Imagine that you’re the administrator of a temple in the ancient middle east in the 4th millennium BCE. Your city is almost entirely focused on agriculture. There are 50,000 people within your walls, and they all need to eat. In order to make sure that everyone gets food commensurate with their position, you pool all of your agricultural products, grain, meat, beer, milk, goats, cows, and more, in your temple at the center of the city. With all of this stuff coming in and going out all the time, you need to develop a system of keeping track of everything. What do you do?
Penn Murfee started as an infielder. He played shortstop at Vanderbilt before an injury meant they he couldn’t field anymore. So he took up pitching in order to stay in the game. And his mechanics did not work. He had no command of the ball, just smearing it all over. Until he made a simple change to his mechanics.
In ancient Mesopotamia, the most common resources were dirt and clay. Buildings were made of mud bricks, so that left the clay for things like pottery. Temple administrators had another use for it, however. They started using clay to keep track of things. Using a piece of reed, they drew a little picture of what they were storing, and then made impressions to count how much of that thing they had. This Sumerian tablet, for example, was an accounting list and features the first recorded instance of a math mistake.
This system worked fine, but there was a lot of smearing. It was very easy for scribes to accidentally smudge what they had already written. The writing originally went vertically, top to bottom, but to prevent smearing, they made a simple change to their mechanics. They rotated the symbols 90 degrees.
Penn’s overarm mechanics did not work. But then he got some advice. A friend of his told him to throw sidearm, just like he was turning a double play from short. And that change eventually led to Mariners fans being treated to strikeouts like this.
AppleTV’s 42% probability may have been a little high. That slider is phenomenal, and that’s the pitch I want to talk about here. John and Isabelle collaborated on an excellent write-up on Penn’s fastball back in August, if you want a deep dive into Penn’s other weird pitch.
So how do we measure Penn’s slider? It’s an outlier in several categories. It moves incredibly far horizontally, but it doesn’t really change too much vertically. It generates a ton of strikes, with a 32.4 whiff%, and induces weak contact with a .181 xwOBA. But we can measure it with a different metric: pitch value. This is a stat that attempts to determine how valuable a pitch is by how it changes run expectancy every time it's used. A typical range is -10-10, with 10 being considered very good. So where does Penn’s slider stack up against other relievers?
Oh! That’s really good. Using this metric, Penn has the 4th most valuable slider in MLB bullpens. He drops down to 10th when using wSL/C, the rate metric version of the stat, since he throws it way more than almost anyone else ahead of him except Edwin Díaz and fellow Mariner Andres Muñoz. So we know that Penn’s slider starts from a weird spot, moves like almost no other pitch, and simply cannot be hit hard. So why is it that this is how we remember Penn from last year?
JEREMY PEÑA BREAKS THE ICE IN THE 18TH! #postseason pic.twitter.com/4EkefQ91U3— MLB (@MLB) October 16, 2022
First, I’m sorry for inflicting that clip on you. I promise to not do it again. Second, it’s pretty clear that Penn ran into some problems down the stretch, and that’s reflected in his stats.
Something happened after August 13th (two days after John and Isabelle’s article), and Penn’s rolling wOBA shot up. So what happened? Well, here’s what Penn’s heat maps look like when his pitches are working.
As we can see, Penn’s stuff mostly goes down and away. That makes sense, when you think about it. He has a low arm slot and a ton of horizontal movement. It works for him, and it’s what gives him one of the best sliders in baseball right now. His fastball lives in the zone, but it’s spread out enough that it’s hard for hitters to get a favorable launch angle from it.
But what does this heat map look like starting August 13th to the end of the year?
Oh no. That’s smeared. That’s smeared all over the place, like a Sumerian scribe got a hold of it. Let’s look at the fastball first. After August 13th, Penn’s fastball was either up and in, or in one little cluster just off the center of the plate. If you were a hitter and Penn threw a fastball to you, if you closed your eyes and swung at that cluster, you had nearly a 50% chance of hitting the ball. That’s part of the reason Penn’s Sweet Spot% shot up as well.
And then you have the slider. Penn’s main pitch lost all sense of command. Instead of being painted down and away like a good slider should do, Penn’s slider was either too far outside to be enticing, or just hanging on a tee down the heart of the plate. The data actually shows that Penn got a little lucky, since his rolling xSLG doesn’t really follow the August 13th trend. It was already high, before dipping down low in the summer.
His rolling xBA does follow the trend, however. This matches the data that shows opposing hitters’ increase in Sweet Spot% but not a corresponding increase in exit velocity. Penn was getting hit around, but not very hard.
This is, I would say, good. No increase in average exit velocity suggests that, in a vacuum, Penn’s stuff is still good. His command issues only led to hitters getting more hits off him, but not really barrelling the ball up. In fact, Penn’s wSL/C is actually higher after August 13th than it was before. It’s his wFA/C that plummets all the way down to -3.19. Of course, pitches don’t exist in a vacuum. If a fastball is bad, it means hitters have more chances to sit and wait for a hanging slider, which was more likely to come later in the season.
So what’s a struggling reliever to do? Well, what did they do in ancient Mesopotamia? Drawing those intricate pictograms into clay tablets proved to be too time consuming, and too easy to damage, so they came up with a simple little change.
Instead of drawing the patterns with the point of their stylus, they used the side to make wedge shaped impressions. This new system, cuneiform - literally meaning “wedge shapes”, had the advantage of being quicker and easier. But it was also further obfuscated from the original picture. Now that it was both rotated and constructed differently, it wasn’t long before the drawing became separated from its original meaning. Eventually, people realized you could use these new symbols to represent the sound of a word, and not just the word itself. And thus, writing was born.
I think the solution to Penn’s problems will be equally mechanically simple. Perhaps most frustrating is the fact that he’s already been great. In fact, his late season struggles only made him league average in xWOBA. His Savant page is still very red. But ending last year he was dragging across middle. Just like the ancient Mesopotamians, he needs to push in from the side.