DeMarcus Cousins, star center for the Sacramento Kings, warming up before a 2012 game. Image via Wikimedia Commons
DeMarcus Cousins, star center for the Sacramento Kings, warming up before a 2012 game. Image via Wikimedia Commons

As Vivek Ranadivé, CEO of TIBCO Software, and his tech industry partners take over the NBA’s Sacramento Kings, they’ve made some comments about applying advanced technology to basketball. But they haven’t really elaborated.
Well, I can provide some clues to the way Ranadivé will likely think about this, and it goes well beyond using Moneyball-style data analysis to find players who are overlooked gems.
Two years ago, Vivek and I were finishing up a book we co-wrote, The Two-Second Advantage: How We Succeed by Anticipating the Future…Just Enough. The book grew, in part, out of one of Ranadivé’s oft-repeated aphorisms: “A little bit of the right information just a little bit beforehand is more valuable than all the information in the world six months afterward.”
Those words are a shot at databases, which store mountains of past information, and the analytics that piece together stored data looking for trends. Ranadivé—along with a lot of other computer people—believes technology has to evolve to comprehend events as they happen, and use something akin to experience to anticipate what’s about to happen next. Searching databases is too slow. Prediction is the new grail.
This idea also happens to reflect the way human intelligence works. We don’t search our database of memories to try to understand what to do next. We constantly use memories to build efficient little models of the world—in other words, to gain experience. As things happen in front of us, we use those models to predict what will happen in the next second or two. It’s how a musician knows what note to play next, how a comedian can time a laugh line, and how you can apply the right pressure to pick up a bottle of water without first testing the weight.
Our book first examined new research into human intelligence and predictiveness, and applied those ideas to computer science. The book then described ways predictive computing could work in real-world situations.
At the time we wrote the book, Ranadivé was a minority owner of the NBA’s Golden State Warriors—so one of the real-world situations we dove into was basketball. Now that he’s the lead owner of the Kings, he’ll have more of a chance to implement his vision.
Some of the more near-term ideas from the book have to do with fans. Season-ticket holders are usually, by definition, loyal fans, so they’d likely be willing to sign up for a loyalty card that they’d swipe anytime they buy something at a game. The card lets the technology “see” the fan and start to understand his or her behavior, and pieces together experiences about that fan.
A more futuristic (but not too futuristic!) concept would be applied on the court.
New capabilities are constantly being invented to “see” and tag movements and events in sports. The San Francisco Giants and other baseball teams have used Sportvision to track the movements of the ball and players on the field, showing patterns in the play of fielders. Some of the world’s top soccer teams use similar technology to track the speed and distance each player runs, which can help a team anticipate when in a game a certain player will get tired.
In time, cameras will be able to watch an NBA game and tag events in real time—so the team’s computer system can have a constant flow of data about what’s happening in a game right now, from player movements to shots taken to fouls committed and more.
In Ranadivé’s predictive-technology vision, the team’s computers will use that data to learn—to gain experience—about not just the Kings, but also their opponents. Then it could start making predictions during games about what is likely to happen.
So an assistant coach could carry an iPad wirelessly connected to the team’s system, which might recognize subtle patterns and conclude that a certain player is tired, or the other team’s star is likely to try a certain shot in the next minute. The system could make some predictions that the coaches, with all their experience, still might miss. The coaches could then adjust their strategy mid-game.
Essentially, the goal is to create technology that has the real-time gut instincts of an experienced coach. That’s a big leap from Moneyball analysis, which looks at past data to find out if a player might be better than most humans think.
As we said in the book, such technology is still years away, but knowledge of how the brain works—its predictive capabilities—is influencing computer scientists at places like IBM, MIT, and labs around the world. Certainly it will come to high-level sports before long.