This article is part of our The Z Files series.
One of my chores around here is generating weekly pitching rankings. I'd like to think they offer a unique perspective since they're mostly formulaic. That said, a system's efficacy is only as good as its inputs, and the primary driver of the rankings is the rest of season expectations for each hurler.
Much like is done with the official rest-of-season site projections, I have a means of adapting my initial expectations to in-season performances. Essentially, I adjust each skill to be a weighted average between initial and year-to-date. Each skill (K%, BB%, HR% etc.) is not adjusted the same, as some reflect a player's current skill level sooner than others.
While back-testing has proven to do a good job of capturing new performance levels, there are always cases in which I was just plain wrong on the original projection, thus the altered version never catches up. We're at the point of the season where I review the pitchers whose current numbers differ the most from my preseason projections. In many instances, I'll overwrite the initial projection to better represent the new level of skills, then I let the rest-of-season engine take over. Here are 11 such pitchers.
Tyler Anderson (Initial ERA: 4.59, Current ERA: 3.30)
Anderson's low placement was justly questioned in this week's pitching rankings, serving as the impetus for this piece. After he posted his best effort of the season Monday, it was clear his initial projection required some tweaking. His current ERA is fully supported
One of my chores around here is generating weekly pitching rankings. I'd like to think they offer a unique perspective since they're mostly formulaic. That said, a system's efficacy is only as good as its inputs, and the primary driver of the rankings is the rest of season expectations for each hurler.
Much like is done with the official rest-of-season site projections, I have a means of adapting my initial expectations to in-season performances. Essentially, I adjust each skill to be a weighted average between initial and year-to-date. Each skill (K%, BB%, HR% etc.) is not adjusted the same, as some reflect a player's current skill level sooner than others.
While back-testing has proven to do a good job of capturing new performance levels, there are always cases in which I was just plain wrong on the original projection, thus the altered version never catches up. We're at the point of the season where I review the pitchers whose current numbers differ the most from my preseason projections. In many instances, I'll overwrite the initial projection to better represent the new level of skills, then I let the rest-of-season engine take over. Here are 11 such pitchers.
Tyler Anderson (Initial ERA: 4.59, Current ERA: 3.30)
Anderson's low placement was justly questioned in this week's pitching rankings, serving as the impetus for this piece. After he posted his best effort of the season Monday, it was clear his initial projection required some tweaking. His current ERA is fully supported by his underlying metrics and is corroborated by ERA estimators.
For what it's worth, Anderson was a global miss, but that's no excuse. It's time to fix the oversight. His skills were bettered across the board, dropping his initial ERA to 4.24. To some, this may still be high. However, Anderson posted a 4.49 ERA over the past two seasons while splitting time in Pittsburgh, Seattle and San Francisco, all prime pitching venues.
Rest of season ERA projection: 4.03
Martin Perez (Initial ERA: 4.94, Current ERA: 1.64)
Perez's xFIP is 3.92, with a 3.69 SIERA. These expected ERAs help normalize allowing no homers in 56.1 innings and a .268 BABIP. Perez's walk rate is a personal best 6.4 percent, as is his 20.2 percent strikeout clip. Even so, his 13.8 percent K-BB is pedestrian.
While it's clear Perez is getting by with smoke and mirrors, landing on a rest-of-season expectation isn't straightforward. Clearly, he'll surrender some long balls, and perhaps he's unduly benefitting from the baseball, but his original home run rate was likely too generous. That said, the adjustments based on allowing no dingers yet this season serves to balance things. This is a case of simply trusting the system, as I'm doing nothing.
Rest of season ERA projection: 4.13
Brady Singer (Initial ERA: 4.26, Current ERA: 1.83)
Singer only has 19.1 innings so far, akin to around three starts. Normally, I don't adjust after only three starts, but Singer has pitched so well since entering the rotation, he merits a look. His expected ERAs are all under 3.00, so his success is from more than just good luck. He's throwing his changeup more and even though the delta between it and his fastball is only 6.6 mph, that's one mph more than last season. Singer's early 5.3 percent walk rate is a career low while his 60.4 percent groundball rate is a career high. His current rest-of-season ERA is projected to be 4.12, so it's already on the way down.
While I like the trends, I've learned not to overreact to such a small sample, so while I made an adjustment, it was only to slightly lower the initially projected walk and home run rates.
Rest of season ERA projection: 4.03
Patrick Sandoval (Initial ERA: 4.09, Current ERA: 1.79)
It can be argued Sandoval's strikeout and walk rate are a bit worse than expected. He's been aided by a .270 BABIP and 80 percent LOB mark. His 3.75 xFIP and 3.96 SIERA are close to my 3.76 rest-of-season ERA, so no adjustments were made.
Rest of season ERA projection: 3.76
Taijuan Walker (Initial ERA: 4.28, Current ERA: 2.70)
Four of Walker's six efforts have been scoreless, but his 14.9 percent strikeout rate is a career low. He's benefiting from a .261 BABIP and 77.6 percent LOB clip. His velocity is down, but I'm intrigued by increased usage of his splitter.
Walker's 4.12 xFIP and 4.46 SIERA portend an ERA correction. No changes were made.
Rest of season ERA projection: 4.15
Cole Irvin (Initial ERA: 4.39, Current ERA: 3.21)
Irvin's already low strikeout rate is trending even lower this season, though his swinging and called strike marks are both higher than last season, so he could see his strikeout rate inch up. Even so, a .269 BABIP and 86 percent LOB mark have suppressed his ERA, as indicated by a 4.28 xFIP and 4.43 SIERA. Irvin's initial projection was left untouched.
Rest of season ERA projection: 4.16
Tyler Mahle (Initial ERA: 3.99, Current ERA: 6.32)
Mahle has a history of struggling at home, where his home run issues are exacerbated. He's again having trouble in the Great American Ball Park, but it's more about control and command than gopheritis. His 12.3 percent walk rate at home, in tandem with a .348 BABIP, have resulted in far too much traffic. His strikeouts are also down.
My concern here is I was aggressive with my initial projection. All Mahle's publicly available preseason ERA projections on Fangraphs were over 4.00. My friends at BaseballHQ shared my optimism, but unabashed Reds fan Jeff Erickson's official site projection was 4.24.
Without an adjustment, Mahle's rest-of-season expectation is a 4.16 ERA. Considering his performance to date, that feels light. As such, I tempered enthusiasm for improved strikeout and walk rates, the two areas I differed most from the field.
Rest of season ERA projection: 4.37
Alex Cobb (Initial ERA: 3.74, Current ERA: 6.25)
It's no secret; Cobb has been wholly unlucky. His BABIP is .411 while his LOB mark is a silly low 49 percent. His xFIP is 2.25 with a 2.54 SIERA. Cobb's 21.3 percent K-BB is a respectable 24th among hurlers with at least 30 frames.
To be honest, an argument can be tendered for Cobb's rest-of-season ERA to be lower than initially projected. However, the little black box pegs it to go up a tad, and I'm going to let it stand.
Rest of season ERA projection: 3.85
Hyun Jin Ryu (Initial ERA: 3.89, Current ERA: 5.48)
Coming into the season, Ryu's strikeout rate was a great example of a projection quandary. Ask someone what number comes next after 4-3-2, and most would say 1. However, someone who does fantasy baseball projections may apply a weighted average, landing on something closer to 3. While Ryu's strikeout rate wasn't exactly like that, he incurred a big dip last season, so the question was if it would rebound or continue downward. So far this season, it's been the latter.
Sure, Ryu has been unlucky with a 63.4 percent LOB mark and a .308 BABIP, but his 12.4 percent strikeout rate is horrible. However, since my projected level assumed a rebound, the current rest-of-season clip is too generous, resulting in a 3.95 ERA. That seems low, even with the bad luck flushed away. Therefore, I lowered Ryu's initial strikeout rate and will let the system take care of the rest.
Rest of season ERA projection: 4.14
Kyle Wright (Initial ERA: 4.85, Current ERA: 2.68)
Wright is a challenge since his major-league track record was limited to a poor 6.56 ERA and 1.69 WHIP in 70 stanzas spread over four seasons. Further, his minor-league numbers weren't special, especially since he was usually old for his level. Still, what he's done thus far is impossible to ignore given his dominant 28.4 percent strikeout rate.
Cutting to the chase, his initial projection was lowered to a 4.22 ERA. If he continues on this pace, the rest-of-season projection will continue to improve, but for now…
Rest of season ERA projection: 3.97
Paul Blackburn (Initial ERA: 5.11, Current ERA: 1.75)
Blackburn's xFIP is a tidy 3.37, and in tandem with a 3.72 SIERA it is apparent he's been both lucky and good. That said, his 13.0 percent K-BB is below league average. Blackburn's bugaboo has always been the long ball, and he's sported high HR/FB marks despite pitching in a division with mostly power-suppressing venues, including his home park.
His strikeouts and walks match his track record, but lowering Blackburn's projected homers seems reasonable, so let's do it.
Rest of season ERA projection: 4.06
