TextQ&A: DePodesta talks stats, research, and predicting the future of human beings

DePodestaMets VP of Player Development and Scouting Paul DePodesta had the following on-air exchange with CNBC earlier today:

CNBC: Is data still influencing the way you pick players?

DePodesta: Oh, absolutely. I think it’s influencing the entire industry, really. Baseball, like a lot of other industries, has a fundamental characteristic of uncertainty. We want to know the future and we don’t know it. And I think about, you know, 15, 20 years ago we started admitting to ourselves that we’re not very good at this prediction business, how can we get better at it? We found data could really help us, at least providing a framework around this uncertainty.

CNBC: Break it down for the average person, how do you use data to better scout and better staff your lineups and pick the best players and make the best teams possible, especially for those of us that don’t remember exactly how it worked in the movie (Moneyball)?

DePodesta: Sure. Again, I think we’re trying to predict the future performance of human beings – often times in situations that those people themselves haven’t even encountered. So, one of the things we really need to do is separate the skill from the luck. … Skill is much more repeatable than luck. Going through the data we try to find out which skills are repeatable and which of the outcomes out there are just luck. And you know, if there’s a guy who hit .300 out there, we need to determine whether or not he’s truly a .300 hitter or whether or not it was a one-time occurrence. The data is not always going to be an exact science obviously in picking those players.

CNBC: Can this be as effective as it once was, just given the fact that so many teams are deploying this whole data breakdown on how they choose their talent, for example, I would assume you’re going to try to bring the same qualities to the Mets?

DePodesta: Sure. Going back 15 years or so ago, one of our biggest challenges was wading through all of the data available to us. Baseball statistics had been around for 100 years. To try to figure out what truly mattered was a real challenge. If we thought that was a challenge then, we had no idea what a huge challenge it would be today. The amount of available information has exploded in the last 10 or 15 years. So now the ability to sift through all that, try to find out the real causal relationships, the truly relevant ones, is a huge challenge. I think it will continue to be a big challenge.  So while all this data presents real opportunities, it also presents, I think, some significant challenges that need to be worked through, both very seriously and very diligently. 

CNBC: You feel like you have great challenges ahead of you with the Mets and do you wish – looking at not only what the Dodgers sold for but the kind of money they have to put to work – that you still had a payroll like that at your disposal?   

DePodesta: I’ve been on all ends of the spectrum, in Oakland, in San Diego, I’ve been in places like Los Angeles and New York. So I’m comfortable really dealing with whatever framework we have and then trying to be as creative as possible with what we have. I think we have great resources in New York. We have tremendous support from ownership. I’m excited about what we’re going to be able to do both on the field and off the field. The resources aren’t just about paying for talent, but it’s also about paying for resources in the front office, for processes for people that will help us make better decisions.