One of the nice things about the Gold Glove Awards this year was that they were based entirely on statistics for 2020. It meant some players who have been overlooked for baseball’s top defensive honor got a chance to shine. It also added an element of transparency and reason to the process that boosted the legitimacy of a process that is often fraught with second guessing and raised eyebrows.
Which brings us to a process that didn’t rely on statistics this year: Cy Young Award voting. And don’t get me wrong, there was a lot of cohesion around the top pitchers in both leagues just like there always is — but there was also the one writer who just had to be different. In this case that writer is J.P Hoornstra from the Orange County Register (CA) and for “reasons” I have some issues with, Yu Darvish just didn’t make his ballot at all.
I want to pause here to say I have a bit of experience with evaluating awards a lot of people think are subjective. In my day job I’ve judged thousands of competitive debate rounds and I’m part of a committee charged each year with differentiating some of the top debate teams from one another to determine the last few spots at nationals. Any system that asks experts to rank the top three, or five, or ten of whatever has some level of subjectivity to it, and someone will always take issue with your methodology. That’s why it’s really important to have a coherent structure for why that methodology exists and valid explanations for any deviation from that methodology. That coherence and consistency is crucial to the legitimacy of the ranking.
Unfortunately for Hoornstra his explanation for his ballot lacks both coherence and consistency. I’d be more impressed with his explanation if he’d just written “I don’t really like Yu Darvish.”
Let’s start with the subheading of the article: “It’s nothing personal. It’s mostly numerical, and in the end, it came down to an aesthetic preference.” Those words barely make sense together. If you have a system based on a certain set of stats that you follow it’s numeric, if you want to reward certain “aesthetic preferences” for Devin Williams’ change-up while not recognizing the same “aesthetic preferences” for Darvish’s stunning array of pitches that seems...I don’t know. Maybe not personal, but definitely like you just woke up and decided you wanted to vote for this guy instead because you wanted to. For what it’s worth, that is totally within the right of any voter, but don’t try to couch it in objectivity:
Yu Darvish, Dirty (literally) 88mph Cutter**...and Sword. ⚔️— Rob Friedman (@PitchingNinja) August 19, 2020
**Subject to Yu correcting me. pic.twitter.com/y3GCWR2wWN
But let’s talk about that “mostly numerical” part, because this is where Hoornstra’s argument falls apart entirely. From his explanation:
The strangest part of 2020 for a voter was the statistics. They disagreed a lot. I started with the number I look at most during a “normal year,” Baseball Prospectus’ Deserved Runs Average. Simply put, it’s the most advanced pitching metric available to the public. You’d need a stack of napkins to write out the full formula, but the basic principle behind DRA is noble. It attempts to account for all the things a pitcher can’t control: how well his catcher frames each pitch, how well his defense performs when he pitches, the ballpark he plays in, the quality of the batters he faced, etc. As with any number, I was curious to see what story DRA would tell about pitching in 2020.
The story was this: the Central Division teams offered weaker competition than their East and West counterparts. There was no inter-divisional play in 2020, so that meant a lot. Trevor Bauer, the winner of the Cy Young Award and the first pitcher listed on my ballot, handily led the National League with a 1.73 earned-run average. ERA didn’t consider anything more than his earned runs and innings pitched. DRA did, and Bauer finished 2020 with a 2.89 DRA. That was higher than five pitchers who threw at least 50 innings this season (Aaron Nola, Pablo Lopez, Jacob deGrom, Corbin Burnes and Zach Eflin).
Got that? The stats were less consistent this year due to sample sizes but he’s going to use DRA from Baseball Prospectus because he thinks it does the best job of isolating factors beyond pitchers controls. Plus, DRA (and it’s counterpart for hitters, DRC+) has systemically determined that the Central Divisions (NL and AL) were weaker offensively than the Eastern or Western divisions, so obviously that will hurt all players in the Central Division equally, right?
Somebody has some explaining to do. pic.twitter.com/K5LW4HpInV— Evan Altman (@DEvanAltman) November 11, 2020
So, somehow, the “mostly numerical” standard can be overridden to keep Trevor Bauer first because:
Still, DRA seemed to disagree with ERA more than it does after a 162-game season. That’s dissatisfying. ERA does a decent job of telling us if a pitcher did his main job – preventing runs – in the context of the actual games on the field. And is it fair to withhold a vote from Bauer (1.73 ERA) and Yu Darvish (2.01) because they pitched against lesser hitters than deGrom (2.38 ERA) and Nola (3.28)?
In Bauer’s case, I decided, the answer was a resounding no. I didn’t need a graphing calculator to see his 1.73 ERA was head and shoulders above the rest of the National League. It was, in fact, 176 percent better than average during the mini-season. Shane Bieber, the unanimous pick for the AL Cy Young Award, had a 1.63 ERA that was 181 percent better than league average. It helped Bauer’s case that he allowed feather-light contact all season. He was, in other words, performing like you’d expect the best pitcher in the league to perform against the hitters he faced.
Wait, what? So we are using DRA to bump Darvish off the ballot entirely but it doesn’t impact Bauer (or Bieber, in Hoornstra’s opinion, although it’s worth noting he didn’t have a ballot in the AL) at all, it just impacts Darvish relative to Jacob deGrom, Dinelson Lamet and Aaron Nola. What’s even more interesting is it’s also apparently not a mark against Devin Williams who, remember, gets an “aesthetic preference” boost.
I am tired after all of those mental gymnastics.
As much as I wanted Yu Darvish to win the 2020 NL Cy Young I happen to believe most of the writers got it correct. There is a stronger argument for Bauer’s Cy Young case than Darvish’s and this piece isn’t a plea to change to voting process or anything of the sort.
However, being a baseball writer with a ballot for postseason awards is a privilege and those writers owe the public a coherent process. On occasion even the best processes will result in unexpected ballots and we can all learn a lot from those explanations. “Mostly numerical” with a splash of “aesthetic preference” isn’t a coherent process. Further obfuscating that process with an inconsistent application of baseball statistics only serves to make fans less informed.