Can a centuries-old sport really be revolutionised? If anything can do it, it may be AI. Here are some of the ways artificial intelligence is shaping horse racing.
How intelligent is AI, really?
Some of the supposed advantages of AI may sound more speculative. One article on the subject states, “One of the most revolutionary changes that AI will bring to the horse racing industry is the injury prevention model.” If you’ve come across the uses of AI in human medicine, it may be believable that predictive models can effectively analyse “various parameters like a horse’s vital signs, movement, and past performances”.
In fact, AI is already being used in injury prevention. HISA, the Horseracing Integrity and Safety Act, has collaborated with Amazon Web Services (AWS) to “apply data analytics, machine learning and artificial intelligence with the goal of enhancing equine safety by analyzing the factors contributing to injuries” (per Thoroughbred Daily News).
AI can monitor horses’ health and optimise training. Predictive analytics and motion sensors help detect signs of injury and fatigue. Monitoring vital signs (respiration, heart rate, etc) can help detect potential health risks before they become major problems.
HISA CEO Lisa Lazarus said the insights from big data and machine learning can help solve horse racing’s “most pressing equine welfare issues”.
Betting
AI’s biggest impact in horse racing, though, will likely be in betting. It is already significantly changing how races are analysed by both bookies and punters. Machine learning models can analyse huge datasets more efficiently than humans.
In 2023, Financial Times’ Oliver Roeder spent time with Scotty “Pick Six” McKeever to learn how he and others were using AI to bet on horse racing. McKeever opts for a more “empirical” approach than bettors who rely on personal relationships with trainers and jockeys. His AI app EquinEdge is just one of many in a sport Roeder describes as “awash in AI-assisted technology.”
CAWS (computer-assisted wagerers) are hugely impacting betting. Although there are only a predicted four “very large” CAWs in the US, they’re dominating pools. CAW teams use AI to place tens of thousands of bets, adjusting to new information (past performance, weather, jockey stats, etc) extremely quickly. Casual bettors now compete against algorithms. The algorithms are far more efficient at identifying value, so they often shift the odds, making it more difficult for regular bettors to find good returns.
The Elite Turf Club (based in the tax haven of Curaçao) are estimated to account for around 20% of horse-racing betting in the US. The president of the computer syndicate, Scott Daruty, describes it as a “wagering platform that services some of the largest bettors in the world”.
Patrick Cummings of the Thoroughbred Idea Foundation has written, “These [CAW] entities bet big because that is what the math dictates. This is Wall Street meeting horseracing”.
McKeever, meanwhile, describes his work as “for the small guy”, the “Joe Q Horseplayer”. He aims to give them all the information he and his 13 employees and contractors have collected so that the “little guy” can be taken care of.
Roeder spoke to other industry experts and concluded the most effective bets were those that layered human expertise on top of the CAWs.
A changing culture
AI has influenced the sport’s culture. For decades, old-school intuition ruled. Now data science has led to a tension between tradition and modernisation.
At times, they meet in the middle. When FT’s Roeder spoke to David Bernsen, the manager of a collective of CAW players, GWG Group, the latter said they “need the public’s input. The best predictor of the win pool is the general public — just wisdom of the crowds.”