Editors and reporters aren’t called to a writer’s life because they love data. So it is not a surprise that more often than not, getting journalists to think about analytics is a hard sell. But there is one corner of the newsroom where data is a mainstay: the sports desk.
At ESPN, that “corner” is the cornerstone of its publishing model, and Data Expert Ben Alamar works to support the network’s editorial output. As a writer, researcher, and academic who has consulted on analytics for a number of NBA and NFL teams, he’s worked with coaches (and editors) who embrace hard figures–and those that don’t.
Learn more about Benjamin Alamar here.
We sat down to talk with Alamar about the ins and outs of data analytics in the sporting world as well as what the data-crunching era means — and doesn’t mean — for media. The short version: data analytics are just one tool among the many that a pro sports team — or a publisher — should have at its disposal. But data isn’t fact. They require interpretation, intuition, and what we in the newsroom call instinct. And that’s where the NBA draft room and the newsroom intersect.
Read on to hear why.
Describe one of your big analytics-based successes in the NBA.
At one team we had a very old-school scout. He was a very nice guy, but the first year I was in the draft room he clearly had no interest in what I was saying. But after a year he ranked a set of players in the draft room and then I gave my ranking of that same set. And he said, “Wait, I want to change my ranking. What Ben said made a lot of sense.” That was a huge win. To know that I’d gotten to the point where he could understand, process, and then act on the information that I’d given him was a great moment for me.
Have you run into a lot of that sort of resistance?
Resistance sort of has a negative connotation to it. I don’t really think it’s that. I think it’s human nature. The people I was working with were extraordinarily successful people. There are only 30 NBA general managers. To get there you’ve been very, very successful. If you’re at that level you’ve gotten there doing things a certain way. And here I come asking you to do it differently. Well, of course you’re going to be skeptical.
So NBA coaches aren’t all data heads by now?
No, not at all. Some coaches are more interested in data than others. Often a staff will have one person who’s really into analytics and will push that piece of the conversation during preparations for games. Sometimes the head coach just won’t really like that kind of information. Sometimes they’ll see it as just one piece of the puzzle and some coaches don’t look at it at all.
Does use of analytics correlate with success in the NBA?
[Legendary NBA Coach] Phil Jackson’s been extraordinarily successful and really does not like analytics at all.
Does analytics use break down generationally?
I don’t know that it does. [Sixty-four-year-old Sacramento Kings Coach] George Karl uses analytics: The [San Antonio] Spurs [coached by 66-year-old Gregg Popovich] are one of the most analytics-based organizations in the league. I don’t necessarily think it’s generational. I think it’s just a question of which coaches are getting the best sort of look into analytics and are able to trust them at first. What I’ve heard from players is that they understand that going forward, if they want to work on the coaching staff or in the front office, then they’re going to need to understand this stuff, that it’s part of the game now.
In what sports are analytics the most useful?
They’ve had the most impact in baseball, for sure. That’s because baseball’s the easiest game to do analytics for. It’s a static game. It’s a game defined by very discrete moments. Most of what goes on in a baseball game happens between two people. In basketball there’s all this interdependency. And football – we don’t even have very good data on football. You know what maybe four or five people at best have done on given play, when there are 22 players on that field.
What’s the biggest misconception that people have about data?
When people see numbers, they want them to be facts. The variability in numbers is hard for people to grasp if they haven’t spent a lot of time thinking about it. That there are error rates for estimates is not something that people are comfortable with, and that some statistics have much bigger error rates than others is hard for people to really grasp. They just want a ranking.
So there’s room to make different narratives out of the same numbers?
Really it comes down to this: What kind of question are you answering? If everybody’s trying to answer the same question using the same set of data – the same analysis of data, I should say – then they should come to the same kinds of narratives. But if you’re using the same analysis but trying to answer different questions, then there could be very different stories that you can tell out of that.
What’s the worst thing a person can do with data?
Use it selectively to support what they already believe. It’s a very, very common thing. It’s a natural thing to do. But it’s a misuse of the data.
What would you like to see analytics platforms do in the future that they can’t do now?
Better support people in making decisions. We’re caught in this dashboard-based situation where if you’re not careful it’s really easy to miss the whole picture. We need to continue to translate analytics into natural language so that you get an executive summary of what the data says in a clear way. People would process that a lot easier than they process real-time analytics, which for a lot of people is overload.
Lots of traditional journalists distrust analytics in their field. They fear it inherently leads to posts about the Kardashians. True?
Analytics can basically support whatever strategy an organization wants to use. If an organization wants to build itself around generating the most clicks and if traffic is all it cares about, we have analytics for that. Analytics are just a set of tools. That’s all they are.
This post originally appeared on Parse.ly’s blog.