About two weeks ago, I wrote about the Humidor and if it was having an effect on the batted ball data.
Today, I am going to briefly present some of the pitching data. It isn’t a lot and this will be a short post. But, essentially, I’m trying to look at the inputs and not necessarily the results (e.g. the amount of home runs). The hope is that looking at the inputs will provide a more accurate representation of the effects rather than the results, which can be quite skewed in small samples.
I will say that when I looked at the batted ball data for hitters, it didn’t really offer a super compelling story. The results seemed mixed and inconclusive to me, though it did seem to imply that there wasn’t really an effect.
Is that different for our pitchers?
Exit Velocity for our Pitchers
The main purpose of the Humidor is to store the baseballs in a climate-controlled environment instead of the hot, dry air of the desert. The lower temperature and higher humidity work together to reduce the bounciness of the ball (and thus, lower the exit velocity) while the extra water weight should, theoretically, reduce batted ball distances flown once hit.
But is that happening?
I have gone over to Statcast and pulled up the average exit velocity for the league as a whole (to adjust to league conditions), then I pulled up the home and away splits for our pitching staff. Behold:
This is really interesting. If we were to look at just at home vs. away splits (red and green), it would appear that the Humidor is having a dramatic effect. But we have that pesky league split (blue) as well: and the home vs league split has barely changed in the 4 years of this data. It’s possible the Humidor is having an effect while there are other factors at play that are affecting the league and away splits.
But if I had to guess, I think our away split is skewed for some reason. This might be worth digging into on its own post (possibly in conjunction with shoewizard’s post about third time through the order).
But what about the second point - are the batted ball distances affected by the increase in water? Let’s look at the flyballs:
Again, not terribly conclusive but it seems to be trending away from an effect. The gap between home and away is practically unchanged but have both increased relative to the league. If the Humidor was having an effect, it would be expected that the gap between home and the league would have shrunk.
The one big variable that’s still out there is the weather. We are just now getting into the summer months in many of the non-Arizona parks. This might bump the league data up relative to the AZ home splits as Chase Field will pretty much always have its roof closed going forward. The fact that league-wide exit velocities are already up almost directly in proportion to what we’ve seen in Arizona does make me a bit skeptical.
Verdict: Inconclusive, needs more data
But What About the Grip?
The first description of the Humidor - less bounciness and more water weight - has actual bearings in science. In practice (e.g. in how it applies to baseball) has some holes and thus why we’ve been looking into this data.
But Arizona didn’t just talk about that. In fact, they spun it with the idea that it would “improve grip” on the baseball for our pitchers. The thing is, how do you quantify something like this? There should be a “measurable” quality or stat that we should be able to relate to pitcher grip.
I’ve come up with two - command/control and pitch RPM. With better grip, you should, in theory, be able to throw it more accurately. You should also be able to grip the laces better and spin the ball more. I also think there might be a negative relationship with velocity - however slight - because improved grip would mean more friction and more friction should lower the speed of the ball (I did not investigate pitch speed this go-round).
So, let’s start with command/control. How do we measure this in a way relating to pitcher grip? The “best” way would probably be to look at where the catcher puts his glove and see the frequency of the pitcher hitting that mark but sadly, we don’t have the tools to do that easily (and no way in hell I have the time to look at 10,000+ pitches). The next best thing would probably be to look at heat maps and compare between years, but again, we don’t have the tools to really do that well.
So, this is quite a challenge. I’ve decided to look at home vs. away walk rates as a rough proxy.
Dbacks Pitching Home Vs Away BB%
This is reasonably interesting. Our home BB% has dropped while our away BB% has stayed nearly the same. The difference between home and away is the second-largest in the 7-year sample that I took. This, of course, does not mean that there is a relationship here; it is merely one piece of the puzzle.
Feel free to suggest other ideas to investigate here, as I think it warrants a deeper look.
Next up is pitch RPM, which is a very understudied aspect of baseball. There are certain teams (cough Houston Astros cough) that have put a very high emphasis on spin rates and they have dominated with their pitching over the last few years. However, this helps us here because it is a metric that Statcast captures on every pitch so we can look at some trends.
I started by looking at the fastballs, curveballs, and sliders through by the five main pitchers we’ve had last year and this year: Zack Greinke, Patrick Corbin, Zack Godley, Archie Bradley, and Robbie Ray. I omitted Taijuan Walker sort of by accident - I didn’t think he pitched enough this year but neither has Robbie Ray. But I don’t think that changes much.
However, comparing them from 2017 to 2018 and by pitcher didn’t really help much. I think I will save that for another article. I did decide to compare the home and away RPMs for that group as a whole on their fastballs, curveballs, and sliders between 2017 and 2018:
I don’t think the pitch RPM theory holds up as all three gaps have shrunk between 2017 and 2018. And then there is that slider gap, which doesn’t make sense at all (and is backwards to what we would expect here).
Verdict: Inconclusive, needs better data quantification
So, as we saw with the hitting data, it’s really hard to draw meaningful conclusions yet about the Humidor. Everything seems to imply that it really isn’t having much effect, but I think we really need more data before we can say that with any confidence. I actually expected a bigger impact that we are seeing, at least thus far.
Please discuss this in the comments! This is not something where I am speaking from a great deal of confidence. We need more and better ways to investigate the relationships here.
I’ll take a look at this again at the end of the season. But I can understand why park factors are so insanely difficult to calculate mid-season. There are just way too many variables.