U.S. Open and IBM: Analytics Grand Slam

Summary:It's US Open season, and the analytics are going wild (not to mention the fans). Here's an inside look at how IBM and the USTA are bringing the data to you, in your Web browser, and on your device.

Back to The Web
You don’t need an iPad to take advantage of the technology though.  For example, just head over to www.usopen.org on your desktop or laptop computer, and you’ll have access to most of the same features and even a few that you won’t get on the iPad.  For example, select VIDEO & RADIO from the top nav bar, then pick US Open Live from the drop-down menu, pick a court, and watch the feed.  As I write this I’m watching Angelique Kerber and Venus Williams play their second round match.  Since I’m on a computer, I don’t just want to watch the linear video and listen to the color commentary; I also want to do some data analysis superimposed over the live video stream.  As you might expect, IBM hasn’t let me down. 

First off, I can click on the Match Stats button and see real-time updated statistics for figures such as first serve percentage, number of double faults and unforced errors, and I can see these numbers calculated for the whole match or a particular set. Towards the upper-left of the screen, there’s a button labeled “Keys to the Match.”  If I click it, some very interesting data comes up.  This feature, which premiered during last year’s Open, can determine the three most important factors in determining victory in a given match, customized for each particular opponent.  You can see that same data on its own (rather than as an overlay on the live video) in the site's SlamTracker feature, available under Scores & Stats:

Fig2 Alt

These measures also include goal and status amounts.  In effect, they are key performance indicators (KPIs), and the Keys to the Match display is a scorecard, bringing the use of that term full circle, back to competitive sports.   But rather than straight Online Analytical Processing (OLAP), these KPIs are derived from performing predictive analytics on the last 7 years’ worth of data from all four grand slam events, totaling 39 million data points.

The analytics favored Kerber through most of the match, and she was ultimately victorious.  Although Kerber has a higher ranking than Venus, the results of the match were somewhat of an upset, which the IBM analytics correctly predicted.

This is really valuable data and IBM works with ESPN to supply the latter's on-air commentators with that data to keep the conversation moving.  But the neat thing is – in this age of the Web and interactive analytics – you no longer need to rely on the commentators to get it.  Instead, you just need a browser and a mouse -- or an iPad and your finger -- to be your own Grand Slam data analyst.

What’s behind the curtain
About two eweeks ago, I wrote a post on IBM’s Big Data prowess , including its product portfolio and the acquisitions that have filled it out.  So naturally, I was interested to know what back-end IBM products and technologies are in use at the U.S Open.  Here’s what I learned:

  • The SlamTracker technology (which includes the match stats and Keys to the Match) make heavy use of the SPSS technology that IBM acquired in 2009.
  • IBM’s stalwart relational database, DB2 is used heavily for scoring data and operations.
  • WebSphere MQ (fka MQ Series), IBM’s foundational message-based middleware is used for scoring delivery, enabling you to get scores even more quickly online than you can over the air.
  • The WebSphere family of technologies is used for the overall services architecture.

What’s most interesting to me about the above list is how all of the technologies in it are more than 10 years old (some of them well more).  Core statistical, relational, SOA and middleware technology has not become less important in this age of data and analytics.  And what’s especially noteworthy is that Hadoop, Netezza and Cognos Business Intelligence technologies haven’t even made the cut.  Hadoop, Data Warehousing and BI are of course important; but IBM’s application of conventional enterprise technology shows that Big Data- and BI-specific technologies are not necessarily prerequisites for good analytics implementations.

More to Come
I’ve shared a couple of screen grabs with you here, but there’s more to come, including some photographs from the USOpen.org Operations Center and more from the site and the apps themselves.  Look for that post ahead of the tournament’s final weekend so you can get yourself set up for multi-mon, multi-device analytics in time for the semi-finals and finals.

Topics: Big Data


Andrew Brust has worked in the software industry for 25 years as a developer, consultant, entrepreneur and CTO, specializing in application development, databases and business intelligence technology. He has been a developer magazine columnist and conference speaker since the mid-90s, and a technology book writer and blogger since 2005. A... Full Bio

zdnet_core.socialButton.googleLabel Contact Disclosure

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Related Stories

The best of ZDNet, delivered

You have been successfully signed up. To sign up for more newsletters or to manage your account, visit the Newsletter Subscription Center.
Subscription failed.