iOS app Android app More


This is one of my favorite times of the summer. New York's U.S. Open Tennis Championships kick off today and build to a crescendo in two weeks with the finals. Every year, IBM's sponsorship of the championships gives us the opportunity to showcase amazing new technologies for some of the most sophisticated tennis fans in the world. During this year's championships, fans and broadcasters alike will be able to enjoy matches with a depth of understanding far beyond anything they have experienced at the tournament before. That's thanks to U.S. Open PointStream, a new match analysis feature on the U.S. Open Web site.

PointStream represents a great leap forward for tennis fans. Last year, fans received a wealth of statistical information about players and matches on the site. But now, thanks to PointStream, they can access deep analysis spelling out what each player needs to do to increase their chances of winning a match, how the match is going in real time and when the momentum is shifting.

PointStream also signals a new level of technical sophistication emerging worldwide that is deepening our understanding of nearly every human endeavor. Thanks to new analytics capabilities, people are able to gather huge quantities of pertinent information about nearly any topic, extract insights, and get up-to-the second updates about what's happening and why. At IBM, we call this the smarter planet.

Here's how you'll experience the magic of analytics on the U.S. Open Web site: When you click on the PointStream feature, up will pop a "dashboard" that displays key dimensions of an individual match, including an assessment of how the two players have performed against each other in the past. PointStream identifies three Keys to the Match for each player. While the match goes on, PointStream not only displays scoring as it happens and statistics such as aces, unforced errors and winners; but it tracks each player's on-going performance against the Keys to the Match. You'll see graphically who is performing as well as they must to win, and who is not. The interface also identifies potential turning points as they happen--and you will see a "confidence meter" that tells how confident PointStream is that it is spotting momentum shifts accurately.

Providing an experience like this is no trivial task. IBM gathered more than 39 million data points from play at the four Grand Slam tournaments going back five years. Engineers built a sophisticated database and use a powerful predictive analytics engine to identify the Keys to the Match.

PointStream is designed for tennis fans, but we expect commentators on the television broadcasts of the U.S. Open to make the most of the feature, as well. Both CBS and ESPN plan on using it. PointStream and its descendents will never replace a cracker-jack analyst like tennis legend John McEnroe, but they'll give him and his brethren a lot more to talk about during those dull moments when players are switching ends, taking long drinks of water or replacing their rackets at court side.

For more information about IBM PointStream, visit IBM's A Smarter Planet blog.