I’m absolutely fascinated by the physics of baseball. Talk about spin rate, multi-axis movement, and moment-of-inertia, and I’m here for it. I’m equally fascinated by how easily the physics of baseball can become distorted or misrepresented by people who don’t actually understand much about them. The Statcast era provides ample data for both intelligent discourse and not-so-intelligent discussion. Too bad most broadcasters excel only at the latter.
Given the amount and variety of raw data involved, it’s a shame to limit Statcast discussions to oversimplified soundbites and clips about exit velocity and launch angle. There will always be a certain fascination with long home runs, but that can only drive interest to a point. Going beyond that point requires a step or two outside the bandbox and into some underappreciated areas of interest.
The new Sprint Speed metric represents one of those areas for me. Mike Petriello put together a great write-up about this metric, and I encourage you to read it for the introduction to the topic and then stay for the Magneuris Sierra video clip. Note that Sprint Speed is measured using “feet per second in a player’s fastest one-second window”. That’s kind of a another way of describing the greatest distance traveled in one second.
In racing terms, we’re talking about a runner’s top end speed in a window of specific length. The distinction to be made here is that the specific window doesn’t have a location in space but rather in time. Therefore, Sprint Speed leaderboards necessarily favor players who have a higher top end speed which is great for the brochures Scott Boras pumps out for his clients. It also looks good on an infographic. Too bad it’s just the tip of a potentially interesting iceberg.
Just imagine where Sprint Speed and the associated technology could lead. Dare to dream of a world where the decision on whether or not to send a runner home requires both quantifiable knowledge and that intangible experience or whatever old people who coach 3rd base like to call it.
Granted, such a decision must be made in real time, under pressure, and with a host of considerations in mind. Sprint Speed could be part of an approach to make the decision easier and more quickly.
For years we’ve been taught that game situation, score, number of outs, arm strength, hit location, fielder positioning, and the runner’s speed are all variables in this incredibly complex computation that only an experienced 3rd base coach can possibly make. To a certain degree, much of that is true, but that doesn’t mean we can’t have a play with the risk/reward factors to make the decision a bit easier. That’s where Statcast can play a role as long as reasonable assumptions are made, and this makes contextual use of Sprint Speed key.
It’s important here to note what Sprint Speed does NOT do. It does not account for baserunning path efficiency, acceleration, time or distance top speed is maintained, average speed, jump, primary or secondary lead, and initial momentum (or lean). Path efficiency determines total distance traveled and greatly impacts a runner’s ability to maintain speed. Think of a runner attempting an inside-the-park home run. His ideal path around the bases should look something like an inverted raindrop. The actual shape of that raindrop can vary slightly, because there is no single perfect travel path. Some slightly longer paths allow the runner to maintain a higher average speed, and some shorter paths allow the runner to lower average speed since he’s traversing a shorter overall distance. Sprint Speed has little to do with path efficiency, but it does relate to average speed.
For the purposes of this discussion, we can actually simplify the equation by comparing two runners hypothetically participating in the same play. Start by setting most of the variables equal with the notable exception of “cruising speed” which I crudely define as “close to and including top speed”.
The first runner is Matt Carpenter, and the play took place last night when Carpenter was thrown out by 175′ on a play at the plate. Actual distance was more like 15′ to 18′, but it was still a terrible send. According to Statcast, Carpenter’s Sprint Speed is 26.0 ft/s. According to my crude stopwatch approach, it took him over 12 seconds to reach home plate starting from the time he took off.
Let’s say that the second runner is someone like Dexter Fowler or Tommy Pham. Both of those players have Sprint Speeds above 28.0 ft/s (28.5 for Pham and 28.4 for Fowler). Now assume that the second runner takes off at the same time Carpenter did and from the same starting point. If you ignore acceleration differences (which decidedly favor the second runner), you end up with at least 9 seconds useful for comparison. For every 1 second of running at or near top speed, Carpenter would lose 2 feet to the second runner. After 9 seconds, he’s at least 18′ behind that second runner.
Replace Carpenter with Pham or Fowler last night, and you’ve got a much, much closer play at the plate. Would either have been safe? We’ll obviously never know, but based on where Contreras fielded the ball, I’d say there is a really good chance of it. It’s not an easy call to make, but Contreras had the luxury of fielding a ball that was slightly up the line while blocking the plate. A runner almost 18′ in front of Carpenter at that time would force Contreras to position himself differently and in a different location. Instead of shifting back to tag Carpenter, Contreras likely has to make a long swipe tag that costs him a half second or more.
If we’re debating whether or not Fowler or Pham makes it, then we shouldn’t even be in a position to have a discussion about Carpenter in the same situation. The former is a tossup, and the latter is a terrible idea.
DISCLAIMER: I took a few liberties in terms of correlating top speed to average speed and “cruising speed”, but I have a high degree of confidence in saying that the gap in top end speed is roughly the same as the gap in cruising speed. Additionally, the acceleration factor creates a bit of a cushion that I left untouched.