(Photo by Juan Salas/Icon Sportswire)
Welcome back to my “Plate Discipline” series. This week, I’m trying something new and exciting – broadening the scope of my analysis by incorporating StatCast data. Previously, contact management was a completely missing element in my ranking system, which led to some strange results and a lot of double-checking other sites to figure out why. StatCast is the best thing we’ve got right now to measure quality of contact, so my goal with this is to make my rankings more comprehensive of a pitcher’s full skill set, and help identify other types of luck beyond just strikeout rates. I’m open to feedback, so if you have any thoughts about the new ranking system please let me know in the comments!
The specific metric from StatCast that I’ve chosen to incorporate into my rankings is xSLG. why xSLG? The answer is simple, but twofold:
1) I wanted this to capture “power luck”. This should cover luck factors like ballpark dimensions, weather conditions, and outfield defense. If a pitcher has given up many “cheap” homers relative to league average, they would be considered unlucky, and vice versa.
2) I wanted this to be completely independent of strikeout rates, which are covered already by the PD metrics. xwOBA would be the other option here, but that does include strikeouts (and walks) so I went with xSLG.
The method I used to update my scoring system is as follows:
1) Take the inverse of the xSLG (1-xSLG) because for SLG, a higher number is bad for the pitcher. We want a higher number to be good.
2) Multiply this by 1.5. This brings it to roughly the same level as PD scores, with almost all pitchers falling between 50% and 100%. Just a handful of the top pitchers can barely top 100%.
3) Add the PD score to this number, and divide by two to bring the number back to the 100 scale.
You may be asking – “Why should plate discipline and xSLG have equal weight?” and that’s a great question. For starters, in the MLB in 2018, strikeouts and walks compose just 31% of plate appearances. The other 69% end with a batted ball, so it’s easy to see that contact management is very important. I considered giving xSLG the full 69% weight, but I wanted to apply a significant discount for two reasons:
1) Going back to the premise of why we care so much about plate discipline metrics in the first place, strikeouts are just the most reliable method of getting outs. Managing contact is great, but depending on this can be a bit more risky for a pitcher. I want to make sure my rankings reflect that.
2) This is for fantasy purposes, and strikeouts are still a fantasy category. A strikeout is just worth more than other outs for fantasy purposes, so I didn’t want to make all outs completely equal.
With that out the way, here is the new-look data table. Data is through Tuesday May 22.
Some new columns have been added this week:
1) ERA, just to help provide context and tie everything back to the real world results
2) xSLG, as explained above
3) SLG – xSLG. This is included for reference to save everyone a trip to the Savant website. It should be used similar to the “K% difference” column, where a positive number indicates bad luck so far (i.e. likely to regress positively), while a negative number indicates good luck so far (likely to regress negatively).
3) “Pitcher Score”. For lack of a better name, this is the new combined ranking metric with a 50/50 weight between PD Score and xSLG. If you have a better name, please let me know in the comments!
A few immediate observations:
1) Jacob DeGrom has been killing it on the contact-management side, enough for him to supplant Scherzer & Sale for the overall top spot in the ranking. He’s been super-elite at both components, ranking 3rd in PD score and 1st in xSLG. He’s just pitching incredibly well this year.
2) How about that Astros rotation? Four pitchers in the top 10 is pretty nuts.
3) Some of my previous darlings, like Castillo and Bundy, have taken significant hits now that contact-management is included. They are both still top 50 pitchers, and they should be owned. But these new rankings should be a much better reflection of their overall value.
4) Likewise, some good pitchers who graded out very poorly in my previous rankings have jumped way up. Carlos Martinez and Johnny Cueto come to mind, who have been excellent contact managers in 2018 but are not striking out a lot of guys.
5) Chris Tillman is dead last, so I must be doing something right.
Nick Pivetta (SP, Philadelphia Phillies)
Something seems to have clicked for Pivetta, as demonstrated by the drop in his ERA from six last year down to just 3.2 so far this year. Strikeouts are up, walks are down, plus hard contact and home runs are down, so at a glance it seems like good news all around. But let’s see if we can break it down further. For starters, to break out the components of his Pitcher Score, basically PD metrics and xSLG have contributed equally to his ranking. He is currently above average, but not elite, in both areas.
On the plate discipline side, he’s posting solid improvements across the board from last year. All four of the metrics I track are up significantly, supporting both the rise in his strikeout rate and fall in walk rate relative to last year’s numbers. So that’s certainly great news and a big part of his success. He’s over-performing his K% slightly, but not so much to be considered a red flag.
When it comes to contact management though, things look pretty different. His xSLG figure is essentially unchanged from last year’s number. It looks like what happened here is that he was unlucky last year, and this year his luck has swung over to the positive side. His 2017 SLG-xSLG was +50, but this year’s it’s sitting at -102. So in total, from last year to this year he has reduced his real SLG by 150 points while allowing the same overall quality of batted balls.
Conclusion: His improvement in strikeout and walk rates is real, but the improved contact management is more a mirage based on good fortune. Overall he’s certainly improved over last year, but not to the extent his drop in ERA would suggest.
Trevor Cahill (SP, Oakland Ahtletics)
Cahill has stood out in the rankings so far early in the season. In last week’s update, he had the third highest PD score of all pitchers. Granted, he had less innings than most pitchers. But it’s hard to look at the 36 innings he’s compiled this year and not be encouraged. He was also very good for a short stretch last year before getting injured, so it’s not completely out of nowhere.
On the plate discipline side, Cahill continues to put up VERY solid metrics. His Contact% and SwStr% are both among the top 10 in MLB, so his strikeout stuff is very real. He’s under-performing a bit on his K rate, as the metrics point towards 30% rather than 25%.
Looking at his contact management, he appears to have benefited from some good luck this year with a SLG – xSLG of -84. His xSLG mark of .423 is still slightly above average (average = .460).
Conclusion: Like Pivetta, it seems to be a mixed bag of real and less-real. Their contact management skills rate similarly, But Cahill has demonstrated significantly better strikeout skills, explaining the gap of 50 pitchers between them in the rankings.
Walker Buehler (SP, Los Angeles Dodgers)
Buehler has been one of the biggest mismatches between my rankings and real performance so far this year. Last week he received a D grade in plate discipline…with an ERA of two point something. Now that StatCast data has been added, we can see why. He’s jumped over 100 spots as a result, due to his elite contact management thus far. By xSLG, he’s been the 3rd best pitcher in MLB, which is pretty impressive stuff for a rookie.
On the plate discipline side, he continues his pattern of completely smashing expectations for his K%. That over-performance of 11% is still 2nd largest in MLB behind his teammate Ross Stripling. But actually, he’s started closing the gap already – it was 14% just a week ago. His actual K% has dropped two percent in that week, and should continue to fall.
Buehler’s SLG – xSLG does indicate a little bit of good luck, but 30 points is not really enough to worry about. After all, he’s still been the 3rd best contact-manager so far even with the 30 points removed.
Conclusion: The new ranking system should help you feel a lot better about owning Buehler. The K% should continue to drop, but the elite contact management means he is still someone you want to own.
Anibal Sanchez (SP, Atlanta Braves)
Just a quick note here, yes the stats look good but it’s 11 innings. This is a 34-year-old Anibal Sanchez we’re talking about, and I’m not buying it. He just gave up 8 runs in a rehab start in AAA and I think that’s all you really need to know. Maybe he might be worth keeping an eye on for very, very deep leagues.
Miles Mikolas (SP, St. Louis Cardinals)
Mikolas is a guy I was expecting to jump way up in the rankings once quality of contact was included. After all, how else could a pitcher with just average strikeout ability put up such a good stat line? However, it just didn’t happen. By PD metrics, he is the #84 ranked pitcher, and by the combined metric he jumped all the way to…#79. In contrast to Buehler, who jumped over 100 spots, this should be concerning to Mikolas owners.
Mikolas is not a big strikeout guy; his K% is just around league average (20%). The PD metrics are perfectly in line with that rate, so no red flags there. He’s got an elite walk rate of just 2.5%. Obviously that’s a good thing, but with those low strikeout and walk rates, almost four out of every five of Mikolas’ plate appearances will end in a batted ball. So you could say that he depends on managing quality of contact even more than most pitchers.
And that brings me to the reason he didn’t jump up the rankings: Statcast says there’s a fair amount of luck involved on his batted balls. His xSLG value of .441 is just barely better than average, but his actual SLG is 100 points lower. With that in mind, we have a pitcher who looks just average in both strikeout ability and contact management, and the only area he is elite is preventing walks.
Conclusion: Mikolas has been depending on weak contact so far this year and it may not be fully sustainable. His profile leaves him even more vulnerable to fluctuations in BABIP than a typical pitcher, which is concerning given the large discrepancy in his xSLG vs actual results. Mikolas’ saving grace is an extremely low walk rate. This should help him limit the damage if and when it arrives, but I’d still consider selling here.
Stephen Strasburg (SP, Washington Nationals)
Strasburg, unlike the previous two guys, is known for his strikeout stuff and not as a contact manager. So it’s not a shock that he took a bit of a hit once the StatCast data was added. However I’m still surprised at his overall low score and ranking just the 64th best pitcher overall. His ERA of 3.36 is still very solid, but I think it’s fair to ask – what’s going on with Stras?
Even on the plate discipline side, he hasn’t been elite this year. He finished 2017 with the 11th best PD score of all qualified starters, but this year all his metrics have dipped slightly in the wrong direction, and is currently ranked just 43rd with a “B” grade. He’s still striking out 28% of batters though, so it’s hasn’t really affected him all that much.
On the xSLG side, Strasburg rates almost exactly average with his mark of .456. This is over a hundred points worse than he was last year. He’s been a bit lucky so far this year, with a -60 “SLG-xSLG” which has helped masked the issue. But the spike in his HR rate (career high) does not appear to be a fluke – he’s been getting hit hard.
Looking at Strasburg’s pitch usage, not much has really changed. However when you look at the individual pitch results, his 4-seamer really stands out as having problems this year. He’s allowing an wRC+ of 192 on that pitch, compared to 113 for his career and 102 last year. Considering he throws the 4-seamer over 40% of the time, that’s a real concern. Looking at the PitchFX movement numbers, these are also down so that could explain it. Hopefully he figures that out soon, otherwise he’s looking like just an above average pitcher, not the elite arm you probably paid for.
Conclusion: Strasburg has seemed to take a bit of downturn this year, in both areas. There seems to be a problem with his fastball, his key pitch, and I might consider selling for the right deal, while his ERA is still nice looking.
Hyun-Jin Ryu (SP, Los Angeles Dodgers) and Vince Velazquez (SP, Philadelphia Phillies)
I’m writing about these two completely unrelated pitchers together, because I noticed going down the list that they have extremely similar stat lines at this point in the season, when it comes to the underlying metrics. Take a look at this:
They are essentially the same pitcher…but Ryu’s ERA is just half of Velasquez (2.1 vs 4.2). The difference seems entirely due to batted ball luck. Ryu’s “SLG – xSLG” is -60. Meanwhile Velasquez’s number is +80. So despite the fact these two pitchers have allowed the same overall quality of batted balls, there is a swing of 140 points in real SLG between their results. Some of this is likely to be park-related, as Ryu does play in a more pitcher-friendly park. But I think it really serves to illustrate just how much of an impact this “power luck” can have on a pitcher.