ESPN Tennis Analytics: Behind the Scenes

roddick-russell small

Entering the ESPN studio a few years ago, our consultant Kenneth Jensen brimmed with excitement. We were introducing a new tennis analytics solution to the U.S. Open broadcast team. Brad Gilbert, a former professional tennis player and coach to stars like Andre Agassi and Andy Roddick, expressed the most interest. In fact, the topic of conversation was the opening round match between Andy Roddick and Michael Russell. Given his history, Gilbert knew both players extremely well.

Our “SlamTracker” application is a coach’s dream. It analyzes 41 million data points collected over eight years of Gland Slam play, including head-to-head matches, similar player types and player surfaces. It then identifies the top three actions a player must achieve against any specific opponent to enhance their chances of winning.

As the meeting started, Jensen opened his laptop and began reviewing the analytics. He quickly pointed out that the most important factor for Roddick was winning at least 32% of his first serve return points. “That’s never, ever, ever gonna happen, because he doesn’t return well, and certainly not the first serve.” said Gilbert. “I don’t believe he has ever won 32% of his first serve return points.” He was adamant.

After what seemed an eternity of quietly listening and taking notes, Jensen felt embarrassed as he left the room. Why did the analytics produce an obviously incorrect result? He went back and started researching the data. He found two startling facts. First, in past Grand Slam tournaments Roddick won at least 32% of his first serve return points more than one-third of the time. Second, Roddick won the set 96% of the time he achieved this measure but only 53% if he did not. The analytics were, in fact, correct.

Did we convince Gilbert that Roddick’s return rate was the hinge point for his success? Would Gilbert or Roddick take advantage of this surprising finding? Watch our case below, along with the outcome of Roddick’s match against Michael Russell.

To learn more about IBM sports analytic solutions visit http://ibm.co/1jXLd1J

Tags: , , , , , , , , ,

Categories: analytics, Business, sports, technology, Uncategorized

About the Author,

Steve Harvey is the Worldwide Leader for Technology and Analytics across IBM's Business Process Outsourcing unit.

 

4 Responses to ESPN Tennis Analytics: Behind the Scenes

  1. I heard China Table Tennis team do have this kind analytics, but it seems they are using simple tools, not analytics, so this solutions really make a difference.

  2. Analytic is amazingly correct! I believe this is a convincing case which would help decision makers to realize the power of Analytic versus the limitation of human intelligence.

  3. Great testimony of IBM’s analytics capability. Analytics in sports are very powerful and under utilized.

Leave a Reply

Your email address will not be published.

Follow

Get every new post delivered to your Inbox

Join other followers:

The postings on this site are my own and do not necessarily represent IBM positions, strategies or opinions.
%d bloggers like this: