Analysis of a Professional Hawt Take

I love baseball. I really do, but when asked why I love baseball, I find it difficult to answer that question. But I get asked this question a lot so I usually respond with two reasons.

The first reason is the reason I loved playing it so much growing up – the drama. There is suspense, heartbreak, triumph and a plethora of other emotions that you can feel on a nightly basis and the real beauty is that you can feel all that for something that has the lowest stakes imaginable – it’s 25 grown men playing a kids’ game.

The second reason is math. This isn’t to say I love math, I’m not that weird. But I’m decent at math and baseball more than any other sport can be understood and/or explained with math. Front offices have always realized this; it’s why there has always been so much data collected in baseball. But what hasn’t always been done is analysis of the data with a systematic approach to really understand what the data says. I love that there is still new analysis being done decades old data that lets us understand the game a little more. Now FOs have invested in technology to collect even more data, which will open up entirely new insights and I find that really exciting. I guess I am a little weird.

I love baseball, I don’t love writing. Between time constraints, crippling dyslexia, and my general love of doing nothing, there are a lot of head winds I have to overcome before I write about anything. 99% of what I write about work related (financial incentives are the best incentives). But 1% of the time I’ll write about baseball. The motivation here is provided that usually I’m writing about something that falls into one of the two categories of things I love about baseball. My two favorite things I wrote about was the Rationalizing Heyman’s HoF ballot Link and quantifying the effect I’d have on the WS odds if I joined the Astros as a pitcher last all-star break Link. The former, was some juicy drama(with some dumb math) the latter was heavy math(with some dumb drama). And since you are reading this, you could probably guess I’ve found something that overlapping section in the math/data and drama baseball Venn diagram.

The Drama:

The story starts with a fairly common occurrence, random person on twitter atting people with hot takes. @Blaze4551 tweeted at Kyle Boddy of driveline baseball asking if the Astros are doctoring balls, which would be the cause of the ‘dramatic’ increase in spin rate’s for Verlander, Cole, and Morton after coming to the Astros. What wasn’t common was the Twitter Verified Boddy replied to the cool name handle with 20 followers; agree that it’s suspicious as personified by the thinking man emoji. But really sparked baseball twitter to erupt was Bauer responding to Boddy’s tweet. Bauer implied in his tweet that the Astros are systematically making pitchers’ better by increasing their spin rate with the aid of a foreign substance.

Before I go any further I should stop and say that I really do like Tyler. He’s definitely a guy that thinks a lot and as a guy that thinks a lot, I can relate. He also has some great insights into pitching, he’ll do and say things that are very different than what you’d here from your typical pitcher, which can be a nice relief from the typical Mike Stanton cliché ridden ‘analysis’ that I’m accustom to hearing on a daily basis.

Anyway, Bauer’s take sparked responses from people Like McCullers, McHugh and Bregman and literally hundreds of others. My opinion on his take is simple. ‘WTF are you talking about?’. To imply the Astros are systematically cheating because a random guy tweeted about increase in spin rates insane for so many reasons. Primarily, it seems like his misapplying some fundamental logic principle. I think in geometry class in high school I learned if P then Q so if P then Q. This is not the same as if P then Q so if Q then P. I remember thinking at the time that I couldn’t believe they gave crazy names to what is really just common sense.

The data:

So what, people post dumb shit on twitter all the time, the comment’s by Bauer alone wouldn’t be interesting enough for me to write about it. I was a bit more interested in this comment though because Bauer, a self-described nerd, and Boddy who’s made a career in the science of baseball pitching, are definitely smart enough to know that without more evidence the assertion is nonsense. So I assumed they had more damning data.

The first thing I saw Bauer address was the evidence behind pine tar increasing spin rate. Which, I mean, I guess is important to have but also, duh. That was never the problem people had with the implication made by Bauer. It’s like if someone didn’t vote for Bagwell because he took steroids, and people were like, wtf are you talking about what evidence do you have? To which the writer would respond, I have plenty of evidence, here’s a bunch that shows steroids increase HR rate.

Luckily, other people were interested in looking into the data as well. Jeff Sullivan looked at the largest increase from last year to this year in spin rate. Of the 30 largest increases, only 2 were Astros pitchers, which would never pass a hypothesis test. This suggests that there isn’t evidence that the Astros started their systematic cheating this year. But it could very well be an established Astros practice started before this year. Eno Sarris over at The Athletic Link took a big step back and actually looked at the original statement made by @Blaze4551 that started all the suspicion. He actually looked a spin rates for guys that have started for the Astros and for other teams: Scott Kazmir, Doug Fister, Gerrit Cole, Charlie Morton, and Justin Verlander. Scott Kazmir and Doug Fister had nearly the same spin for Houston and Non-Houston teams. How about the three guys that saw ‘dramatic’ increases? Verlander’s spin outside of Houston was 2535 and for Houston its 2591. So I guess it’s only two guys that had a dramatic increase. Charlie Morton’s non-Houston spin was 2103 and now it’s 2244, pretty significant. But Eno took a little bit deeper look and explained “Charlie Morton’s spin increase is interesting, but spin is related to velocity, and when Morton’s velocity was up in his abbreviated season with the Phillies, his spin rate was also up. So really, his spin was up to around 2200 right before the Astros picked him up.”

That just leaves Cole. Who’s spin rate has gone up from 2165 (non-Houston) to 2332 (according to Eno’s article). After reading Eno’s article I began to doubt that Bauer has any real evidence that Houston Is doing anything suspicious. Unless there is really something strange about that Cole spin rate increase that can’t be explained by anything else. So I decided to try and take a stab at understanding what could impact spin rate. Release point, inning, pitch location, stadium (trackman calibration) and other more obvious things (pitch type, velocity) are what I’d want to test to see if they have an impact on spin rate, an see if it could explain Cole’s change relative to random variation.

Anytime I embark on such a journey, it’s easy to start with things you know are likely to be statistically significant. So I simply looked at Cole’s fastballs in 2016 & 2017 and compared the spin to 2018. My data matched closely with Eno’s. 2016 and 2017 Cole’s spin rate averaged 2158 and in 2018 it averaged 2324. I then ran a couple quick linear regressions to control for fastball type (FT and FF) for velocities impact on spin rate. This should be a first step for understanding how much the changes in the variable that impact spin rate contributed to the change in spin rate.

This was indeed a good first step because as it turns out, it was also my last step. Using the regressions from 2016 and 2017 to estimate Cole’s 2018 spin rate I get 2316 (compared to the actual of 2324). Holy crap, that’s it, that’s the answer. The observed increase in 2018 fastball spin rate is explained changes to his repertoire and velocity. At first I was happy that it took like 15 minutes to do something I assumed would take hours/days. That feeling quickly faded. I really hope I made a mistake in my math, or they truly know something we don’t, because if not, this would literally be two douches (I’m not counting @Blaze4551) who obviously know that not only isn’t there evidence that the Astros are doing some type of systematic cheating, there isn’t even evidence supporting the very basis of their suspicion.

Update:

Since I’ve done the work, Zack Meisel posted an article on the Athletic Link with a follow up from Bauer. It’s pretty clear that he wanted to start a discussion about how foreign substance and he was just using that thread to draw attention to the work. Now that’s all I’m comfortable saying with the evidence I have.

But as long as we are taking liberties to call out people’s character for no particular reason without much evidence, I think Bauer’s motivation wasn’t really to draw attention to the relationship between pine tar usage and spin rate. Rather, I think he wants to draw attention to his work to show everyone how smart he thinks he is. If Bauer didn’t do the work, and was presented with the same results from the work, he would not give a shit. I’ve gone to school with a hundred kids just like him, they think people are super impressed with how smart they are and all they talk about is about fancy sciency shit they are learning or doing and they talk about it in the smuggest way possible because they legitimately think people are interested in their intelligence. A real quote from the Meisel article.

“I’ve tested sunscreen and rosin,” Bauer said. “I’ve tested pine tar sticks. I’ve tested the liquid pine tar. I made my own non-Newtonian fluids. I sat down with a chemical engineer to understand it. I’ve melted down firm-grip and Coca-Cola and pine tar together. I’ve tested a lot of stuff.”

As a chemical engineer myself, I can also provide my professional opinion that talking about learning about and producing non-newtonion fluids to a baseball reporter makes you a gigantic douchebag.