SoxProspects News
|
|
|
|
Legal
Forum Ground Rules
The views expressed by the members of this Forum do not necessarily reflect the views of SoxProspects, LLC.
© 2003-2024 SoxProspects, LLC
|
|
|
|
|
Forum Home | Search | My Profile | Messages | Members | Help |
Welcome Guest. Please Login or Register.
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,925
|
Post by ericmvan on May 19, 2023 15:33:27 GMT -5
As promised ... the rankings are as of 2 days ago 
(I don't expect this to spawn a lot of discussion, but I wanted it to be easy to find in the future.)
His xwOBA the firs three innings: .214. That ranks 3rd among the 162 starters who have faced 40+ hitter, behind de Grom (.165) and Max Fried (.203). But Fried has faced just 57 hitters versus Houck's 88. If you regress Fried by adding 54% more action, is there any projection for him that doesn't add .011 xwOBA or more? I think Houck is really second here.
He leads everyone in wOBA, .167.
-----
From the 4th inning on he's .405. That ranks 127th out of 149 starters with 30+ PA. (He is tied for 32nd in most batters faced, so his hook may be slow.) He's 136th in wOBA, with .441.
-----
So what's causing this? There are two possible factors: lack of stamina and increasing hitter familiarity. Splits by pitch count give you the former. Splits by times around the order give you both combined. I think that some people miss that.
There's some reason to believe that the hitter-familiarity effect is real, and that the larger a pitcher's arsenal is, the less of a factor it is. Houck now has a large arsenal: two fastballs that are as different from one another as any pitcher in MLB has, a slider, splitter, and cutter.
So our starting point is to blame stamina. We have an eyeballs case that suggests that, the Jays game where he couldn't get the last out in the 5th, gave up the three-run bomb on pitch 24 of the inning, had a nice rest (23 pitches, all but the first 4 with runners on, and a pitching change) and was much better in the 6th. So let's see if there's data to supports this.
Alas, there's no easy way to get either times-around-order or pitch count splits in Statcast. Furthermore, the b-Ref pitch count splits are warped by all the relief appearances, where all the best guys rarely go past 25 pitches.
However, we do have the pitches 51 to 75 OPS+ for MLB, and those are almost all by starters. MLB average is a .768 OPS allowed ... Houck is 1.112. This is an astounding decline from his 1-25 numbers.
.113 / .200 / .156 (1-25) .205 / .314 / .227 (26-50). At this point Houck is on average 1 out into the 4th.
.423 / .429 / .692 (51-75)
Note that on average, the first line is against 1 through 6 hitters while the second line is against 7 through 9 and 1 through 3. So the decline -- from a 2 OPS+ to 56 -- is despite facing weaker hitters (7 through 9 instead of 4 through 6). It looks like he's already starting to tire.
Now, here's a tiny sample size that will still be helpful: in 6 of his 8 starts he has faced, in the 5th inning, the 8 and 9 hitters a second time, and then the leadoff hitter a third time. The #1 hitters are much better than the 8 and 9, so we already expect them to hit Houck harder. To make us consider a familiarity effect, you'd need a honking big difference.
In fact, the 8 and 9 hitters have a .515 xwOBA while the subsequent leadoff guys managed just .315. That sure looks like he bears down extra against the leadoff guy ... using energy he saved facing the apparent easy outs. This may well be unconscious -- increasing fatigue, then a shot of adrenaline.
So, yeah, this is a stamina problem.
I'm pretty sure that there's a way to improve it via off-season training, because Josh Beckett seems to have done it after every off-year and skipped it after every good one. ERA's of 5.01, 4.03, and 5.78 in his even-numbered years withe the Sox, and 3.27, 3.86, and 2.89 in the odd ones. There was no difference at all in pitches 1 to 25 and then in the bad years there's a big decline as he goes deeper, while in the good years the splits stay much flatter.
Is there a way to build this stamina during the season? I'm skeptical. I don't think they could be unaware of this, so we can look for improvement if they keep him in the rotation.
It should also be possible, with all the data they have, to build a model that tells you not when he's toast, but when the toaster-oven has just been turned on. It's worth noting that the infamous pull of Blake Snell in the 6th inning of Game 6 of the 2020 WS was data-driven, and someone at Fangraphs found a bunch of danger signs in just the public data. I find the whole "yeah, but they were so relieved to see him out of there" criticism to be inane ... as if they wouldn't have been equally as relieved if Snell started missing over the middle, and with reduced velo to boot.
If they could add an inning of stamina, and used a data model to pull him before he gets hit hard ... that might make him more valuable as a starter than as a reliever.
In any case, there's no reason why he shouldn't add a good deal of stamina this winter, which will make him a CY contender. Seriously.
|
|
|
|
Post by jimmydugan on May 19, 2023 16:04:22 GMT -5
|
|
manfred
Veteran
Posts: 11,397 
Member is Online
|
Post by manfred on May 19, 2023 16:14:46 GMT -5
Honestly, if it is stamina, there is something wrong with this guy. He came to camp as a starter, so presumably his whole off-season was geared towards it. No starter… not even in high school… should be tired at pitch 50. If he is, he either has underlying physiological issues, far too taxing a delivery (or he is so maxed out that his stuff requires short outings), or is so amped up he’s blowing energy unnecessarily. There is no reason to be tired that fast.
|
|
shagworthy
Veteran
My neckbeard game is on point.
Posts: 1,492
|
Post by shagworthy on May 19, 2023 16:20:42 GMT -5
Honestly, if it is stamina, there is something wrong with this guy. He came to camp as a starter, so presumably his whole off-season was geared towards it. No starter… not even in high school… should be tired at pitch 50. If he is, he either has underlying physiological issues, far too taxing a delivery (or he is so maxed out that his stuff requires short outings), or is so amped up he’s blowing energy unnecessarily. There is no reason to be tired that fast. I don't think it's fatigue honestly. I think it's focus and he needs ritalin. It's not like his stuff suddenly changes as the game progresses, he just doesn't seem to place it well or sequence it correctly. Look at the games Cora left him in after the bad inning, he suddenly turns back into the guy the first time through the order. Also, just want to point out that game one hasn't even been played and there are already 4 pages of comments. We all need to find alternative hobbies. |
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,925
|
Post by ericmvan on May 20, 2023 0:11:43 GMT -5
Honestly, if it is stamina, there is something wrong with this guy. He came to camp as a starter, so presumably his whole off-season was geared towards it. No starter… not even in high school… should be tired at pitch 50. If he is, he either has underlying physiological issues, far too taxing a delivery (or he is so maxed out that his stuff requires short outings), or is so amped up he’s blowing energy unnecessarily. There is no reason to be tired that fast. It's true that he hardly needs to be the second best pitcher in MLB (for three innings), so it may well be true that he needs to pace himself more. But most SP's go more or less all all out until they run out of steam. If we see him beginning to have really good opening innings (instead of insanely great) while fading in the 5th and 6th to a normal degree, we'll know he's made a pacing adjustment -- and was one of the rare guys who needed to make one.
As far as offseason work to get into better shape, it's pretty much a universal human truth that when people are told "it's going to take a lot of hard work to achieve your dream" they later report that they had no idea just how much work it would actually take. It's credible to me that the team (and Houck) are just now recognizing the disparity between how much the F.O. wanted him to work on stamina and how much Houck actually did.
Houck learned a splitter in the off-season and it's become his #3 pitch (after the slider and sinker) and far and away his most effective; it's at 92.1, just 2 mph slower than his four-seamer. Again, it's credible to me that this was his primary focus this winter.
|
|
|
|
Post by GyIantosca on May 20, 2023 0:23:03 GMT -5
Some guys are made to start and some guys are made for the pen. All I know is he was one half of an unbeatable team in 2019. Of course Whitlock was the other. They were our nasty boys.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,925
|
Post by ericmvan on May 20, 2023 1:49:38 GMT -5
And comes down heavily on the side of stamina, with graphs showing the decay in pitch quality.
I can tell that he wrote the story as he did the research (I do the same thing usually) and when he discovered that it was likely stamina-driven, he of course didn't label the already-written data based on familiarity as probably not relevant.
|
|
patford
Veteran
Posts: 2,402
Member is Online
|
Post by patford on May 20, 2023 9:31:13 GMT -5
And comes down heavily on the side of stamina, with graphs showing the decay in pitch quality.
I can tell that he wrote the story as he did the research (I do the same thing usually) and when he discovered that it was likely stamina-driven, he of course didn't label the already-written data based on familiarity as probably not relevant.
Hopefully it is a stamina issue as that would be the most correctable problem. The "he needs another pitch" theory never made a lot of sense to me with the "starters pitching like relievers" trend of success in recent years. Even before adding the split finger and the cutter Houck had a three pitch mix as his two fastballs were radically different. Houck also does not seem to get any benefit from his split finger. He adds it second time through but evidence suggests it's not doing anything for him.
|
|
|
|
Post by pappyman99 on May 20, 2023 10:04:50 GMT -5
For me the constant moving between the bullpen and rotation has never really let him figure it out.
All pitchers are worse split wise later in the game, but his are worse.
Sometimes I think he gets careless or loses focus because he will be cruising and then bam 3-4 runs
Can also be they need to clearly mix his pitches up differently, his sliders seem to find way more or the plate the longer he goes.
For me I’d let him stay in the rotation until the all star break (he does have a 4.09 FIP). Then back to the bullpen if it’s clear nothing will ever change (he is also only 26).
Sale
Bello
Paxton
Whitlock
Houck
That’s a good 1-5, and if we are worried about innings we can just switch Crawford and houck after the all start break to manage the innings for both
|
|
|
|
Post by umassgrad2005 on May 20, 2023 12:18:15 GMT -5
He was basically a two pitch pitcher, he's now throwing four pitches. His best two pitches per fangraphs are way down, yet his new pitch is now his best. I was always keep him in the pen guy, yet him adding a new pitch and doing well with it is a great sign. Going to take time to figure out using them, gain that muscle memory. I'd say it's a success so far, he just needs more time. He could be key for a 2nd half run!
|
|
|
|
Post by Guidas on May 20, 2023 12:45:01 GMT -5
I'm hopeful, but the same thing that's plagued him his whole career continues to skew his results: Left-handed hitting.
In fact, this year, even with a new pitch, he's .259 AVE/.323 OBP/.482 SLG with a .344 wOBA vs. lefties. Until he fixes that, he's going to be a 5/6 or a pen guy.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,925
|
Post by ericmvan on May 24, 2023 3:24:54 GMT -5
I'm hopeful, but the same thing that's plagued him his whole career continues to skew his results: Left-handed hitting. In fact, this year, even with a new pitch, he's .259 AVE/.323 OBP/.482 SLG with a .344 wOBA vs. lefties. Until he fixes that, he's going to be a 5/6 or a pen guy. Guiadas had the fact I missed: the platoon splits. So I looked them up (xwOBA).
Through inning 3 he had no split at all.
In 4 through 6 he was a bit worse against righties, about as expected, but he got killed by lefties ... until his last start.
After discovering that I did a 20-way breakdown -- R vs.L, by 1-3 innings vs. 4 through 6, for each of his 5 pitches.
What I found suggested that pitch selection against LHB was contributing to the problem. I though he needed to simplify things. That was close, but not quite right.
After the great last start I realized I had to look at LHB in innings 4-6 start-by-start.
And I found something really cool. In fact, I found, for innings 4 through 6, a correlation between pitch selection pattern and success vs. LHP in a given start that had p =.01 -- hugely significant.
More tomorrow.
|
|
|
|
Post by philsbosoxfan on May 24, 2023 8:17:53 GMT -5
Based on his last start, maybe the Sox beat you to the punch.
|
|
|
|
Post by philsbosoxfan on May 24, 2023 8:38:11 GMT -5
If this is in fact a pitch selection issue, to whatever extent pitch selection is from the analytics group, somebody should be looking for a new job or perhaps career.
|
|
patford
Veteran
Posts: 2,402
Member is Online
|
Post by patford on May 24, 2023 21:04:21 GMT -5
The article at fangraphs showed Houck introducing the split finger the second time through the order and it does not seem it's an effective pitch for him. I suppose the thinking is he needs to change things up second time through and he does but that wasn't working so maybe he's better off sticking with his three best pitches and the cutter.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,925
|
Post by ericmvan on May 28, 2023 13:17:53 GMT -5
If this is in fact a pitch selection issue, to whatever extent pitch selection is from the analytics group, somebody should be looking for a new job or perhaps career. I don't think it was an obvious one. Maybe even counter-intuitive, until you think it through. Short version here (as I remember the numbers, rather than trying to find them in a pile of notes)
Against righties Houck remains a two-pitch pitcher. More than 80% sinkers and sliders, and given the tremendous quality of both, why not? Innings 4 though 6 swap the two, with sliders outnumbering sinkers.
He's been a 3-pitch pitcher against lefties, but it took him a bunch of starts to figure out which 3. He settled on:
The cutter instead of the sinker, for a hard pitch with downward movement that will bore in on hitters
The splitter, for an offspeed pitch that will run away from hitters, like the slider does to righties
The slider, because it's his slider.
He's had two good starts in innings 4 through 6 and what they had in common was a relatively even distribution of the three main pitches. In his other starts, he was going with one pitch (which one, varying from start to start) much more predominately. The latter is the very common "identify your best pitch in this outing and challenge guys to hit it" method.
Mixing the pitchers up instead makes sense two ways. That hitters cannot guess is the obvious one. But one of the most fundamental truths of pitching is that the more you throw a pitch, the better you command it. And yes, the opposite is true, but the improvement of command from repetition is much stronger than you'd expect if guys were simply throwing their better-commanded pitches more often. *
Mixing the pitches more evenly may be a tool for preventing the mistake that kills you.
All of this is very tentative. I wouldn't commit to a word of it. But it's a nice working hypothesis.
(What I haven't yet done is break down pitch use vs. LHB in innings 1 to 3).
* IIRC, I discovered this by creating two different models to project changeup effectiveness from pitch movement and velocity relative to FB -- one including pitch frequency and one without. That was a dozen years ago, though!
|
|
|
|
Post by Chris Hatfield on May 29, 2023 8:27:23 GMT -5
I wonder about the cause/effect at play here though. It's it that he's choosing not to throw all 3 pitches to LHB or is it that he lacks feel for one or two that outing?
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,925
|
Post by ericmvan on May 29, 2023 11:59:46 GMT -5
I wonder about the cause/effect at play here though. It's it that he's choosing not to throw all 3 pitches to LHB or is it that he lacks feel for one or two that outing? Yeah, the confound here is whether he needs all three pitches to be working well in order to succeed past the 3rd inning, or whether he can get a better feel for them by throwing them more often, e.g., in innings 1 to 3 when the worst case result would be a solo homer. That's why I need to compare innings 1 to 3 vs. 4 to 6 pitch use, start by start, and I currently only have the latter. |
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,925
|
Post by ericmvan on Aug 2, 2023 14:35:57 GMT -5
As I said yesterday, I expect them to add a starting pitcher to give them 6 starters including Murphy-Bulk. Sale supplants Murphy as soon as he has a sharp rehab outing, and stretches out start by start. Sale, Bello, Paxton, Pivetta, Crawford, Acquisition. Then when Whitlock comes off the IL, he gets the 6 spot and the new guy may well be optioned.
Houck? As I've been saying all year, his problem is stamina, not times facing the hitter. He has a .504 OPS allowed the first time around the order, but a .454 allowed through the first 3 innings. In the third inning he has been better against the top of the order than he was the first time he faced them. He has an .812 allowed the second time facing, and a .901 the third time, but a .962 starting in the fourth inning.
As a three-inning reliever he'd make Winck look like a AAA guy. And you'd seriously think about just piggybacking him on Pivetta. Six innings of Pivetta all out and three of Houck, in some order -- if Pivetta doesn't turn back into a pumpkin, you could start game 1 of a playoff series with that combo.
You’re saying that because Houck has fared better the second time through an order than the first, his problem is stamina, not being a two pitch pitcher with a splitter in progress. How large are Houck’s time through the order samples? Are they large enough to confirm there isn’t noisy variance baked in? What is the confidence interval for his splits so far to be what we see going forward? Is it common or uncommon for any given pitcher to fare better the second time through the order than the first? Pitchers as a whole do worse; how many instances are there of any given pitcher doing better in the second time through - for a single season? For a career? Mitchel Lichtman, among many others, has written about the three times through the order penalty. In this article, he finds two pieces that tell me it’s at best premature to draw the conclusion you’re drawing about Houck: (1) pitcher deviations from league average TTO penalty have a .03 correlation and require a sample of 1650 IP (8-9+ full seasons as starter) to deduce a statistically meaningful deviation in skill here from the overall pitcher pool. (2) taking pitchers as a whole, there is no statistical correlation between pitches thrown and time through the order penalty, suggesting that it is in fact about hitters being exposed to the pitcher’s arm slot and arsenal, rather than some sort of conditioning or fatigue issue. It doesn’t mean that we aren’t sometimes identifying meaningful things in small samples before it’s “responsible” to conclude them, but along the same lines, you’re demanding us to accept a conclusion at face value that the limited data at hand may or may not support. You’re insisting a hypothesis is a factual conclusion. SOURCE: Mitchel Lichtman c/o Baseball ProspectusFurther, this finding strongly suggests that TTO penalty is all about exposure:Have you looked at P/PA for Houck on first PA for hitters? Perhaps some of what you’re seeing is that Houck is throwing very few pitches to hitters on the first PA on average, which in turn decreases the second PA penalty due to limited exposure. being a two pitch pitcher with a splitter in progress
Houck this year innings 1 through 3:
Pitch Use wOBA
Slider 35% .184
Sinker 28% .190
Cutter 16% .193
4-seam 11% .475
Split 10% .111
So, actually, four dominant pitches (admittedly, 2 for RHB and 3 for LHB -- but the latter is where he gets hammered after three innings).
I should look at the specific PA for second time faced but in 3rd inning. We know that on the aggregate, they are better, and that's opposite of what the times-around-order thesis would suggest. And yes, it's a tiny sample size, but your hypothesis is that that Houck's .454 to .962 OPS split is driven by TTO. You should not be able to find any instances where the number went lower, let alone a few. I'm fairly certain that a t-test of the data against a hypothesis of a split that big would fail big-time. We may see!
Here's the argument. We know for a fact that pitchers can reach a point where fatigue sets in, they lose command and make mistakes in the zone, and suffer a precipitous drop in performance.
It's also believed that getting up and down (and throwing your warm-up pitches) makes a small but real contribution to that fatigue. (I bet that's been studied, but I've never seen it.) So "hitting the wall" at the start of an inning is quite credible.
We also know for a fact that there is a real times-around-the-order effect. It's actually bigger than it seems because guys going well in a game ate allowed go go deeper. There was an excellent set of papers trying to correct for that selection bias on Baseball Prospectus, maybe a year ago. The thing is, no one has ever suggested that a pitcher might go from the best in baseball to one of the very worst because of a second or third look. We know how fatigue can do that. How could that happen for a second look at a pitcher with dominating stuff?
(And moving this argument to this thread reminded me of the FG article that reached the same conclusion.)
|
|
|
