Tag Archives: StatCast

Statcast Debuted Last Night, but It’ll Be a While Before It Really Matters.

Yes, very essential, but what does this all mean? (h/t Fangraphs)

The actual Statcast tool has been around for a little over a year now, and we’d seen snippets of it here and there since then, particularly during last year’s World Series.  However, yesterday’s Cardinals/Nationals game on MLB Network was the first in which Statcast was used throughout a game as a device fully integrated into a game broadcast.

The tool, which is being audaciously touted as “one of the largest advances in instant replay that we’ve seen in the last 50 years“, is intended as a way for us to finally quantify the game in a way which is impossible to do with the naked eye: completely, even exhaustively. With multiple HD cameras set up in each of the 30 major league parks, providing a complete view of the field and its adjacent environs, and connected to a system that will immediately spit out all the related data accurate to a fraction of an inch, absolutely action, no matter how small, will be left unseen and free from analysis from this point forward.

And that’s a good thing, right? Well, that depends on how you view the game of baseball.

The real constituents of the system, the major league teams themselves, will absolutely love this tool.  By being able to analyze and quantify every movement made by every player on the field, they will be able to evaluate their own players in context against all other players in the majors at that time, and should be able to properly value their performance as it occurs during live game action.  This should in theory help them more accurately determine what they should be paying their talent.  I’m pretty sure that as the cost of this system drops over time, organizations will be arranging Statcast systems to be set up at their minor league affiliates’ facilities as well.

The players themselves will benefit or be penalized versus the prior expectations of them that had been set for them under less perfect observational or statistical methods.

As for the fans, StatCast is sure to drive a yet deeper wedge between the “sabers” and the traditionalists.  The former group will welcome this development with open arms and spreadsheets; the latter will likely moan that one more nail has been driven into the coffin of baseball as the game of romanticism, poetry, and whatever other ethereal qualities define for them the pastime they love.

As for me, I like that the data are now being made available during broadcasts, but I am more looking forward to the day that we can review the results within the proper context of league norms.  Since we are in the infant stage of the technology, analysis of the data is very immature.  All we have at the moment are raw numbers and the inevitable SWAGs that will follow in their wake, at least in the short term.

Here’s what I mean: in the middle GIF above, Jon Jay’s first step to get to that batted ball was timed at 0.3 seconds, his top speed was clocked at 14.8 MPH, and the distance he covered was measured at 31.8 feet.  To which I ask: are these good numbers?  Is his first step better than average?  Average?  Worse than average?  Same with his speed and distance: are these results good, bad, or so-so?  I have no way of knowing this yet.  You probably don’t, either. These are just numbers so far.  I mean, sure, Jay caught the ball, but does that mean that he necessarily has an earlier first step and/or faster top speed and/or greater distance-covering ability than average? Of course it doesn’t, because we don’t know what those averages are yet. Maybe Jay was just lucky this time that the ball was hit near enough to him for him to reach it.

You can see that I have acknowledged that because we are in Statcast’s infancy, we don’t have this context available to us yet, so I am in no way making an unrealistic demand that I be provided this context, like, yesterday. But I think I can safely say that until enough data to provide this context are gathered, the information will be more like “somewhat interesting” than it will be “really useful”.  It will probably be a while before we get to that point, likely measured in years, but until we do, I have limited interest in the Statcast data, at least as of April 22, 2015.

I do hope, though, that as with HITf/x and PITCHf/x data, StatCast data are made available outside of MLB organizations, if not freely available to the public, then at least on a licenseable basis in which websites such BrooksBaseball.net can obtain it, dissect it and disseminate at least filtered versions of the it to the public.  That’s the really the only way the data will be meaningful to fans who are interested in it.  Otherwise, it’s just a bunch of numbers.

In the end, MLB may well be right that this will be “one of the largest advances in instant replay that we’ve seen in the last 50 years.”  If that turns out to be true, I think it will take a few years before any of us see that to be the case.