One Way to Use Statcast as the Future of Evaluating Hitters

Thanks to the wealth of new Statcast data that’s entered the public sphere over the last two seasons — including information on batted-ball exit velocity, direction, and location, as well as this year’s rollout of previously unavailable launch-angle data — the metrics we currently use to evaluate hitters are not the same metrics we’ll be using to evaluate hitters, say, 10 years from now. We’ll need several years of complete Statcast data, at least, before people much smarter than myself can start devising these metrics with any sort of confidence, but that won’t stop us from playing make believe until we have them.

The more I thought about the idea behind this post, which was originally intended to be trivial in nature — just looking at some of the season’s more unique singles — the more I realized there was to it. That’s not to say this is any sort of groundbreaking work — there are no major findings in what’s to follow, and in fact there aren’t any findings at all. But a simple base hit recently reminded me of how close we are to the next leap of better understanding the game. In the time it takes to get through this post, we’re just going to watch some clips, and think about those clips and what they may mean, and dream on the future. Nothing wrong with any of that.

I was watching a baseball game the other night and a player hit a baseball with a bat. That’s gonna happen. The player was Yan Gomes, and he hit the baseball like this:

Now, that type of hit is called a single. It’s been called a single for more than 100 years, and it’s still called a single now. It’s a single according to batting average, it’s a single according to total bases, it’s a single according to OPS, it’s a single according to wOBA and wRC+, it’s a single according to WAR. A single. No more, no less.

Y’know what else is a single? This is a single:

Austin Romine singled! So did Yan Gomes. Two singles. In batting average, in total bases, in OPS, in wOBA, in wRC+ and in WAR, these two hits count for the same thing. Romine even got an RBI! Throughout almost all of baseball history, what Romine did with his baseball bat would be considered better than what Gomes did with his baseball bat. In a newspaper box score, Romine drove in the runner from third with a single. Gomes just singled.

I don’t write for a newspaper anymore. I write for the Internet, and the Internet is where Statcast lives. And so, thanks to the Internet and thanks to Statcast, I can learn more about these hits. I don’t even have to see them! The single with the RBI came off the bat at 48 mph and traveled 3.4 feet before its flight was interrupted due to a launch angle that had a negative sign in front of it. The single without the RBI came off the bat at 99 mph, at a near-ideal launch angle of 25 degrees, and it went more than 350 feet. That single was the kind of single that makes you sympathize with the batter over his lack of extra bases.

It seems weird to say, and of course this is an abstract thought, but the point of hitting isn’t actually about being credited with a single or a double or a triple. Those are all just made-up things. If you’re a poker player, this concept will immediately make sense — poker isn’t about making hands, it’s about always doing what yields the highest expected value. The money is made through the process, not the results. At the plate, it can be boiled down to trying never to do what Austin Romine did, and about making the ball come the bat more like the way it came off Yan Gomes’ bat as often as possible.

Or Kris Bryant’s when he hit this single:

Or when Dioner Navarro got this one-base hit:

Can’t forget about Chris Davis hitting half of one double:

As far as I’m concerned, those guys all hit home runs. As far as Statcast is concerned, they pretty much did, too.

Take Navarro’s, for example. Off the bat at 104 mph, launch angle of 21 degrees, went 378. Ideal hit. Of the 100 most similar batted balls this season, 50 of them went for doubles, and another 23 went for homers. There was a triple, and this was one of just four singles. In other words, batters should expect to slug something like 1.970 on balls hit like this. Navarro slugged 1.000. Maybe that’s too complicated. The point is this: Navarro’s batted ball was worth twice the result he received, it’s always been worth twice that result, and before long, enough information will exist for evaluations, public and private, to properly reflect that.

This isn’t a revolutionary concept, and I’m certainly not the first one to put thoughts to words on the subject. Jared Cross, proprietor of the Steamer projection system, has already gotten to work on using exit velocity to improve his projections. Others in the industry are actively working on ways to turn raw batted ball data into expected run values, separate from the actual results of those balls in play.

It took some time for the public to come around to Fielding Independent Pitching, another process-oriented way of thinking, rather than one that’s results-based, and in fact the public is still coming around to it. If we’ve agreed that the better way to evaluate pitchers is through process rather than results, shouldn’t we do the same for hitters? We’ve just never had the means to do so. Soon enough, though, Yan Gomes will receive extra credit beyond a single for his batted ball, like he should. For more than a century, we’ve dealt in inexact absolutes. Change is on the horizon.

We hoped you liked reading One Way to Use Statcast as the Future of Evaluating Hitters by August Fagerstrom!

Please support FanGraphs by becoming a member. We publish thousands of articles a year, host multiple podcasts, and have an ever growing database of baseball stats.

FanGraphs does not have a paywall. With your membership, we can continue to offer the content you've come to rely on and add to our unique baseball coverage.

Support FanGraphs




August used to cover the Indians for MLB and ohio.com, but now he's here and thinks writing these in the third person is weird. So you can reach me on Twitter @AugustFG_ or e-mail at august.fagerstrom@fangraphs.com.

newest oldest most voted
scotman144
Member
scotman144

What in the world is up with MLBAM telling BP to take down their run expectancy based on exit velocity/launch angle tool? Why would they make the data readily available and mine-able via baseball savant but then suppress a tool putting all the data together?