IBM’s artificial intelligence has taken a break from digesting health records and directing films to bone up on tennis ahead of this year’s Wimbledon tournament. Among several announcements at a press conference this morning, it was revealed that the supercomputer has been fed 53,713,514 tennis data points, 6,349 Telegraph articles, and 11,208,192 words’ worth of Wimbledon annuals, interview transcripts and social media commentary in an effort to find what makes a great tennis player.
“From all of that, we’ve identified six traits of what it takes to be a great champion of Wimbledon,” explained Sam Seddon, Wimbledon client and programme executive at IBM. With the help of analysis from tennis experts, this came down to six points: passion, performance under pressure, fast or accurate serves, stamina, adaptability and the quality of the return.
The discovery that stamina, passion and racquet skill play a major part in things may not be the most stunning revelation for veteran tennis aficionados, but it’s encouraging to see that Watson can keep pace with the paid pundits after such a short spell reading up on things. And that’s just the beginning of Watson’s involvement in the annual tennis jamboree.
The IBM SlamTracker, along with providing fans with live real-time stats of the matches, also takes advantage of insights collected through historical data. Not only will fans be provided with specific player tactics to look out for ahead of a game, but close attention is paid to “pressure situations” within a match – how a player tends to perform when down 0-40 and down two sets, for example. “Within SlamTracker, there will be components to allow a fan to know what the likely outcome is if a player loses this point,” explained Seddon.
But Watson’s involvement in the tournament isn’t limited to that of being a back-seat pundit. First up, it’s getting back in the director’s chair to edit highlights of all the action from the six main show courts. That’s an average of 18 matches per day to sift through, which would be a serious undertaking for human editors. These highlights will be assembled using Watson’s assessment of player movement, match data and crowd noise to ensure the right highlights make the 90-second final cut – although a human will still be consulted to make sure nothing odd slips through the net (figuratively speaking).
Finally, Watson will be putting on its concierge outfit in the form of “Ask Fred” in the official Wimbledon app. Named after Fred Perry, the bot will be on hand to offer visitors advice on finding their way around the venue and dining options, should you be fed up of strawberries. As if anyone could ever be fed up of strawberries.
“Research has shown us that the people are either coming for the first time, and therefore don’t know necessarily what there is to do, or they’ve been here for multiple years and are very used to coming through the same gate and going to the same places they’ve always been,” explained Alexandra Willis, head of communications, content and digital at the All England Lawn Tennis Club (AELTC). The hope is that Fred will help both newcomers and old-timers to get the most out of their Wimbledon experience. “One of the great things about Watson and using cognitive in this way is that it learns, and hopefully it will pick up on the types of questions being asked that will then enable us to change what we do for next year.”
As someone who has spent a lot of time exploring the technology aimed at athletes lately, IBM’s continued involvement with Wimbledon is a timely reminder that pro wearables are only one side of the puzzle. Without fan enthusiasm, the money dries up, and pro sports analytics ceases to be the lucrative industry it has become. Watson’s growing interest in tennis will hopefully ensure that this is the most talked-about Wimbledon yet, no matter what happens on Centre Court.
Disclaimer: Some pages on this site may include an affiliate link. This does not effect our editorial in any way.
