It got lost among the strike calls and late-inning dramatics, but Porcello pitched another very solid game last night. He looked a little shaky early on (perhaps because he was having as much trouble figuring out Kulpa's strike zone as the batters were), but ended up giving up just 3 runs in 7 innings. That's a very respectable start. He's given the team a chance to win every time he's started this year.
It got lost among the strike calls and late-inning dramatics, but Porcello pitched another very solid game last night. He looked a little shaky early on (perhaps because he was having as much trouble figuring out Kulpa's strike zone as the batters were), but ended up giving up just 3 runs in 7 innings. That's a very respectable start. He's given the team a chance to win every time he's started this year.
Meanwhile, Wade Miley is being marginally serviceable for Seattle. How many people this past off season would have preferred Miley to Porcello to start for Boston this year?
This article from August Fagerstrom at Fangraphs about Porcello's pitch mix is very interesting.
Whatever the case, Porcello went back to the sinker being his primary pitch, and he hasn’t looked back since. But the four-seamer is still there. And the way he’s using it now is making it more effective than ever. The idea to employ a four-seam approach may not have gone as smoothly as originally planned, but it looks like it’s working itself out anyway.
We can start with the basics: through nine starts this season, Porcello has pitched as something like the best version of himself. He’s got what would be his lowest adjusted ERA for a season with peripherals that rival his career-bests, including a 23% strikeout rate that dwarfs anything he posted in Detroit. This is the guy I think a lot of people have wanted to see Porcello become: he’s still exceptionally stingy with walks, he still gets ground balls at an average-or-better rate, and now, he can sit a batter down with a punch-out when he needs it. Before, there were just too many balls in play — grounders or not — for sustained success to seem realistic.
It's interesting, but I'm not sure it's based on real numbers. As far as I can see, he's taking PITCHf/x categories and assuming they're real, which is always wrong; PITCHf/x is only mediocre when it comes to distinguishing even quite different pitch types, and is always terrible at distinguishing sinkers vs four-seam fastballs. Even for Porcello, who has fairly clear separation between those pitch types, PITCHf/x mixes them up all the time.
The numbers in that article don't even come close to BrooksBaseball, which re-categorizes pitches according to its own, much better, algorithm (e.g. the FG article says Porcello threw 876 and 972 four-seams in 2014 and 2015 respectively; Brooks says 761 and 732). Nor is it close to my own hipsterish, artisanal recategorization that I did earlier this year, which came up with 637 four-seams for 2015. If three different evaluators can come up with a range from 637 to 972 for a pitch type, I don't think there's any sense at all in trusting his claim that there's a difference between 29% and 46%.
If you're interested, here's one of the (many) charts I used to manually re-categorize Porcello's pitches. The chart on the left here (raw PITCHf/x) shows a lot of four-seams (red dots) in what is clearly the two-seam (blue dot) cluster. It even calls some things that are almost certainly sinkers, changeups (orange) and sliders (purple). One step in my recategorization is shown on the right.
Again, I'm not saying this recategorization is right, just that the raw numbers that the FanGraphs article uses are wrong. There's no true right answer, because the sinker and four-seam are very similar pitches that smear into each other. You could draw the lines somewhat differently, come up with somewhat different numbers (e.g. anywhere from 637 to 732), and still not be wrong. But 972 is certainly too high by 25-30%, and using that wrong number to calculate a precision of 29% and 46% is not very helpful.
(Edit: In this article at the .com I talk about Porcello's pitch use in 2015 and early 2016. With more charts.)
Brooksbaseball has Porcello using his fourseam 23.4% of the time in two-strike counts, up from 20.0% in zero strike and 17.6% in one strike counts. So that's a more moderate effect with that set of classifications than Fagerstrom offered, but still detectable. What do your classifications suggest about that hypothesis?
Brooksbaseball has Porcello using his fourseam 23.4% of the time in two-strike counts, up from 20.0% in zero strike and 17.6% in one strike counts. So that's a more moderate effect with that set of classifications than Fagerstrom offered, but still detectable. What do your classifications suggest about that hypothesis?
That's Porcello's lifetime pitch usage, not his most recent.
I quickly ran numbers on the recategorized pitches from 2015 and 2016, and there's actually a very striking effect -- much greater than mentioned in the FanGraphs article, and much greater in 2016 than in 2015.
... since I can't figure out how to format tables in the new system, here are the tables as images ...
In 2015, he threw about 19.4% four-seams with no strikes, and about 30.4% with 2 strikes.
In 2016, he throws about 3.6% four-seams with no strikes, and about 25.1% with 2 strikes.
I'm surprised to see such a big difference. Reassuringly, my numbers seem pretty similar to Brooks, if you set to just the 2015 or 2016 seasons.
Obviously not many of us watched last night, but he did notch a QS- which is something. Anyone who was able to catch this have any thoughts?
If Wright can continue pitching like he has and Porcello, EdRod and Price live up to expectations then our headbanging search for a "no. 5" isn't terrible and we can actually deal with the ups and downs of Kelly, Clay, whoever else without losing any prospects for a shitty rental.
Great stuff, Iayork. Based on that data, in 2015 and 2016 it appears Rick relies on his two-seamer early in the count and when he falls into hitter counts (likely to induce weaker contact, then uses his four-seamer as his put away pitch. I'd be interested to see how this compares to his time in Detroit. There could be a potentially related increase in his k/9 and FB%, which began during his last season in Detroit.
If Wright can continue pitching like he has and Porcello, EdRod and Price live up to expectations then our headbanging search for a "no. 5" isn't terrible and we can actually deal with the ups and downs of Kelly, Clay, whoever else without losing any prospects for a shitty rental.
Given we're about 3 weeks from the mid point of the season these are a lot of significant "ifs" aren't they? Wright has been mostly solid. A real bonus for sure, but it's rare that a knuckler has the consistency to be the best pitcher on your team which is what he's been so far. Price has been better, but he's still not the lock down ace that we hoped for. Porcello started the year as well as any of us hoped he might, but has sputtered lately. That will happen to anyone over the course of the season, but it does raise concerns given his first half of last season. And Rodriguez's progress is still unknown to this point. Honestly I'm more concerned for numbers 1-4 than the 5th slot. I think the rotation as a whole needs to be strengthened and would personally want to see DD find a solid 1-3 type and slot everyone else down to satisfy the 5th spot in the rotation.
Obviously not many of us watched last night, but he did notch a QS- which is something. Anyone who was able to catch this have any thoughts?
If Wright can continue pitching like he has and Porcello, EdRod and Price live up to expectations then our headbanging search for a "no. 5" isn't terrible and we can actually deal with the ups and downs of Kelly, Clay, whoever else without losing any prospects for a shitty rental.
I watched this morning via DVR and except for one awful inning Porcello was amazingly good. Great rhythm - nice action on all pitches. Best I've seen him minus that inning.
Have to take last night's start with a bit of salt, though. That was a very weak NL lineup that San Fran ran out there.
Their #3 hitter rocks a sub-.650 OPS, both pinch hitters they used were sub-.550 OPS guys, and I think Brandon Belt is the only guy who starts for the Sox when comparing lineup to lineup.
Agree that his stuff looked pretty decent, but he seemed to have a hard time putting guys away early. Was good to see him settle in later.
To be fair to Giants hitters, SF plays their home games in maybe the best pitcher's park in the league. Span has been 90th in wRC+, Crawford 95th and Panik 112th. You are right that Boston has better players at those positions though.
It's funny - Porcello right now has exactly the same FIP- in 2016 as he did in 2015, but this year he's gotten there by being steady rather than getting bombed in the first half and then recovering. I guess it's a more comfortable ride.
He gives this lineup a good chance to win most times out. A 4.04ERA (and the 4.26FIP), a 1.08WHIP, and 8K/9 seems about right to me, given his career. He's averaging over 6 innings per start. There have been and will be games where he wiggles out of trouble instead of giving up a 3 run jack, and we will really enjoy those games. There will also be games like last week in Baltimore, and we will not like those games. But he's pretty steady and predictable to me this season. He's gonna be better at Fenway, but keep you in the game wherever. And on this team, this year, with this offense, I'll take it. Hoping his HR rates for June go down, but otherwise, he is who most of us thought he was.
Hi! It looks like you are not logged in. SoSH has a bunch of content only available to users who have registered so we suggest creating one.
If you are seeing this because of a supposed issue using the site, all site outage notices will be posted to our official Twitter account at
SoSH on Twitter.
Thanks!
