SP Plate Discipline Update #3
(Photo by Patrick Gorski/Icon Sportswire)
This week, instead of picking fan favorites to discuss and then presenting the data dump as an afterthought, I’m going to flip things around and start things off with the raw data. Then I’ll just go down the list and pick some guys to who stand out to me to talk about. So without further ado, here are the updated plate discipline grades and rankings, along with their K% anomalies. Data is through Tuesday May 15.
A new contender has seized the top spot from Max Scherzer, though just barely. I haven’t talked about him yet, so it seems as good a time as any. It may come as a surprise that Sale actually over-performed his metrics a bit last year, with a K% of 36 while the metrics pointed to a rate around 30. Based on this, one of my preseason predictions was that Sale would not lead the league in strikeouts again. But if he keeps it up at this rate, it’s becoming much more likely. He’s currently posting career-high marks in Contact% and SwStr%, while maintaining the elite O-Swing%. So his 34% K rate is very much in line with the metrics, and more sustainable than last year’s performance.
Sale’s pitch mix is very similar to last year, but his pitches are performing better. It has most to do with his changeup, which is getting the best results of his career. In every previous year, his changeup had positive value for vertical movement, indicating less drop than gravity alone would provide. But this year it’s a negative number, indicating more drop. Drop is generally a good thing for changeups, as they are commonly thrown near the bottom of the zone intended to get whiffs or induce weak grounders. And so the fact that he’s allowing a wRC+ of just 18 on his changeup, with a 24% whiff rate, is not surprising. Of course his other pitches are fantastic as well. Sale is looking every inch like the stud you expected, and even a little better than that.
[Update – Sale pitched yesterday which is not included in the data, but it was another typically solid performance and shouldn’t change anything]
#4. Patrick Corbin (SP, Arizona D-Backs)
I covered Corbin fairly extensively last week, so I’m not going to do that again. But I do think it’s important to mention that even with his reduced velocity the past three games, his plate discipline metrics still point towards a K% in the upper 20s, which is really not bad, especially considering Corbin is also a strong contact-manager with a 50% groundball rate. During that three-game stretch he still put up a Contact% of 72, which is top 20. Furthermore, his fastball velocity is ticking back upward in each of those starts. So there are certainly some reasons for optimism here, and I’d caution against panic selling. He’s still ranked #4 overall, and it seems unlikely you’re going to get value in return anywhere close to that.
Early in the season it was difficult to pin a value on Ohtani, as he went through some swings in performance, and was also much slower than most SP to accrue a decent sample size of innings. But now after six starts, the metrics are painting him as very much the real deal. His strikeout rate of 33% is in line with expectations, so he’s certainly looking elite in that area. His contact management skill also grades out nicely so far, with an xSLG of .344. That’s is the neighborhood of Arrieta, Sale and Corbin. One area where he’s not quite so elite is his O-Swing%, so we should expect walks to continue to be somewhat of a weakness.
Ohtani so far in his MLB career is mainly a three-pitch guy. He throws his fastball about 45% of the time, then a slider and splitter about 25% each. The key pitch for Ohtani is his splitter. His slider is also pretty good, but his truly elite pitch is that splitter. It currently carries a 29% whiff rate, and is allowing a wRC+ of negative 76. Accordingly, it shouldn’t come as a surprise that it’s rated the best splitter in MLB in terms of pitch value, and one of the best pitches period. I’m curious what would happen if he threw it a bit more often. Subbing out the occasional slider for his more dominant splitter may be a way for him to improve even more.
Typically when a pitcher doesn’t even have a FanGraphs profile for the past two seasons, that’s not a good sign. As a contact-oriented pitcher who played for the Rockies, it’s understandable that he has been ignored for fantasy purposes in prior years. But so far this year, he hasn’t looked like the same pitcher. His strikeout rate is sitting at 25%, compared to just 15% for his career. What do the metrics tell us about that strikeout rate? Well actually…it may be legit. His Contact% and SwStr% are both sitting at career highs, in line with his K%, and furthermore his O-Swing% is up to an impressive, elite level. These improvements, plus a move from Coors to Petco, make him suddenly quite intriguing. It may also come as a surprise that he is only 27 years old.
The main concern here is the small sample size of starting innings. Lyles began the season out of the bullpen, and has only started two games. Typically a pitcher adds a bit more velocity in relief, which in turn tends to improve the PD metrics and strikeout numbers. In Lyles case this is partially true – his velocity was higher in relief. But actually his metrics got better when he moved to the starter role. However, those excellent metrics were mostly a result of completely dominating the Rockies in one game, to the tune of 1 hit and 10 strikeouts over 7.1 innings. This may come as a surprise, but the Rockies actually have the worst offense in MLB when you remove park effects (using wRC+). So there’s definitely some strength-of-schedule concern, in addition to the small sample size. But he still seems like someone worth keeping an eye on right now.
One note, is that you’ll notice his predicted K% shows a fairly large anomaly. Those numbers are just from his two starts, where he did indeed overperform his metrics. But his complete body of work including the relief work does support his overall K% near 25.
It’s not often that a pitcher reinvents himself at age 34, but on the surface that appears to be exactly what Verlander did last year upon being traded to the Astros. Since the trade, he has been completely, utterly dominant. There’s no denying that. But at the same time, he’s also been very lucky in some ways. It seems like a worthwhile endeavor to see if we can determine which types of luck he is benefiting from, and which is he isn’t. On one hand, the huge spike in Verlander’s K rate does not appear to be lucky, in the sense that they are driven by real improvements in his plate discipline metrics. He is currently blowing away his career highs in Contact% and SwStr%. On the other hand, he’s put up incredibly lucky LOB and BABIP numbers, so his minuscule ERA is not fully to be trusted either.
ERA luck aside, let’s focus on the K rate. It would appear something has changed to support such a big improvement up to a career high of 34%. So, what did the Astros do to get so much more out of Verlander? Honestly, having examined the data, it’s not entirely clear that anything significant changed. One trend has been decreasing his changeup usage and adding more sliders instead, to the point where the changeup is now essentially eliminated from his repertoire. He is basically just throwing a fastball, slider, and curve now. But Verlander’s changeup and slider are both similarly good pitches, so that wouldn’t appear to make such a huge difference. Mainly, it just looks like he is getting much better results across the board of those three pitches. All three have lower contact rates and higher swinging strike rates than he’s ever got before on those pitches. He throws his 4-seamer by far the most often, so the bulk of the overall improvement is coming from that pitch. For his career, the four-seamer has allowed a wRC+ of 113. But this year, it’s generating a wRC+ of just 41. The pitch isn’t showing more velocity or movement than in the past, so this seems like a bit of a red flag to me. This led me to examine his game log to check the strength-of-schedule. If contact is down across the board with no real explanation, he could just be playing bad teams. As it turns out, of his nine starts this year, seven have been against below-average offenses. So we can add one more type of luck to the list he is benefiting from – schedule luck. In particular, he has been very fortunate to face the Rangers three times, who are the 3rd worst overall offense and strike out the most of any team.
Overall, there seem to be plenty of reasons why you might want to consider selling high here. Don’t get me wrong, he is absolutely pitching great, but expecting it to continue at this level just isn’t realistic.
[Update – Verlander pitched another gem yesterday (CG shutout) which is not included in the data. He played against a better offense (the Angels), and struck out 20% of the batters rather than 30+. He got it done with a lot of soft contact, but also he’s currently over-performing his xSLG by over 60 points. His BABIP in that game was .200, and the stars still seem to be aligned]
Like his teammate Verlander, Charlie Morton has been damn near unhittable lately, taking his pitching to another level in his mid-thirties. And that’s not where the comparison ends, either. He’s currently sporting a career high strikeout rate of 32%, which is backed up by also career-high Contact% and SwStr% numbers. But foul balls and umpires seem like the only type of luck he isn’t benefiting from. His LOB% is over ninety, and his BABIP just .230. Furthermore, Morton has had a fairly easy schedule so far. The top three teams in MLB who strike out the most are the Rangers, Padres and D-Backs. Half of Morton’s starts are against those three teams.
Unlike Verlander though, Morton has made some meaningful changes to his pitch usage this year. His bread-and-butter has long been his sinker, which for his career he’s thrown 46% of the time, and 41% last year. But this year he is reducing its usage down to just 28%, making up the difference mostly by throwing more 4-seamers up in the zone. Along with the increased 4-seam usage, his infield fly rate has soared to almost 20%. Popups are as good as strikeouts, so that’s nice. His third pitch is his curveball, which is his best, allowing just an wRC+ of 11. All three of those pitches are roughly equal in their usage. That balance between two different types of mid-90s fastballs, and his curveball at 80 MPH seem like a very effective mix of pitches. He also throws occasional cutters and splitter, which both sit around 88 MPH – almost exactly halfway between his fastballs and curveballs. So hitters really have no idea what to expect against Morton – it could be high, low, fast, slow, or in between. That seems like a solid recipe for success, even if he does fall off a bit from his current pace.
#19. Blake Snell (SP, Tampa Rays)
Here is a breakout that could have been predicted from plate discipline metrics. In 2017 he made very real improvements over the course of the year, with markedly better metrics in the second half than in the first. He improved his O-Swing% by 8 points, Contact% by 4 points, and SwStr by over 3 points. Unfortunately, these types of things can be completely hidden among the full-season stats, if you’re not looking for them, which I wasn’t. Maybe identifying similarly improved players will be a fun project for the next off-season. Anyway, this year he has continued to build on those improvements, and his stat line stands out more. His K% of 26 is fully supported – actually he could even go a bit higher, as the metrics point towards a mark around 30.
The only change in his pitch selection this year seems to be a few less changeups and a few more curveballs. His curveball is looking great this year, showing improved movement from previous years. At the same time, his changeup is less effective both in terms of movement and results. So that seems like a good adjustment to make. But I do wonder why he throws so few sliders. Snell’s slider has a whiff rate in the same neighborhood as Corbin or Bundy’s (i.e. excellent), but he throws it just 13 percent of the time. Even with the reduced changeup usage, he’s still throwing that changeup the second-most of any pitches, behind just the fastball. Given that his changeup and slider have the exact same velocity, but the slider is just better, I’m really curious to see what might happen if he cut down a bit on the changeups in favor of sliders.
Despite the breakout, there is still at least one reason to hesitate here. His BABIP sits at just .233. That is probably not sustainable, especially given the fact that his contact management skill seems less than elite. His xSLG mark falls in the neighborhood of Robbie Ray and Masahiro Tanaka – guys with excellent strikeout stuff, but who get hit hard and give up a lot of home runs. A lot of that can be masked by pitching at home, but it seems likely that the occasional blowups on the road will continue to be a problem for him, especially playing in the AL East.
Wait…who? Actually, don’t worry, it’s not another name you need to learn. I am just mentioning Velazquez to quickly explain why you can gloss over him. He grades out highly from two starts in which he posted some pretty good whiff rates, but then in 15 innings of relief he struck out just eight batters. His overall metrics are in line with his K-rate of just 17%. Those two starts seem like small sample noise, nothing more. If that isn’t enticing enough, he also just went on the DL with a back injury.
Without looking at his numbers, I would have guessed Severino was higher on the list. With last year’s superb breakout year, followed by his excellent work thus far in 2018 resulting in an ERA and FIP of just two, he seems like every inch a fantasy ace. So why is he only ranked #25 around the likes of Kenta Maeda and Jon Gray? There are two reasons. The first one has to do with his plate discipline metrics and K rate. Last year he outperformed his K%, very slightly, by about 2.5 points. That is within the margin for error, i.e. some pitchers have the skill to do that, so no worries. This year though, his metrics have all taken a small step backwards, but his strikeout rate is actually slightly up. With the two sets of stats moving in opposite directions, that 2.5 percent gap has now doubled to about 5 percent, which is at the very top end of the margin for error. So he is now looking more like a likely regression candidate on his K rate, though not as glaring as some others.
The second explanation for his success with a relatively low grade is probably the much bigger one, and has to do with his contact management. He is currently looking incredibly lucky in that area. For example his hard-hit% is essentially the same as last year, but his HR/9 rate has cratered to just one-third his career rate. If you prefer StatCast data (you should) how about this: his xSLG of .395 is over 100 points higher than his actual SLG of .270. That is a sizable difference, one of the largest among MLB starters. And it’s not like he plays in a pitcher’s park which could explain it. Severino is a very talented young pitcher, but I suspect some owners who are banking on him as a 100% safe #1 Ace might end up a little disappointed.
Like many fantasy owners, typically I just gloss over anyone who pitches for the Rockies, so I admit I don’t know much about Anderson. I talked about Jon Gray last week, but Anderson is actually neck & neck with him as the highest ranked Rockies pitcher. Let’s see if he might be worth a closer look. First of all, yes, his plate discipline metrics are trending upward this year. His “predicted K%” is up about 3 points from last year, to about 27%, while his real strikeout rate is just 24%. That is an indication of more strikeout upside, though not a particularly strong one.
Next we can ask: “But why did his metrics improve?” This is where I typically look at velocity and pitch usage. For Anderson, his velocity is one reason to believe. He started out the season sitting around 92 MPH, but it’s quickly improving over the young season. His average was up a full tick to 93 MPH by his last start, and he’s even hit 96 on occasion. That can’t be bad.
When it comes to pitch usage, it’s more of a mixed bag. He’s definitely made changes, but I don’t think I like them. The main change he made was throwing less “regular” fastballs and also less changeups, while throwing a new sinker instead. He’s throwing that sinker 15% of the time, but it really doesn’t seem to be working at all. Sinkers are supposed to get groundballs, but he’s getting just a 22% GB rate on it. Instead, it’s getting almost 40% line drives, and overall allowing a wRC+ of 150. He also can’t seem to control it, as it has the highest walk rate of any of his pitches. Basically, it just looks bad and he should stop throwing it. On the plus side, he has good alternatives to choose from. There was really no reason for him to go away from the changeup in the first place, as it sports a solid 20% whiff rate for his career. He also mixes in a curveball on occasion (4%) which looks promising. He’s thrown just 30 of them this year, but not given up a single hit or walk. If I was the Rockies pitching coach, I’d have him abandon the sinker immediately in favor of more changeups and curveballs. For now I guess he’s a streamer, thanks to Coors.
Just looking at his stat lines the last couple years, and then this year, something clearly seems off. Gibson has a career ERA of 4.6, and K% of just sixteen. But this year he has raised his K% up ten points to 26%, and his ERA has fallen to a respectable 3.4. I can assure you that his increased strikeouts is not a mirage. He has increased his Contact% and SwStr% significantly, to career bests (by far).
As to how this is happening, I think Nick hit the nail on the head in Gibson’s FanGraphs profile for this year. He correctly noted that Gibson was much better towards the end of last season after he made some changes to his approach with the slider, and it had the potential to carry through into 2018. That seems to be exactly what is happening, with the slider continuing to get very impressive results. His slider is currently the 5th most valuable in all of MLB on a per-pitch basis (wSL/C). It’s sporting a 33% whiff rate, a 32% contact rate, and allowing an wRC+ of negative 41. He’s thrown 110 sliders this year and given up literally ONE hit. It’s hard to imagine a pitch being much more effective than that. He’s only throwing it about 15% of the time, so again there could be some room to improve if he throws it a bit more often.
The Angels sixth starter, I was surprised to see Barria this high on the list. His strikeout rate around 20% is supported by the metrics, and maybe even a bit low, but that still doesn’t explain his B+ grade. The main contributor to his score is actually O-Swing%. His O-Swing% currently sits close to 40%, which would be tops in all MLB if he qualified. That is just…weird. Maybe it’s just a small sample, I’m not sure, but he seems to be doing something right. Unfortunately it doesn’t appear to be enough to make him fantasy relevant. With a 20% K rate, you need to be a very good contact manager. Barria has been incredibly lucky this year on that side of things, with his actual SLG% 150 points under his Statcast xSLG, an even larger difference than Severino. I would stay away here unless you are desperate for a streamer.