Three years ago, Major League Baseball executive Bob Bowman used the MIT Sloan Sports Analytics Conference as the venue to unveil Statcast, a new system that would track every movement on a baseball field and promised to alter fundamentally how the game was viewed and discussed.
In two seasons since Statcast became a staple of MLB coverage, the change to the game has been considerable. A concept like exit velocity is now a staple of discussion about power-hitting prowess, and other concepts (launch angle, defensive route efficiency, pitcher spin rates, and more) have likewise crept into the lexicon.
“[The dialogue change has happened] in some sense a little more quickly than I would have thought. Two years in baseball time is very little,” said Mike Petriello, an analyst for MLB.com who is in many ways the site’s voice of Statcast. “Exit velocity took off very quickly. It’s very easy to understand. It’s ubiquitous. You see it in a lot of in-stadium scoreboards now. You see guys talking about it constantly now. Kris Bryant talks about exit velocity and launch angle. Mark Trumbo has talked about his launch angle. … It’s informing the way the game is being talked about and also [how it’s being played] on the field.”
Over the weekend at this year’s Sloan Conference, Petriello and Greg Cain (the senior director for sports data at MLB Advanced Media) unveiled Statcast’s latest innovation that could reframe our understanding of what happens on the field – and which outcomes represent luck vs. skill.
The duo introduced two new metrics that Statcast will use for the coming season: Hit probability (the likelihood that a ball in play turns into a hit based on its exit velocity and launch angle) and catch probability (available for outfielders only, it will examine the likelihood that a ball is caught based on the distance an outfielder must cover to reach the ball and its hang time).
The use for such measures is considerable. For instance, when Jackie Bradley Jr.’s 29-game hitting streak ended, he hit one ball with a 100 miles per hour exit velocity and a 30-degree launch angle – which had a hit probability of .593. He hit another with a 101 m.p.h. exit velocity and a 29-degree launch angle – which had a hit probability of nearly .700.
Knowing those pieces of information can open up worlds. They are revealing about the quality of Bradley’s at-bats (he should receive offensive credit for hitting balls that often would have yielded hits). They say something about the pitcher’s performance, as Rockies righthander Jon Gray was in no small measure lucky that a couple of hard-smashed balls ended up in his outfielders’ gloves. And with catch probability, it would also be possible to measure the role that Colorado’s defense played in the end of the streak – and in the role it played on turning a hard-hit ball into an out.
Such information, in turn, can play a considerable role in redefining how we view a player’s overall performance – whether his skills are eroding despite numbers that look unchanged on the surface, or whether he remains as good as ever (and merely the victim of bad luck) despite seeing a decline in, say, average or slugging percentage.
For years, it’s been reasonable to discuss the role played by luck in transforming a hard-hit ball into an out or a meagerly struck six-hopper into a hit. Increasingly, Statcast will permit real-time means of measuring that phenomenon.