Will machine learning and AI take the fun out of racing?
There is a lot of talk in certain technology circles about the impact of machine learning and artificial intelligence in sports. Mostly, there is a consensus that sports will be disrupted by new technology, and that consensus sees it as broadly beneficial. In the case of horse racing, you might imagine a situation where a trainer relies on an algorithm to tell them to pull a horse from a race due to the extreme likelihood of injury. Similar technology is already being used in football. The tech could go a lot further, too, as machine learning develops into a more sophisticated industry.
There is, of course, always a lot of reticence about adopting new technologies. Look at how initial perceptions of Martin Pipe's training methods were so negative, yet everyone follows his lead now. For some, there is an admission that scientific application somehow dilutes the artistry of the sport. We might have sat back and marvelled at the technology behind the Virtual Grand National over the last few years, but it’s clear that most horse racing fans prefer the real thing.
Technology will skirt the edges of horse racing – for now
The good news is that most of the impact of machine learning and AI will most likely be seen on the periphery of horse racing, at least for the medium term. We've already seen it in broadcast innovations for the sport, for example. We mentioned trainers using data-driven models to determine whether or not to run a horse. While sceptics might say it’s better for an expert trainer to judge for themselves, such technology would be complementary to common sense and experience. But it’s easy to see the use cases, particularly when it comes to animal welfare.
We are also likely to see machine learning involved in racing tips. Today, of course, the best tipsters use data to make decisions, but the capability of computers to crunch the numbers is extraordinary, and it’s getting better. We have seen evidence already with IBM’s WATSON supercomputer. It uses machine learning to analyse millions of pieces of data before making predictions on fantasy football (American football). Other supercomputers are being employed in other sports.
Of course, that begs the question – would such technology take all the fun out of racing. The average punter heading to this year's Avon Vale will grab their race card, perhaps a copy of the Point-to-Point Planner, and then parse out the information before making a selection. It’s part of the ‘game’ that’s played between punter and bookie on every racecourse around the world. If the attendees at the races simply consulted a computer before each bet – wouldn’t that such the joy out of the day at the races? Many would think so.
Hard to beat the real thing
Indeed, going back to what we said about the Virtual Grand National, the event has been a tremendous success since its inception a few years ago, even replacing the real thing in 2020. And some punters warmed to the idea of trying to beat the algorithm. But, for the majority of us, our love of racing comes from the unexpected – the stuff you can’t squeeze into an algorithm.
There is certainly no reason to fear technological disruption in horse racing. “Disruption” is not synonymous with replacing. However, there is also an understandable need to pay attention. Billions are pouring into the development of virtual sports products, much in preparation for the so-called metaverse. Man City, for example, is working with Sony to build a virtual stadium – might the same happen for Aintree or Ascot?
In the end, though, the world’s great races like the Melbourne Cup and Grand National will always take place, and those events will be able to live with any tech disruption. But technology has particular challenges for the grassroots level of horse racing. Will AI be affordable to connections looking for an edge over their rivals, and what will it add? Most notably, will it impact the traditions and values of those who support the structure from the bottom up?
