The annual convergence of the end of NCAA basketball season, the onset of the NHL and NBA playoffs, the lead-up to the NFL draft, and the kickoff of the major league baseball season have player stats and sports analytics front of mind for many fans and data nuts – myself included. But rumbles of discontent portend a possible storm on the horizon. It seems not everyone’s a convert to the data-driven future of sports (or business).
Two weeks ago, New York Times columnist Steve Kettmann penned a nostalgic op-ed critiquing the modern game for becoming a slave to statistics and data analysis. In many ways, Kettmann mirrored the sentiment we heard in the NBA back in February, when erstwhile TNT commentator “Sir Charles” Barkley called out Houston Rockets GM – and well-known stat head – Daryl Morey for his reliance on analytics on the court. Barkley professed to caring little about “analytical BS” and instead relying on “the style of play, looking at matchups” to make his judgments about how teams will fare in the league. Kettmann, for his part, says largely the same, arguing that today’s “advanced stats” get in the way of appreciating the nuance of the game. Both arguments should be familiar to sports data fans as well as corporate consultants who counsel executives on a daily basis wrestling with the concepts of big data and analytics, and how to use them to inform business and marketing decisions.
What Kettmann and Barkley, as well as many business executives, miss is that this isn’t an either/or situation – you don’t need to dismiss data analysis outright if you rely on instinct or experience in making (game time) decisions, nor is personal judgment a relic of the past in the age of big data.
Take this example: Before I founded HPA, I ran a small, but highly successful, sports analytics business predicated upon a very data intensive model that ran thousands (and thousands) of permutations overnight, every night. But someone with the experience that comes from watching thousands upon thousands of games (me) still manually reviewed all the results the servers generated each morning and made adjustments to, or overrode entirely, several of those outputs.
That’s because the computer doesn’t know that it was just announced that Player X has failed a drug test, or Player Y had a baby the night before, or Player Z is still favoring his bum ankle even though he’s starting, or that the Coach is not there because his father passed away. There’s data about the human element of the games that can’t be quantified but still impacts the outcome.
I love data. I started the analytics business because of my love of sports and data. At HighPoint, I use data to better understand our business and how to grow it. But at the same time, there’s an experience aspect to everything that needs to inform decisions – when I had kids and HPA started to take off I couldn’t devote all my free time to watch every game each night, so I shut down the other business. Why? If you can’t apply human insight to the data, performance (on the field or in the boardroom) suffers and what you end up with is a bad product.
The question we need to tackle, as sports fans or business leaders, is how to meld analytics and experience to get the best of both. How do you leverage the mass of data – collected from more sources, with greater detail than ever before – without becoming a slave to it? Alternatively, how do you see the mountain of data and avoid the temptation to throw up your hands and say ‘fuhgeddaboutit’?