Batted Ball Velocity, Mark Trumbo, and the Safeco Field Effect

With the introduction of Statcast, it seems like batted ball data is the next frontier in baseball analytics. More granular data is becoming available to the public but no one really has any idea what to do with it yet. A couple of weeks ago, I took a look at how Safeco Field affected batted ball velocity and whether or not it had an effect on the Mariners. In the end, there are very few definitive conclusions we can make using this data. It’s incomplete at best and erroneous at worst. But it’s there, it’s available, and we can still attempt to see what we can glean from it.

After looking at team performance in this area, I wanted to see what a deep dive into a single player’s performance could look like. What better batter to examine than Mark Trumbo, the pinnacle of Jack Zduriencik’s tenure with the Mariners. We all know the story; Trumbo is acquired by the Mariners at the beginning of June and proceeds to post the worst offensive line in all of baseball during that month. The rest of his season was much better (there was nowhere to go but up), as he posted a slash line of .295/.351/.479 from July 1 onwards.

In 2015, Trumbo ranked 16^th in average batted ball velocity, just ahead of his teammate, Nelson Cruz. After pulling his batted ball data from Baseball Savant, we find that 72% of his batted balls have velocity readings, just below the league average. Here’s his average batted ball velocity with the league average and the average posted at Safeco Field:

So far, so good. Trumbo was hitting the ball harder much harder than league average this year. As I hypothesized in my previous article, Zduriencik was focused on building an offense that could thrive in Safeco Field and used batted ball velocity as his bellwether stat. Trumbo certainly fits the bill.

When examining batted ball velocity, the oft missing piece is the angle off the bat. The outcome of a batted ball greatly varies depending on the angle at which it is hit. Unfortunately, the data provided by Statcast, and curated by Baseball Savant, often doesn’t include the launch angle off the bat. In fact, Mark Trumbo’s dataset includes a launch angle reading for just 5% of the data. Angle is important for our analysis because of the donut hole.

Alan Nathan, a baseball physicist, built this graph to show how batted ball velocity and launch angle are connected. That deep blue area between 70 mph and 90 mph, those are fly balls that die in the outfield. At a certain point, around 95 mph, launch angle begins to have less of an effect on the outcome of a batted ball. Since we don’t have launch angle data, we’ll have to make do with what we’ve got. Batted ball classifications seem to be the closest we can get to estimating launch angle.

First, let’s take a look at Trumbo’s batting average on balls in play across a range of velocity bands:

The sample size is pretty small across these velocity bands but a cursory glance reveals something very interesting. That very distinct "W" shape to Trumbo’s home BABIP is very abnormal. As a team, the Mariners performed better than average when hitting the ball between 81–90 mph, but Mark Trumbo takes this performance to the extreme. What’s even stranger is his performance away from Safeco; on anything he hit below 90 mph, it was almost guaranteed to be an out.

What the heck was going on at that velocity range that’s right in the middle of the donut hole? Let’s take a look at Trumbo’s batted ball mix for that velocity band:

As we might have expected, Trumbo hit a ton of line drives with a velocity between 81–90 mph at home. These are soft line drives hit over the head of an infielder which fall in in front of outfielders. Take a look at his spray chart (both home and away) for this range band:

A bunch of grounders, a few pop ups, and line drives sprayed across the field. It’s interesting to note that all of the fly balls hit in this velocity band were hit to the opposite field. We don’t have any historical data to tell us whether or not this was Trumbo getting lucky on a bunch of liners but because of the small sample size, I’m willing to bet we won’t see this kind of shape to his BABIP graph next year.

Because Mark Trumbo usually hits the ball very hard, his batted ball outcomes will be greatly dependent on his batted ball mix. His line drive rate and his fly ball rate were the highest of his career this year and that certainly helped him post his highest wOBA since 2012. Whether or not these batted ball trends are an indication of a new approach or just fluctuations within his established norms is hard to say. With his elite batted ball velocity, more line drives and fly balls certainly can’t hurt. Still, even in the upper reaches of his velocity bands, Trumbo was underperforming against the league average while batting at home. That's almost certainly an effect of the oppressive environment in Seattle and not exactly a good sign for Trumbo's continued success.