Artificial Intelligence Hot Topic in Racing Circles

Considering the players in the racing and breeding industry who best understand information have always been poised for success in this industry, it's safe to say artificial intelligence figures to impact this sport in big ways. From the breeders choosing stallions for mares, to bloodstock agents understanding every detail in a sales catalog, to horsemen organizing their workforce, to racing secretaries writing race cards, to bettors finding overlay selections, AI already is shaping the racing industry and that trend figures to rapidly increase. That reality, as well as the question of how quickly racing will be able to effectively embrace AI, has made artificial intelligence a hot topic for the industry. This column brings many of those thoughts together, from both events I listened to as well as press releases issued from the events. At the National Horsemen's Benevolent and Protective Association conference earlier this month, John Julia brought expertise from both sides of the equation. On the equine side, Julia is an owner/breeder who previously served as a board member for the Pennsylvania Thoroughbred Horsemen's Association. On the technology side, he's also a senior vice president at Pinnacle AI. Julia outlined the strengths of artificial intelligence, which very much sound capable of solving some of the mysteries of Thoroughbred breeding. "Machines are very good at analyzing data at massive scale, finding patterns and correlations between different data points to yield an output. Sometimes that output is predicted in nature," Julia said. "But what machines are not good at is developing strategy; understanding how to take those business insights that we're getting from the machine and apply them to everyday life." Last fall at the Asian Racing Conference in Saudi Arabia, Kitman Labs founder Stephen Smith outlined the phenomenal changes in human athletic performance in the past 30 years as well as the lack of a proportionate progress in Thoroughbreds, pointing to the fact that Secretariat still holds the record time for the Kentucky Derby (G1) in 1973. Kitman Labs develops training methods to improve human athletic performance. Smith said breeders and trainers who embrace tailored, data-driven pathways could enjoy a significant advantage over competitors still using traditional approaches. "I think we can very, very quickly take the learnings we have taken from the human performance world and apply them to equine performance," he added. "We can look at progression rates from every stage of an equine athlete's career and start to analyze the masses of information that have been collected." At the HBPA conference, racing executive Rick Hammerle said AI could deliver information that could improve condition books. "Let's say that you've been writing a race for every 18 days," Hammerle said. "It now suggests that 23 days is the (better) number between races for that, something along those lines. Maybe you thought about writing the race going a mile. Maybe it suggests that if you had it a mile before, maybe you should have it back at a mile-and-a-sixteenth. The more data we get, this could help in improving condition books, races moved maybe two days or a week." Also at the HBPA conference, Equibase president and COO Kyle McDoniel said the company has used AI to create more uniformity in chart caller terminology. He noted that North American racing's commitment to information gathering could position it well as, comparable to other sports, participants and fans call for more data. "How can we come up with more stats to make horse racing easier and better to understand?" McDoniel said. "Stats like ROI (return on investment); they're just not really good stats. There are ways to improve upon them." On a February conference call with analysts and investors, Churchill Downs Inc. CEO Bill Carstanjen noted that because pari-mutuel wagering takes place against other players in the pool, and not the house; CDI would like to see its bettors on the TwinSpires advance-deposit wagering platform win more often. He said with that goal in mind, the company is putting AI tools in customers' hands. "For those of you with a TwinSpires account, you'll see the little button in the right-hand corner. That's an AI product that gives you an analysis of each race. It gives you some of the attributes and some of the indicators to look for in each race," Carstanjen said. "We rolled that out to five tracks and we plan on expanding it and improving it and making it even more robust. "We're also working on another tool that we expect will be delivered at some point in the future, which is completely interactive. You'll be able to talk and ask any questions that you have on a specific race or a specific horse. So those are big priorities for our business, and things we think we can deliver to our customers to make them better. We have every incentive to want to do that." Beyond assisting wagering customers, Carstanjen said CDI is looking to use AI in several areas. "In terms of as a business as a whole, we're looking forward to what vendors and what providers can do out there to make us better at marketing, to make us better at cost management, to make us better at acquisition of everything that goes into our business. So we embrace AI," Carstanjen said. "We think it's incumbent on us to find ways that it can help our business and we'll continue to do that." According to experts, it's imperative that racing move forward in these areas. But for anyone who remembers press releases from tracks boasting about going to HD video signals about five years after that stopped being newsworthy, the industry doesn't always have the greatest track record for adopting new technology. At the Asian Racing Conference, AI expert Danilo McGarry said it's imperative racing is not left at the gate this time around. "You don't have forever to transform. I would say most of your organizations have about 2-4 years to transform because the people that you serve are changing and they want different things tomorrow than they did today," McGarry said. "There are new sports coming up every 6-10 months. New sports that understand attention, marketing, and entertainment, and they are taking the attention away from your potential new customers that would like to go and see horse racing. "You are competing with more sports than you ever were before and this will not stop. What you do as organizations to transform the way you market and transform the way you create that immersive experience has to accelerate. I believe if you don't do this over the next few years, you will start to see a massive decline in the amount of interest in horse racing as new sports become interesting and market themselves better." Julia said it's great to see the industry thinking about ways it can embrace AI, but he noted that it needs to make up ground in a number of technology areas. "We are woefully behind as an industry from a technology standpoint, and it's great that we're sitting here talking about artificial intelligence, but technology in general, even before we got to this point, has lapped us multiple times over," Julia said. "There's a habitual tendency when something new comes out, like a shining object. We're talking about artificial intelligence running around saying, 'Hey, where can I apply AI? Where can I apply AI?' And that's really not an approach that I would recommend. "What I would suggest is that artificial intelligence is a very promising and powerful tool. The insights and the things that we can do from understanding data around us, regardless of where we get that data from, the internet, from sensors, whatever the case, allows us to make more intelligent decisions. But there's a lot of things that we can do right without using AI to modernize."