AI experiment in halfpipe judging at X Games will give snowboarders a glimpse into the future
The X Games will experiment judging halfpipe runs this week in Aspen using artificial intelligence, the cutting-edge technology that could someday play a role in the way subjectively judged sports are scored.
Long a trendsetter in action sports, the X Games and its new CEO, freestyle skiing great Jeremy Bloom, teamed with Google founder Sergey Brin to build the technology.
Using Google Cloud tools including Vertex AI, Bloom thinks this experiment has potential to change the game on halfpipes, then maybe on slopestyle courses, skating rinks and anywhere a judge is used to score a contest.
“Part of subjective sports, we see it all over the place, is that even at their best, humans can get it wrong,” said Bloom, who was a freestyle skier at two Olympics while also playing college football at Colorado. “Sometimes getting it wrong has huge implications. What if we could give judges superpowers and they could see things they couldn’t see with the human eye, and this technology could help inform them?”
The specter of a judging mistake lingers over every high-stakes contest, and even with its more laid-back vibe, snowboarding, which is now a fixture in the Olympic program, is no exception.
At the last Winter Games in Beijing, the sport narrowly avoided a potential scandal in the men’s halfpipe final. Japan’s Ayumu Hirano landed the most difficult trick in the sport — a triple cork — as part of a strong, top-to-bottom run, but was ranked behind another rider who didn’t do the trick after two rounds.
Snowboarding experts were aghast on social media. Had Hirano not pulled off the trick again on his third round, his score from Round 2 wouldn’t have been enough for the gold medal he eventually won.
In another episode, Canadian slopestyler Max Parrot acknowledged not grabbing his board on a run that earned him a gold medal, a key element that judges missed but that could be picked up on close review of the video.
The AI at the X Games this week won’t have any impact on the official scoring, but will be a gauge of what’s possible in the future. Bloom said thousands of hours of halfpipe footage, along with the judging criteria, have been loaded into a system that will be shown on the TV telecast and be made available to the live judges.
The AI will be programmed to watch haflpipe practice and predict the top three finishers. Then its powers will be used to judge and commentate on three different riders as they go down the halfpipe.
“It’s early days, but the technology sort of blows your mind,” Bloom said. “It’s the power of what it can do when you give it clear direction. It’s pretty amazing what this thing can do.”
In snowboarding, judges look at elements like the height of the jumps, the difficulty of the tricks and how well they’re executed. They ultimately deliver scores on a 100-point scale based on how runs stack up against each other.
It’s a nuance unlike figure skating or gymnastics, where individual tricks have specific point values. Bloom says all that is being taken into account as they work through the experiment.
He does not envision a future with no human judges.
“I don’t think this replaces the judges,” Bloom said, “but I think it gives them power to ensure objectivity.”
___
AP Winter Olympics: https://apnews.com/hub/winter-olympicsAP