Potential Starting Pitcher HR Rate Improvers — May 26, 2021
Earlier this month, I reviewed the hitters who had most underperformed and overperformed their Statcast xHRs. But Statcast doesn’t only calculate xHRs for hitters, it does so for pitchers as well. So let’s review the starting pitchers that have allowed at least two more home runs than Statcast has calculated that they should have been expected to allow.
Player | Actual HR | xHR | HR-xHR |
---|---|---|---|
Trevor Williams | 8 | 5.4 | 2.6 |
Triston McKenzie | 8 | 5.6 | 2.4 |
Dean Kremer | 10 | 7.7 | 2.3 |
Patrick Corbin | 10 | 7.7 | 2.3 |
Anthony DeSclafani | 6 | 3.8 | 2.2 |
Adam Wainwright | 9 | 6.8 | 2.2 |
Yusei Kikuchi | 10 | 7.8 | 2.2 |
Chad Kuhl | 4 | 1.9 | 2.1 |
Lucas Giolito | 9 | 6.9 | 2.1 |
With a 23.5% HR/FB rate, it’s not too surprising to find Trevor Williams atop the list. He’s actually a great example of why SIERA is far superior to ERA over a one year sample when forecasting a pitcher’s future performance. In 2018, he massively outperformed his SIERA and explanations of his supposed ability to induce soft contact, leading to a .261 BABIP, were everywhere. Sure enough, those supposed skills disappeared quickly, as he hasn’t posted a sub-.300 BABIP since and his ERA has spiked even higher than his already poor SIERA. Even when his fly balls start flying out of parks at a lower rate than they have been so far, he’s still not someone you want to roster in any sort of league, though the increased strikeout rate makes him a bit more interesting, if it lasts.
Although he was demoted to the minors over the weekend, I wanted to discuss Triston McKenzie anyway. After a strong 33.1 inning debut last year, McKenzie endured a bizarre start to this season. While his 19% HR/FB rate is high, it’s not extraordinarily so, so you wouldn’t necessarily think he’s been the second unluckiest starting pitcher in home runs allowed this year according to Statcast. Oddly, his BABIP is significantly below the league average again. I always struggle to explain how a pitcher could suppress BABIP but allow a high HR/FB rate. When the pitcher in question makes a mistake, it’s like super mistakey?
Aside from the weird BABIP/HR/FB rate combo, McKenzie’s control deserted him as he walked an astounding 20.8% of batters faced. He also allowed an absurd 60.9% fly ball rate. He essentially embodied the three true outcomes, but as a pitcher. Encouraging is that his strikeout rate remained above 30%, but with his fastball velocity down and a low called strike rate, it’s hard to imagine that lasting. I honestly don’t know what his future looks like.
Baltimore is not a good place to pitch half your games when you’re an extreme fly ball pitcher like Dean Kremer. Both his strikeout rate and SwStk% are well down from last year’s debut, though both are over a small sample size. Even with an improvement on his 20.4% HR/FB rate, I don’t see the upside here to worth risking your ratios.
Patrick Corbin’s velocity has rebounded off last year’s low, but his strikeout rate and SIERA still sit at the worst levels of his career. Obviously, that 25% HR/FB rate is going to decline, which Statcast already tells us he’s underserving of. But without the strikeouts, it won’t matter, as his 4.72 SIERA confirms. He probably deserves better than the mid-teens strikeout rate, but I would still bet the way over on Steamer’s RoS ERA projection.
Anthony DeSclafani has only allowed six homers on a 10.9% HR/FB rate and yet he still manages to make this list as one of the unluckiest. That’s pretty crazy. As an owner, I shouldn’t have been too surprised about the 10 run blowup, as eventually his ERA was going to rise to meet his much less impressive SIERA. All that regression just so happened to come during one game. His home park could certainly help him keep a below average HR/FB rate, but I wouldn’t bet on a single digit mark the rest of the way. Whatever it ends up, he should remain at least a bottom tier shallow mixed league asset all season.
Adam Wainwright’s career HR/FB rate trend is fascinating. From 2016, his first season with a reasonable number of innings, he recorded eight straight seasons with a single digit mark, plus a ninth in 2015 when he didn’t allow a homer in 28 innings. Then beginning in 2016, he has allowed a double digit HR/FB rate every single season and his current mark would represent a career high. I still can’t believe that he’s still pitching, and that his skills remain respectable enough. Do realize that any improvement in HR/FB rate could be matched by regression in BABIP, so don’t expect his ERA to actually improve, as his ERA is just barely below his SIERA.
With 10 homers allowed and a 24.4% HR/FB rate, I’m surprised Yusei Kikuchi didn’t rank higher on this list! With even higher fastball velocity than last year, which itself represented a surge from his 2019 debut, another jump in SwStk%, an increase in strikeout rate, and a partial rebound in walk rate, you would think he was in the middle of a massive breakout. His ERA has come down to nearly meet his SIERA, but all that has done is bring his ERA right in line with the league average. I’m still holding him in the one league I own him and continue to start him each week. However, I’m not sure he possesses the command to improve much more.
The spike in ERA would have most fantasy owners think Lucas Giolito hasn’t pitcher at anywhere near the level of the last two seasons. However, his SIERA hasn’t increased much, with the majority of that ERA jump driven by that HR/FB rate spike that Statcast thinks is due to bad fortune. There are other small issues as well, like a slight decline in fastball velocity, a drop in strikeout rate to below 30%, and an uptick in walk rate to cross into double digits again. But really, one gem with a high strikeout rate and suddenly his full season stats will look exactly like we all expected.
Mike Podhorzer is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year. He produces player projections using his own forecasting system and is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. His projections helped him win the inaugural 2013 Tout Wars mixed draft league. Follow Mike on Twitter @MikePodhorzer and contact him via email.
Gotta tell ya, the way StatCast calculates this is not useful for fantasy purposes. All they do is examine all balls in play that would have been out of any stadium by their black box calculation, divide the number of stadiums that it would have been a HR in by 30, and add all of those up. It’s crazy inaccurate.
Let’s take an example from Trevor Williams. On May 2, Nick Castellanos hit a fly ball 103.3 MPH off the bat at a 36 degree launch angle that went for a HR. StatCast says this would only have been out of 10 ballparks, therefore it only counts at 1/3 of a xHR. There are 884 FB from 102 to 104 MPH and 35 to 37 degrees of launch angle over the course of the StatCast era. 460 of them have gone for HR, which is 52%. Now, this is all FB, regardless of hitter handedness, pull/straight/oppo, or park.
Furthermore, I have noticed that StatCast underestimates the value of pulled FB in particular. For all pulled FB in the StatCast era, the xwOBA is .613, while the actual wOBA is .869. Castellanos’s HR is in the pull category for StatCast. There have been 396 pulled FB with the aforementioned EV/LA parameters in the StatCast era. A whopping 338 of them have been HR, good for a 85% HR rate. Again, this is regardless of stadium or handedness.
I could go further, but I think I’ve made my point. Don’t get me wrong, I absolutely love StatCast, but I believe there’s a few simple tweaks they could make to improve their calculations immensely.
And that doesn’t deal with outlier distortions. Coors is on the extreme xHR side as an outlier pitchers park because of the OF dimensions and is also the extreme outlier for actual HR.
Ignore what that does for Rockies pitchers – or really anyone who pitches there. It also is a significant enough outlier on both ends to distort league data.
For sure. There are so many unaccounted for variables as to basically render the entire table useless. Furthermore, if someone plays zero games in a particular park throughout the year, the fact that a particular batted ball would be a HR in that park is irrelevant to their HR total. If this were weighted by, say, team games played in each park, it would at least be a step in the right direction. For example, the Astros, Rangers, A’s, and Mariners are the only AL teams to play in Colorado. So a HR Jose Abreu hits that wouldn’t be out of Coors Field means zero for his “expected” HR total.
The concept of the table is fantastic, the execution, not so much.