next evolution of our business.”
William Hill appears to be embracing the use of predictive analytics in sports betting. A team of data scientists from Stats Perform worked closely with the sportsbook’s analysts for about a year to ensure that the companies had the technological compatibility to complete the partnership.
“I think there is a lot of entrepreneurial activity around this space, as there should be,” said William Hill CEO Joe Asher. According to Asher, customers largely bet on the outcome of a game or a defined subset such as whether a team will score a run in the next inning. At some mobile sportsbooks, prop bets can comprise roughly 15-20 percent of an operator’s NFL handle on a given Sunday, the National Football Post reported.
It is also too early to gauge interest levels on microbets that ask customers to predict whether Steph Curry will hit his first 3-point attempt or if Justin Verlander will throw a strike on his next pitch.
“Clearly some people will be interested, but will it be a big segment of the population?” Asher said. “Time will tell.”
