Spring training can be a tough time for baseball fans. The games don’t count, injuries change the course of the season before it even really starts, and it’s hard to pull any meaningful insights out of the performances on the field. Sample size in spring training is just too small to really draw any conclusions.
Or is it?
That’s the beautiful thing about Statcast; it measures the very granular parts of the game that actually do become reliable fairly quickly. Batted ball data - exit velocity and launch angle - for a given hitter or pitcher tends to become reliable right around 40-50 batted balls, and it’s the pitching side that I really want to hone in on for this post. Consider that the big starters in baseball tend to get around 120 or so batters faced in spring, and even the relief pieces get to see 50 or so batters.
Pitching is often where the big focus of spring is - which relievers get the last couple slots on a roster, whether the veteran in his last season looks like he’ll go out on a high note, etc. The Yankees have a couple of those cases in Ben Heller, J.A. Happ and others.
This is where Statcast can really help analysis. Building block data, like swing and chase rates or spin rate, is reliable in even shorter samples than batted ball data, and can be used to better evaluate a pitcher than ERA over four spring training starts.
Take Happ, for example. One of the reasons why I’m so skeptical of him being much use in 2020 is because his fastball is getting demonstrably less effective as time goes on, and it’s a pitch he leans on a lot:
The spin rate on Happ’s fastball, which he throws about 50% of the time, has steadily declined, and there has been a proportional change in his exit velo allowed. Spin rate is a great gauge for deception - Happ’s primary pitch is fooling fewer hitters, and they’re able to make better contact with it as a result.
Unlike the early days of Statcast, we now know that you can actually increase spin rate, or at least, some pitchers can. For a guy like Happ, one of the worst fulltime starters in baseball last year, being able to bring deception back to his pitches would go a long way in making him a reliable starter again. Providing Statcast data to the public would be the best way for us as fans, and for the Travis Sawchiks of the world, to better evaluate how the players on our teams are doing.
There’s also an entertainment aspect to Statcast that the Yankees and MLB can better leverage. One of the reasons why Statcast has proven so popular among all fans is that it’s pretty simple stuff - how fast does a guy run, how hard does he hit the ball, etc. The criticism of WAR as “fake winz!” is a little silly, but it does highlight an issue with statistics that can only really be told with a spreadsheet.
Statcast, meanwhile, is just so much more narratively pleasing. In the regular season, YES incorporates the AWS breakdowns of Statcast metrics into replays and commentary, in ways that help to further tell the story of the game. As fans, we’re going to struggle to stay invested in what may arguably be an overly long spring training, so you might as well make the broadcasts as interesting as possible.
All of this is feasible because MLB clubs already have full Statcast setups at their main spring training ballparks. The tech to help fans get more out of spring training is already there, just let us use it.