The Interplay of Velocity and Effective Velocity

I went to the SABR Analytics Conference in Phoenix about a month ago, and one of my favorite presentations was by Perry Husband, creator and champion of the concept of Effective Velocity.  The presentation actually jogged my memory for an article I read on SB Nation by Jason Turbow back in 2014 that introduced me to the concept and chronicled some of Husband’s efforts to get MLB teams to buy into his idea.

I recommend that you read both of those links for a detailed (and highly enjoyable) explanation, but I’ll also provide a simple summary here.  The idea of Effective Velocity is that the location of a pitch affects the velocity a batter perceives that pitch to be.  That is because, when a ball is thrown up and in to a batter, he cannot make solid contact with the ball if his bat is square to the front of the plate.  Instead, he has to turn on that pitch and make contact with the ball with his hands extended in front of the plate with his bat square toward his pull side of the field.  And because the batter needs to make contact with up-and-in pitches in front of the plate, he has less time to work with, and those pitches play faster than their actual velocity, at a rate of about 2.75 mph faster per six inches according to Husband.  Meanwhile, the opposite is true on pitches low and away.  Batters make solid contact on those pitches with swings square toward their opposite field, which means they have more time to make those swings and those pitches appear slower.

When I first read about Effective Velocity in 2014, I found the idea really captivating, but Husband’s presentation last month helped me realize that I now had the means to actually test it.  Before I detail the approach I took, let me first point out that Husband’s work on this topic is much more sophisticated than what I’ve come up with, and he has various work and services available on his site for anyone interested.

To my mind, the biggest potential benefit to the existence of Effective Velocity would be its importance to pitch sequencing.  Pitchers presumably select their pitches in part to vary their velocity and location and keep hitters off balance during their plate appearances.  Effective Velocity suggests that hitters may perceive two pitches with different velocities in different locations to be the same velocity and may perceive two pitches with the same velocities in different locations to be radically different velocities.  To explore that idea, I’ve calculated a basic Effective Velocity of every pitch from 2013 to 2016 by first drawing a line that connects the low-and-in corner of the plate to the high-and-outside corner.  Then, I applied Husband’s rule of 2.5 mph per six inches to pitches with locations off of that diagonal to get an estimated Effective Velocity.

With those estimates of the Effective Velocity of pitches calculated, I then looked at sequences of consecutive pitches to hitters and tested the swinging strike rate (swings and misses per pitch), swing and miss rate (swings and misses per swing), and hard contact rate of those second pitches.  For each stat, I created a heat map with the Effective Velocity differential on the x-axis and the actual velocity differential on the y-axis.

First, swinging strike rate:

 

When I look at this first heat map, I see an arrow that points from the upper left-hand corner down to the bottom right-hand corner.  The diagonal shape of that arrow suggests that velocity and Effective Velocity are having mirrored impacts on hitters.  When the difference in either velocity or Effective Velocity from one pitch to the next increases but the other does not, hitters become less likely to swing and miss.  However, when the difference in both velocity and Effective Velocity from one pitch to the next increases, hitters become more likely to swing and miss.  I’m not sure what to make of that trend, so I’ll put a pin in that for now and move on.

Next up is swing and miss rate:

 

This heat map shows a pretty clear increase in swing and miss rate as the velocity differential increases but not as the Effective Velocity differential increases.  The former makes sense related to the thought that when pitchers vary their pitch velocities on consecutive pitches, it keeps hitters off balance.  However, this is not evidence that Effective Velocity is a relevant concern, so let’s move on.

Finally, hard contact rate:

 

I had to stare at this one for a little while before I saw it, but notice how the faintest region of the heat map is toward the right side in the vertical middle.  In that region of the map, the Effective Velocity differential between pitches is higher than the actual velocity differential, and that is leading to weaker contact on average.  Meanwhile, as you move left across the heat map, the hard contact rate is increasing.  In particular, the three squares with velocity differentials between four and six mph but with no Effective Velocity differential show pretty much the highest hard contact rate on the map.  That’s exactly what we were looking for.  Effective Velocity suggests that even though those pitches have noticeably different real velocities, hitters perceive them to be the same velocity, and they are squaring those pitches up.

I think that discovery is real evidence that Effective Velocity has merit, which makes it an exciting avenue for potential research on whether pitchers’ adherence to the concept in their pitch sequencing might have a tangible impact on the performance of their batted balls allowed.  Hopefully, that is something I can dig into over the course of this season.