Keith Law And The Business Of ‘Smart Baseball’

When Keith Law began his career in Major League Baseball in 2002 as a grunt in the Toronto Blue Jays analytics department. A large chunk of his job was spent gathering data on college players to present to then-GM J.P. Riccardi. This analytical information would be used when deciding which players to scout, which to draft, and which to pass on completely.

In the early-2000s, only a handful of teams were mining data from college team websites. Few teams were even really paying attention to the stats of their own players besides the old tried and true measuring sticks like batting average, long balls and runs batted in. Then Michael Lewis published Moneyball and the “data boom” in MLB began.

Then in 2004, Michael Lewis published Moneyball and the Sabermetrics “data boom” became just as important to teams as the scouts out in the trenches.

Law moved on to cover baseball for ESPN Insider after creating the insanely popular Baseball Prospectus. He now works in front of the camera on ESPN’s Baseball Tonight.  Law’s new book, Smart Baseball: The Story Behind the Old Stats That Are Ruining the Game, the New Ones That Are Running It, and the Right Way to Think About Baseball, is out this week. In this book, Law attempts to take a baseball bat to a century’s worth of accepted wisdom in calculating the true value of player. Armed with “concrete examples from different eras of baseball history, logic, a little math, and lively commentary,” Law proves how the allegiance to certain stats is firmly rooted not in a ridiculous adherence to the old ways of thinking about the sport.

Law answered some questions for us about the book, his love of stats and how sabermetrics has changed the way MLB teams operate.

Which team took the longest to come around to this way of seeing the sport?

Arizona was the last holdout, but when they fired Dave Stewart in September and hired Mike Hazen, setting up a proper analytics department was one of his first priorities.

Do you think this approach to data could be used in other businesses? 

It’s being used all over. Big Data’s been around for at least a decade, and it’s changing many industries. Becoming comfortable with this kind of data-driven thinking is critical for the next generation of white-collar workers; even if you’re not a coder or an analyst, you have to be able to think in these rational terms.

You mentioned that this new approach could help with preventing injury. How do teams handle the issue of “if this player continues like this he’ll get hurt” vs. “we need this player for a playoff push.”

I can’t speak to any specific teams facing that question, but it will be something they have to deal with going forward. If I were the GM, at least, I’d want to have the player involved in that conversation, and ensure that he’s comfortable with any risks we might be taking – and if there’s a question of compensating him further (hey, we’ll pick up your club option, or add a year to your deal), do that in a way that all parties walk away feeling like the risk and reward are in balance.

Are there any drawbacks to this push to “big data”? 

Job losses. I worry about the state of scouting long term, as data streams like Statcast provide objective information on things scouts have long been asked to evaluate in subjective terms. You can tell me that a certain pitcher can really spin the breaking ball, but now we can put a number on that in revolutions per minute. The nature of scouts’ jobs is changing, at the very least, but I hope teams still see the value in first-person observation and evaluation.

Can you think of any players from any generation who would have benefitted from this data? Maybe someone who doesn’t have the gaudy numbers but really was an exceptional player.

I have a chapter in the book where I highlight some players who were better than contemporary or historical accounts would say, and others who were overrated for similar reasons. But I also think going forward we will see different players get opportunities because, say, Statcast tells us that they have great exit velocity (speed of a batted ball) or a high spin rate on a curveball, so teams will look at them and say “let’s play this guy more, or use him differently” whereas in a previous generation they might not have gotten the chance. There was probably some pitcher in the 1980s who had a high spin rate on an average fastball and didn’t get enough of a chance because he didn’t throw that hard. That same pitcher today will get his opportunity, and now teams are actively scouring Statcast data for players with standout skills like that.

In the final section of the book, you mentioned the amount of money and manpower MLB teams now dedicated to analyzing data. Do you see that becoming the most important area front office area in the future? 

It’s certainly the newest area in which teams are looking for competitive advantages, but I would stop before saying “most important.” It is now as critical as having a good scouting department, but right now, I think the best model incorporates analytics, scouting, and people management so that the right information is getting to players in a way that they can use it and buy into using it.

How much money do MLB teams allocate to these jobs? 

I know entry-level analysts are paid $50K-$100K, roughly, although the teams offering salaries at the bottom of that range have been turned down by folks who can make 2-3 times as much outside of sports. If you’ve got a PhD, you shouldn’t take a pay cut of 50% or more just to work in baseball. That’s the old mentality that says it’s a privilege for you to work in the sport so you should make less money.

Smart Baseball: The Story Behind the Old Stats That Are Ruining the Game, the New Ones That Are Running It, and the Right Way to Think About Baseball is available now.