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.
|
Post by patford on Oct 26, 2020 18:52:53 GMT -5
I claimed nothing of the sort. I said that he had 70 movement (factoring in velo) on his slider and 85 movement on his sinker.
There's not just a world of difference between a pitch that grades as a 70 and pitch movement that grades as a 70, there's the entire freaking universe. And I just explained that, too.
To call these strawman arguments would be an insult to actual strawmen. Both Dorothy and the Scarecrow are spinning in their graves.
This is verbatim from your post at 3:49 a.m. on Saturday: "And what we know is that he seems to have the velocity and movement on both his slider and sinker to project as 70 pitches (the sinker analysis should be up tomorrow), and of course he fared way better than that with them in these starts.
"Did anyone on the staff think that Houck could display a 70 slider and a 70 sinker over his first three MLB starts? It's a plain fact that he did, and I don't believe he was supposed to be able to do that ever, let alone immediately."I don't know how anyone can interpret that as anything other than a claim that he showed a 70 slider and a 70 sinker. And I therefore do not feel that I owe Dorothy and Scarecrow an apology. The qualifiers in that quote are, "he seems to" "project" "over his first three MLB starts."
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Oct 26, 2020 19:13:14 GMT -5
I claimed nothing of the sort. I said that he had 70 movement (factoring in velo) on his slider and 85 movement on his sinker.
There's not just a world of difference between a pitch that grades as a 70 and pitch movement that grades as a 70, there's the entire freaking universe. And I just explained that, too.
To call these strawman arguments would be an insult to actual strawmen. Both Dorothy and the Scarecrow are spinning in their graves.
This is verbatim from your post at 3:49 a.m. on Saturday: "And what we know is that he seems to have the velocity and movement on both his slider and sinker to project as 70 pitches (the sinker analysis should be up tomorrow), and of course he fared way better than that with them in these starts.
"Did anyone on the staff think that Houck could display a 70 slider and a 70 sinker over his first three MLB starts? It's a plain fact that he did, and I don't believe he was supposed to be able to do that ever, let alone immediately."I don't know how anyone can interpret that as anything other than a claim that he showed a 70 slider and a 70 sinker. And I therefore do not feel that I owe Dorothy and Scarecrow an apology. OK, I'm guilty there of not including "with good command" in that first sentence, because I took it as a given.
"And what we know is that he seems to have the velocity and movement on both his slider and sinker to be projectable as 70 pitches ..."
Again, I've been talking only about movement and velo, and everyone here knows that command is much more important.
Commanding a pitch well over three straight starts tells you a lot less about his command going forward than plus-plus movement over three starts tells you about movement going forward. It's absolutely possible that he was in a groove with his mechanics and that he is prone to losing those mechanics for many starts at a time.
Next, in these three starts he certainly did show a 70 slider and 70 sinker. He got results that exceeded that easily and he had the movement to establish that he could get 70 results without luck.
You can display a pitch of a given quality over any time frame. If that weren't true, then someone "flashing a plus slider" would make no sense. Saying someone can flash a plus-plus slider is not saying he has a plus-plus slider, and that is so obvious that you should feel that I'm talking to you as if you were a child (which is not my intent). And I think you also get that showing a plus-plus slider or sinker over 17 innings doesn't remotely mean you have those plus-plus pitches.
You already quoted my reason for pointing this out and why I believe it requires upping Houck's propect status. It's the second thing I bolded above.
There is a simple and beautiful analogy for what Houck has done here that is so good that I hope it puts this misunderstanding to rest permanently.
A hitting prospect has, according to the SP report, average raw power that might project as plus.
He gets called up to MLB in mid-September and hits three homers of 440'+ with exit velocities of 112 and up.
OK, the scouting report is wrong. He showed plus-plus game power in this trial.
This doesn't remotely mean that he's going to become one of the best hitters in baseball, right? Of course not. It just means that a key tool is better than we thought, a good deal better.
All I'm trying to do here is measure his pitch quality tools objectively.
|
|
|
Post by philsbosoxfan on Oct 26, 2020 19:24:07 GMT -5
I claimed nothing of the sort. I said that he had 70 movement (factoring in velo) on his slider and 85 movement on his sinker.
There's not just a world of difference between a pitch that grades as a 70 and pitch movement that grades as a 70, there's the entire freaking universe. And I just explained that, too.
To call these strawman arguments would be an insult to actual strawmen. Both Dorothy and the Scarecrow are spinning in their graves. This is verbatim from your post at 3:49 a.m. on Saturday: "And what we know is that he seems to have the velocity and movement on both his slider and sinker to project as 70 pitches (the sinker analysis should be up tomorrow), and of course he fared way better than that with them in these starts.
"Did anyone on the staff think that Houck could display a 70 slider and a 70 sinker over his first three MLB starts? It's a plain fact that he did, and I don't believe he was supposed to be able to do that ever, let alone immediately."I don't know how anyone can interpret that as anything other than a claim that he showed a 70 slider and a 70 sinker. And I therefore do not feel that I owe Dorothy and Scarecrow an apology. To me, he essentially said the equivalent scouting terminology as flashed a plus slider and sinker. That said, the 47% whif rate on the slider says it's more than flashed but that's SSS.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 17, 2020 2:36:23 GMT -5
(First of two parts, because I forgot to hit "Create Post" last night!)
So I just realized that Houck's pitch values are measured in runs per 100 pitches! That makes an adjustment for the quality of his opposition trivial.
So, adjusting every pitch he threw to every hitter, based on how well that hitter hit that pitch all year ... the average pitch he threw was to a hitter produced 0.66 runs above average per 100 pitches.
He threw 257 pitches, so that's 1.7 runs above average in 17 innings of work, which is 1.6 ER, which is 0.82 points of ERA.
So his opponent-adjusted xERA, and his pERA (my metric) are 3.01. That's a .259 xwOBA. And that does not credit him for his skill in inducing GDP's (the big weakness in wOBA and xwOBA for pitchers).
How good is that? Adjusting for virtual league:
2.52 Bauer
2.72 DeGrom 2.93 Lamet
2.97 Fried
3.01 Hauck
3.06 Bieber 3.07 Javier
3.09 Means 3.09 Sanchez, Sixto
3.11 Cole
3.12 Bundy
I decided to do that.
The guys he faced in GDP situations, over the course of their careers, hit into a GDP in .097 of their PA. MLB average in 2020 was .098, so there's no bias there at all.
Using xBA on all the balls in play in GDP situations, the 4 GDP's he induced become 3.2. The expected number of GDPs given the hitters were 1.26, so he had 1.95 extra GDPs.
The average change in Run Expectancy for a GDP versus a SO in 2019 (last year we have the data for) was .384 runs. So that's 0.75 runs saved, which in 17 IP is exactly 0.40 off your RA or ERA.
Now, the +0.66 R/100 pitches I calculated earlier for his opponents is based on the actual pitches he threw hitters, and he did pitch some guys backwards. But he got away with it! Ideally, you'd want to also know how good these hitters were in general, as a group. So let's do that.
The hitters he faced had a 120 wRC+. The MLB ERA was 4.44, so 20% extra of that is 0.89 of ERA, which is larger than the 0.82 I got by doing it pitch-by-pitch. So we'll stick with the latter.
So now he's got a 2.71 real, true, adjusted for everything ERA in his first 17 MLB innings, an ERA- of 61.
The two questions you have to ask yourself are this:
What are the odds that a guy with 4th starter upside (not projection, upside) can do that in his first 17 MLB innings? Remember, we have already made a huge adjustment for luck on balls in play.
And in those three starts, he had better shape and speed on his sinker than any MLB starting pitcher, and top 3% shape and speed for his slider ... is that more compatible with a 4th starter upside, or something much better?
Yes, it's just 17 innings. But the odds of a guy with a 4th starter upside doing both of these things are essentially nil.
We're going back to the basic principal that sample size is meaningless without looking at results, just like the opposite.
I still haven't posted the full FB analysis, BTW. I have to see how close I got to finishing the writeup.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 17, 2020 4:37:51 GMT -5
Where Houck ranks start-by-start plus overall on the two pitches. The slider model has been tweaked by adding vertical movement in addition to total movement. Slider. I don't have a good translation yet to turn this VMQ (Velocity-Movement Quality) into a scouting grade, because the distribution isn't normal. The number is runs saved per 100 pitches. Note that plenty of guys might be able to place near the top of the leader board when they're on their game. Houck clearly didn't have his good stuff against the Yankees and his Slider VMQ would still rank 10th out of 140 starters.
Name VMQ Corey Kluber 2.53 Houck 1 2.26 Houck 3 2.04 Jakob Junis 2.02 Houck all 1.87 Mike Clevinger 1.85 Max Fried 1.71 Collin McHugh 1.60 Sonny Gray 1.57 Yu Darvish 1.52 Walker Buehler 1.44 Garre. Richards 1.33 Houck 2 1.22 Marcus Stroman 1.17 And here's the sinker. The reduced effectiveness against the Braves was hard to notice, and that he didn't get hit harder suggests that he had better command of it.
Here we do see a nice normal distribution, as evidenced by Soroka's 81 grade (representing a fraction more than 3 standard deviations). Name VMQ Grade Houck 1 1.09 93 Houck all .89 86 Houck 2 .85 84 Mike Soroka .74 81 Sal Romano .67 79 Jose Urena .52 74 Vince Velasquez .52 74 Dallas Keuchel .51 73 Houck 3 .43 71 Jesse Hahn .43 71 It's impossible to tell from 3 starts whether he had 2 unusually good ones for each pitch, or 1 uncaharacteristic off one. But the off ones are still 65 (likely) and 70 grade VMQ.
You can see from these numbers that VMQ is much more important for a slider than a sinker. This matches what we know. Fastball command is the single most important pitching skill. In contrast, a good slider is such a nasty pitch that good command is almost just a bonus; you see guys chase pitches that missed their target by a foot.
|
|
|
Post by orion09 on Nov 17, 2020 5:06:02 GMT -5
Where Houck ranks start-by-start plus overall on the two pitches. The slider model has been tweaked by adding vertical movement in addition to total movement. Slider. I don't have a good translation yet to turn this VMQ (Velocity-Movement Quality) into a scouting grade, because the distribution isn't normal. The number is runs saved per 100 pitches. Note that plenty of guys might be able to place near the top of the leader board when they're on their game. Houck clearly didn't have his good stuff against the Yankees and his Slider VMQ would still rank 10th out of 140 starters.
Name VMQ Corey Kluber 2.53 Houck 1 2.26 Houck 3 2.04 Jakob Junis 2.02 Houck all 1.87 Mike Clevinger 1.85 Max Fried 1.71 Collin McHugh 1.60 Sonny Gray 1.57 Yu Darvish 1.52 Walker Buehler 1.44 Garre. Richards 1.33 Houck 2 1.22 Marcus Stroman 1.17 And here's the sinker. The reduced effectiveness against the Braves was hard to notice, and that he didn't get hit harder suggests that he had better command of it.
Here we do see a nice normal distribution, as evidenced by Soroka's 81 grade (representing a fraction more than 3 standard deviations). Name VMQ Grade Houck 1 1.09 93 Houck all .89 86 Houck 2 .85 84 Mike Soroka .74 81 Sal Romano .67 79 Jose Urena .52 74 Vince Velasquez .52 74 Dallas Keuchel .51 73 Houck 3 .43 71 Jesse Hahn .43 71 It's impossible to tell from 3 starts whether he had 2 unusually good ones for each pitch, or 1 uncaharacteristic off one. But the off ones are still 65 (likely) and 70 grade VMQ.
You can see from these numbers that VMQ is much more important for a slider than a sinker. This matches what we know. Fastball command is the single most important pitching skill. In contrast, a good slider is such a nasty pitch that good command is almost just a bonus; you see guys chase pitches that missed their target by a foot.
Nice analysis. Your data backs up what was obvious to those of us who watched Houck's starts: his raw stuff was exceptional. The question that begs answering is how common is it for a pitcher to flash plus-plus stuff over a 3-start period before regressing? Put another way, is there a subset of pitchers that display high-variance in their VMQ due to mechanical inconsistency?
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 17, 2020 12:50:16 GMT -5
Nice analysis. Your data backs up what was obvious to those of us who watched Houck's starts: his raw stuff was exceptional. The question that begs answering is how common is it for a pitcher to flash plus-plus stuff over a 3-start period before regressing? Put another way, is there a subset of pitchers that display high-variance in their VMQ due to mechanical inconsistency? Precisely the question. And just as Arby's has the meats, I have some data -- every pitch thrown by a Sox pitcher in 2008 that was tracked by pitch/fx. It'll go a long way to answering this question.
Since the slider VMQ is based on just velo, horizontal movement, and vertical movement, I can actually track it pitch-by-pitch. I can measure each pitcher's overall variance, within-outing and across-outing variance, variance over a given number of pitches thrown, and so on. I can do those for the three separate components plus net movement, as well as VMQ.
That this data includes Justin Masterson's debut is just gravy. I already looked at his slider VMQ for the season and it was nothing special.
Oh, an addition to the last post -- my best guess for what an 80 VMQ for a sinker does for the pitch's overall grade is that it turns a 50 into a 60. IOW, it gives you a plus pitch even if you had average command. If Houck actually has an 86 VMQ, he'd still need 76 command to have an 80 pitch. But 60 command would give him a 70, and that's the hope. I think the odds of the pitch being a 65 are pretty good. That's 55 command of an 80 VMQ.
(I know that people are cringing over the high BB rate, but 4 of his 9 BB went to hitters with a 177 wRC+ or better, and one of them, to Ozuna, came after he absolutely struck him out when he was squeezed egregiously.)
|
|
|
Post by philsbosoxfan on Nov 17, 2020 14:45:24 GMT -5
After the Braves game, Snitker referred to him as nasty and effectively wild. I assumed the effectively wild refered to his slider but maybe it doesn't ?
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 18, 2020 3:57:31 GMT -5
Slider. I don't have a good translation yet to turn this VMQ (Velocity-Movement Quality) into a scouting grade, because the distribution isn't normal. The number is runs saved per 100 pitches. Note that plenty of guys might be able to place near the top of the leader board when they're on their game. Houck clearly didn't have his good stuff against the Yankees and his Slider VMQ would still rank 10th out of 140 starters.
Name VMQ Corey Kluber 2.53 Houck 1 2.26 Houck 3 2.04 Jakob Junis 2.02 Houck all 1.87 Mike Clevinger 1.85 Max Fried 1.71 Collin McHugh 1.60 Sonny Gray 1.57 Yu Darvish 1.52 Walker Buehler 1.44 Garre. Richards 1.33 Houck 2 1.22 Marcus Stroman 1.17
An even better revision. This one fixed a problem with the velocity component.
Name VMQ Corey Kluber 2.61 Houck 1 2.17 Mike Clevinger 2.03 Houck 3 1.93 Jakob Junis 1.87 Houck all 1.79 Collin McHugh 1.63 Max Fried 1.49 Yu Darvish 1.43 Sonny Gray 1.39 Jhoulys Chacin 1.38 Walker Buehler 1.37 Houck 2 1.15 Marcus Stroman 1.14
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 18, 2020 4:58:05 GMT -5
I looked at every pitch location, and how they were called, according to Gamecast's graphics ....
Edit: which, however, as Chris' skepticism revealed, are scaled all wrong at ESPN and are kind of meaningless. Never again!
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 18, 2020 12:40:35 GMT -5
Stactcast reveals that Houck got no gift strikes but was squeezed four times. Of course, when you get squeezed matters a lot. Failing to call a strike 3 is big. Missing a pitch on a full count -- calling strike 3 ball 4 -- is even bigger. Both of those happened to Houck, and both while pitching to Marcel Ozuna, in the 1st inning on a 1-2 count, and in the 4th. Ozuna ended up singling and, of course, walking.
Against the Marlins, he was squeezed on first pitches to Brain Anderson in the 4th, and Corey Dickerson in the 5th.
The correct was to put a number on this is to use the run value of a generic bad call on that count.There's a nice set of tables online with values.
The four pitches totaled .64 runs. That doesn't include the fact that failing to get a K on Ozuna and needing to face him with a 2-2 count instead is more damaging than it would be for the average hitter. So it's a conservative figure. Which is 0.34 of ERA.
We started with 3.83 (by both Statcast and my even deeper metric), knocked off 0.82 for facing hitters who collectively had a 120 wRC+ (also a conservative choice, as I did it 2 different ways and picked the smaller figure), knocked off 0.40 for the ground balls he allowed in GDP situations, and now we have another 0.34 for getting squeezed.
Which gives us 2.27. Better than anyone in MLB pitched over a full season.
That's a 2.27 ERA adjusting for luck on balls in play, for defense, for umpiring, for ballpark, and for sequencing. It's based entirely on the run values of correctly called strikes and balls, plus the run values of hardness of contact with launch angle of balls put into play (done two different ways, which agreed), while assuming an average distribution of batted ball direction. Now, you have to ask yourself: what are the odds that a guy who projects to be a 4th starter, which is to say, post about a 4.70 ERA, can put up an actual 2.27 over his first 17 MLB innings? And while showing plus-plus movement on his two best pitches?
The actual here is important. We're used to seeing a guy who projects to be a back-of-rotation starter post a number like that in his first three starts, and the proper response is always "well, he was lucky on balls in play, he had great sequencing karma which in the long run is almost certainly luck, he faced weak hitters, he got some gift strikes ... it's not that hard for a below-average pitching prospect to put up numbers like that over there starts."
But what happened here is that our guy actually pitched 2.27 ERA ball and his BABIP luck and sequencing karma reduced that to 0.53. And the 2.27 is not mysterious, as he showed plus to plus-plus movement on his two best pitches. The movement is good enough to add 10 to 15 grade points for each if he can throw them with 50 command.
I'm guessing he's going to be bumped up to 7th or 8th (passing Ward, Groome, and maybe Song), but after Casas I find it very difficult to rank the next 6 guys. Given his stuff, he can lose a lot of command and still maintain a 4th starter floor (the need for an effective third pitch is, I believe, much less if you have a deadly slider). He has years to refine his command, work on the splitter and maybe add a cutter, and that gives him borderline ace upside (think Lester).
I would change his projection to 5, and since you don't do a 4.5 - 6.5 range of outcomes, that stays at 4-6, but the ranking should be based on an understanding that he's at the upper end of both numbers.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 18, 2020 16:26:10 GMT -5
Repeat of my breakdown of his walks ... which I had forgotten I did already!
|
|
|
Post by Chris Hatfield on Nov 18, 2020 17:43:10 GMT -5
I'm not trying to be a Debbie Downer, but the mental gymnastics in play here are baffling to me. Help me out here, Eric: In a sentence or two, what specifically are we trying to show here? He walked 9 guys. He needs to improve that. It's possible he got squeezed a bit. I have a hard time with the idea that most of the walks were him getting squeezed. The idea that in 17 innings the umpires cost Houck more than three runs fails to pass the sniff test the same way your calculations that showed Christian Vazquez was one of the most important players in baseball because of his pitch framing (or whatever your point was at that time) failed to. I post below his pitch charts (click to expand them). there are a lot of non-competitive balls in all three games. These are not the pitch charts of a pitcher with plus command or whatever it is you're trying to say he had. There are definitely blue dots in the zone here - indeed, even some egregious ones (look at that one middle-middle in the Marlins start! Good lord.). But I have a hard time with the concept that those misses cost him as much as you're trying to say they did.
|
|
|
Post by patford on Nov 18, 2020 17:52:58 GMT -5
I'm not trying to be a Debbie Downer, but the mental gymnastics in play here are baffling to me. Help me out here, Eric: In a sentence or two, what specifically are we trying to show here? He walked 9 guys. He needs to improve that. It's possible he got squeezed a bit. I have a hard time The idea that in 17 innings the umpires cost Houck more than three runs fails to pass the sniff test the same way your calculations that showed Christian Vazquez was one of the most important players in baseball because of his pitch framing (or whatever your point was at that time) failed to. I post below his pitch charts (click to expand them). there are a lot of non-competitive balls in all three games. These are not the pitch charts of a pitcher with plus command or whatever it is you're trying to say he had. There are definitely blue dots in the zone here - indeed, even some egregious ones (look at that one middle-middle in the Marlins start! Good lord.). But I have a hard time with the concept that those misses cost him as much as you're trying to say they did. View AttachmentView AttachmentView AttachmentFunny coincidence. I just looked at the forum and saw Tanner Houck was the most recently commented on post and that Chris had made that comment, and I instantly thought before opening the thread, "This will be Chris being Debbie Downer. " Now Chris is probably correct that Houck will probably not be anywhere close to what I am hoping for. I'll admit that I fully expected Houck to get shelled in each of his three starts and expect him to get shelled in his next. I am, after all a long time Red Sox fan (so long time that my father passed away without ever seeing the Sox win a WS). I just thought the coincidence was funny.
|
|
|
Post by patford on Nov 18, 2020 18:00:59 GMT -5
Maybe Eric, or someone, could take the time to see exactly where in a count bad calls happen. Because a bad call at 1-0 is entirely different from a bad call which would have ended an AB, or ended an inning. So if by chance an umpire was a fan or not a fan of the Sox and had some conscious or subconscious inclination to fudge it would be much more when there was a bad call than the total number of balls and strikes. I don't think it is possible to overstate how big one pitch can be.
|
|
|
Post by Chris Hatfield on Nov 18, 2020 18:20:59 GMT -5
I mean, Houck pitched really freaking well in the three games. It provided hope that he can stick as a starter.
This idea that 3 starts showed he's a future number 2 or 3 or whatever is nuts. It's a small, positive sample. I look forward to getting a larger one next year.
|
|
|
Post by philsbosoxfan on Nov 18, 2020 19:08:55 GMT -5
On a different note, something I didn't know: Tanner Houck has been an eager student at Cressey’s facility, soaking up everything Scherzer, Verlander and other veterans tell him. He made significant arm angle and delivery alterations in their alternate site this summer, made three impressive September starts and Statcast lists his sinker movement in the top 3 in the game. theathletic.com/2202331/2020/11/18/gammons-state-of-pitching-2021-red-sox-rays-yankees/ADD: I did know that Max Scherzer and Tanner Houck went to the same college and that He sent a message to Tanner before him major league debut. I didn't know they trained together.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 18, 2020 22:56:29 GMT -5
I'm not trying to be a Debbie Downer, but the mental gymnastics in play here are baffling to me. Help me out here, Eric: In a sentence or two, what specifically are we trying to show here? He walked 9 guys. He needs to improve that. It's possible he got squeezed a bit. I have a hard time with the idea that most of the walks were him getting squeezed. The idea that in 17 innings the umpires cost Houck more than three runs fails to pass the sniff test the same way your calculations that showed Christian Vazquez was one of the most important players in baseball because of his pitch framing (or whatever your point was at that time) failed to. I post below his pitch charts (click to expand them). there are a lot of non-competitive balls in all three games. These are not the pitch charts of a pitcher with plus command or whatever it is you're trying to say he had. There are definitely blue dots in the zone here - indeed, even some egregious ones (look at that one middle-middle in the Marlins start! Good lord.). But I have a hard time with the concept that those misses cost him as much as you're trying to say they did. View AttachmentView AttachmentView AttachmentIt's quite possible that there's a huge bug in Gamecast's graphic, which purports to show the precise same thing. I will look into it.
If that is the case, that's really f-ed up. People look at those pitch location charts for good reasons ... it's immensely easier to see where a specific pitch went with Gamecast then with Statcast.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 18, 2020 23:56:21 GMT -5
The GameCast graphics at ESPN, at least, are scaled all wrong ... and in fact there were just four squeezes and no gifts. But that still has a significant effect on his ERA, just not a silly one. I'm editing the posts to correct them.
I watched those games but didn't score them pitch-by-pitch. I couldn't recall him being squeezed terribly, but my memory during that time period was acting very funky (much better now thanks to a dietary supplement that actually works) so I just trusted GameCast.
|
|
redsox04071318champs
Veteran
Always hoping to make my handle even longer...
Posts: 15,684
Member is Online
|
Post by redsox04071318champs on Nov 19, 2020 0:31:47 GMT -5
I mean, Houck pitched really freaking well in the three games. It provided hope that he can stick as a starter. This idea that 3 starts showed he's a future number 2 or 3 or whatever is nuts. It's a small, positive sample. I look forward to getting a larger one next year. Thank you! I could only imagine the "research" being done to show that Billy Rohr was going to be a star after dominating the Yankees his first two starts in 1967 and barely missing out on a no-hitter. I guess the sky's the limit for Juan Pena. And of course Clay Buchholz has future Cy Youngs in his future after that dominating no-hitter against the Orioles in his second start. And based on his minor league record, you'd think that he'd break out for sure in 2008. I'm being a little facetious here of course. I get the main point. I remember Roger Clemens striking out 15 and walking none against the Royals in 1984, and that was kind of a confirmation that the absolute dominance he displayed in the minors was real, because who has a 15-0 K/BB ratio in a game? But in Houck's case, he has a spotty minor league track record - and yes I acknowledge that he spent a lot of time working on certain things that impacted his numbers, but still...it was promising. He had three starts where he went about 5 or 6 innings. I can't get overly worked up about 3, count 'em, 3 starts. What it does is make me really want to see more - and it gives me hope that maybe, just maybe he could be a mid-rotation starter. It's just as possible that he winds up in high leverage in the pen as teams stack their lineups against with lefties or perhaps if the hits fall in enough, the walks exacerbate the issue. I have trouble with the - he only walked guys in certain situations so let's not count it too much. I'm sure you can make that argument for most pitchers, so I don't see why you'd filter out those walks as if they really didn't matter. At some point, he's going to have to attack the great hitters, too. Look, I don't want to sound "negative" on Houck, because maybe, just maybe there IS something there beyond a possible starter, likely high leverage reliever profile. Maybe that stuff he shows will translate to him being much more. But my God, we're talking 3 kind of short starts. Let's see a lot more and those 3 starts MAKE me want to see more. For me, this is like projecting a hitter based on 50 ABs. I guess you can say Rudy Pemberton should be a great hitter based on those Sept 1996 stats.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 19, 2020 1:02:00 GMT -5
I mean, Houck pitched really freaking well in the three games. It provided hope that he can stick as a starter. This idea that 3 starts showed he's a future number 2 or 3 or whatever is nuts. It's a small, positive sample. I look forward to getting a larger one next year. Thank you! I could only imagine the "research" being done to show that Billy Rohr was going to be a star after dominating the Yankees his first two starts in 1967 and barely missing out on a no-hitter. I guess the sky's the limit for Juan Pena. And of course Clay Buchholz has future Cy Youngs in his future after that dominating no-hitter against the Orioles in his second start. And based on his minor league record, you'd think that he'd break out for sure in 2008. I'm being a little facetious here of course. I get the main point. I remember Roger Clemens striking out 15 and walking none against the Royals in 1984, and that was kind of a confirmation that the absolute dominance he displayed in the minors was real, because who has a 15-0 K/BB ratio in a game? But in Houck's case, he has a spotty minor league track record - and yes I acknowledge that he spent a lot of time working on certain things that impacted his numbers, but still...it was promising. He had three starts where he went about 5 or 6 innings. I can't get overly worked up about 3, count 'em, 3 starts. What it does is make me really want to see more - and it gives me hope that maybe, just maybe he could be a mid-rotation starter. It's just as possible that he winds up in high leverage in the pen as teams stack their lineups against with lefties or perhaps if the hits fall in enough, the walks exacerbate the issue. I have trouble with the - he only walked guys in certain situations so let's not count it too much. I'm sure you can make that argument for most pitchers, so I don't see why you'd filter out those walks as if they really didn't matter. At some point, he's going to have to attack the great hitters, too. Look, I don't want to sound "negative" on Houck, because maybe, just maybe there IS something there beyond a possible starter, likely high leverage reliever profile. Maybe that stuff he shows will translate to him being much more. But my God, we're talking 3 kind of short starts. Let's see a lot more and those 3 starts MAKE me want to see more. For me, this is like projecting a hitter based on 50 ABs. I guess you can say Rudy Pemberton should be a great hitter based on those Sept 1996 stats. Ordinarily I'd agree with every word of this.
The thing you don't mention is that he had better movement / velo on his sinker in these three games than any active starting pitcher in baseball, and better movement / velo on his slider than 97% of the guys who throw one.
That was not in his scouting report.
Here's a thought experiment for everyone to try: imagine that his report here, based on an update from the alternate-universe 2020 AAA season, said he had gone from flashing a plus-plus sinker and plus-plus slider to being able to throw them for several games at a time.
Where would you have ranked and projected him? How much salt would you be taking his MLB performance with, if what he'd done with those two pitches had merely matched your reasonable hopes?
Because that is now the correct scouting report.
My enthusiasm is almost entirely driven by the pitch shapes. The results are secondary.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,941
|
Post by ericmvan on Nov 19, 2020 6:13:36 GMT -5
Does he have an off-the-scale sinker? I'm going with "no" for that one. This is the heart of our disagreement. Chris, what's your rationale for saying that? Looking at the movement?
My rationale, of course, is analytic. And I never posted the sinker analysis. It's way cool and then some.
---
What makes a sinker effective, for guys who throw both it and the 4-seamer? I now have a full model, and it's nifty. [1]
There are three inter-related factors: velo of course, the difference in movement from the 4-seamer, and the minimization of rise, which is to say, maximizing the perceived sink.
Let's start with the latter. The average sinker of the 141 starting pitchers who throw both fastballs has 6.1" of rise, +/- 1.5. (The average 4-seamer has 8.8" +/- 1.0.)
Clayton Richard has the most raw sink of these 141 pitchers, at 1.9". That's 2.8 SD, or a 78 grade for the sink component. The guy with the worst sink, Nathan Karns, grades out as a 29. Guys with worse don’t throw it. So the performance here—for starters who throw both pitchers—seems to be close to a normal distribution.
(Note that there are relievers who get more sink because of funky deliveries, like Darren O’Day.)
Tanner Houck had 0.0." Yes, that's 4.0 standard deviations better than average. You can say that Tanner Houck has 90 raw "sink" on the 20 to 80 scale (relative to this group, an important caveat) and you would not be talking nonsense.
The second major factor is the one I discovered and noted earlier—it's the difference in movement between the sinker and 4-seamer. If you throw both pitches in an average fashion, how far apart do they end up when they reach home plate? That's what matters.
One cool thing I just discovered is that the worst possible value to have is the average value of 4.6." The same thing is true of slider movement, and of changeup velocity differential. Furthermore, that worst value increases slightly as velocity rises, which probably reflects the fact that harder throwers tend to have more movement.
Let's look separately at the vertical and horizontal components of his differential. (Why? Wackiness ensues.)
The average pitcher has 2.9" +/- 0.9 of difference vertically. Houck has 3.6", exactly a standard deviation better than average and hence a 60 grade, and it puts him in the 84th percentile.
The average sinker has 8.3" +/- 1.1 of armside run. Houck has 8.9", which gives him a 56 grade and puts him in the 71st percentile. Given his arm slot, that's less than expected. It seems solid but nothing special.
The average 4-seamer (remember, it's the difference between these two that counts) has 4.6" +/- 1.8 of armside run. Now, Houck had an extra 0.6" on the sinker, so you'd expect 5.2" here, right?
He had 3.1." He is in the 22nd percentile for armside run on his 4-seamer, with a 42 grade. How is that even possible given his low arm slot? He has 2.1” less armside run than he should ... which increases the contrast with the sinker.
It seems as if his release is designed to minimize the natural armside run generated by the arm angle. You could do that by cutting the ball very slightly, which I don’t think is easy. Cut it more than slightly, and you lose velocity.
And they taught him this fastball. We're about see why.
By cutting the 4S a little—straightening it out intentionally—he gets 5.8" of difference in armside run. The average pitcher gets 3.6" +/- 1.0. He's got a 72 grade and is in the 97th - 98th percentile here; only Mike Soroka, Brad Keller, and Dallas Kuechel have more.
For overall difference in movement between the two pitches, the average is 4.6" +/- 1.0. Houck gets 6.9", which is third to Soroka and Keuchel (each 7.0), and is a 73 grade.
The velocity is actually a bit below average (91.5 versus 92.3 +/- 2.2), 46 grade, but he didn't have his best velo against the Yankees. He was absolutely average in his other two starts.
Now, put these three things together in the model, and what pops out is 3.6 SD above average. The best figure in the dataset is Mike Soroka, who is 3.0. That gives Soroka 80 sinker movement (in concert with velo), and Houck, so far, 86. Without the extra cut on the FB? It’s a 64. And you can't see the difference when you look at the sinker.
Now, the joker in this deck is that he grades out as a 41 in my 4-seamer model. It’s insanely straight. Is that the price he pays to increase the contrast with the sinker?
Surprisingly, no. Without the extra cut, it's a 38! The bad grade is from the innate lack of rise, which goes hand-in-hand (although only one is actually involved) with the extreme sink of the sinker. The contrast with the sinker goes a long way to nullifying the straightness; the 4S is playing off the sinker, and the contrast helps both pitches. The cutting action, which makes it unexpectedly even straighter, thus actually works a bit to his advantage.
What seems hard to explain at first is that he threw the 4S more often than he did the sinker. But that makes sense, I think. Both pitches set each other up, but it's the extra cut that turns the sinker into a beast. So you train the hitter's brain on the 4S, and they expect a given amount of armside run on the sinker and get two extra inches, which apparently they can't handle at all. If you do it the other way around, they get used to the sinker, and the extra unexpected cut on the 4S is apparently much easier to handle. This dichotomy is not unique to Houck; it's built into the models for the two pitches. And it makes perfect sense that unexpected extra movement running away from home plate is tougher to deal with than unexpected movement running towards the center.
Now, he got plus results with this 41 grade 4S, against seriously good hitters!! Am I going to look into that? No, I am not. There’s some deep game theory going on that I plan to just admire rather than understand.
I will guess that his major weakness going forward will be giving up the long ball on the 4-seamer, which he does need to throw with some regularity to make the sinker deadly. Commanding it will be important, but throwing it when they’re not looking for it will be even more so. Can you say "tunneling and sequencing"? Sure you can. Remember that they move so differently as to function as two different pitches.
So Tanner Houck seems for all the world to be better than you can tell by just watching him. His pitch repertoire has been designed to interact with itself to create an invisible but real advantage.
[1] The data set here begins with the 213 pitchers with 200 IP as a starter from 2015 to 2019. It’s the 141 of those who throw each type of fastball at least once per average inning.
|
|
|
Post by Oregon Norm on Nov 19, 2020 18:11:38 GMT -5
I think the likelihood is that Houck has positioned himself in a starting role. That said, the walk rate informs where that might go off the rails. He'll eventually build up a (video) data record that batters can scrutinize for giveaways on his pitches. Many of his pitches dart out of the zone just as they get to the plate. If players can figure out how to identify those, they'll be able to convert PAs to even more walks and he'll be in trouble. Now, there are pitchers such as Kluber who are capable of embedding all their pitches in the same delivery. He appears to be developing that sort of "tunneling", but it's very early in the reformation. So I'm with Hatfield and the others who've insisted on the need for a bit more evidence before he heads off to Cooperstown.
There's a bit to like. He works hard, he takes to suggestions about mechanical adjustments, and he can implement them. That's the sort of focus that can take him a long way. But he's not there yet.
|
|
|
Post by Chris Hatfield on Nov 19, 2020 19:17:29 GMT -5
Why are you using a data set of pitchers with at least 200 IP to evaluate a pitcher who threw 17 innings? You're making a data set of only pitchers who have thrown a ton to evaluate a player who threw the sinker all of 65 times. I don't get that. Houck's sinker definitely moved a lot! Here's a chart showing he had the second-most drop on his sinker of anyone who threw at least 50 of them this year: baseballsavant.mlb.com/leaderboard/pitch-movement?year=2020&team=&min=50&pitch_type=SIFT&hand=&x=diff_x_hidden&z=diff_z_hiddenHere's a chart sorted by whiff percentage on sinkers in 2020. Houck's 16.7% whiff rate on sinkers ranked 169th out of 398 pitchers. Although I typically think of a sinker as a pitch-to-contact offering, if his sinker was this movement unicorn, presumably he'd have been inducing all kinds of swing-and-miss with it, and he wasn't. baseballsavant.mlb.com/leaderboard/pitch-arsenal-stats?type=pitcher&pitchType=SIFT&year=2020&position=undefined&team=&min=1&sort=11&sortDir=descHere's the chart sorted by Hard Hit %. Houck was 37th, and that's with a bunch of guys ahead of him who threw a handful. Much better. In other words - a very good standard sinker. baseballsavant.mlb.com/leaderboard/pitch-arsenal-stats?type=pitcher&pitchType=SIFT&year=2020&position=undefined&team=&min=1&sort=17&sortDir=ascSo in other words, he didn't give up hard hits, but he also wasn't missing bats. It's a sinker so that makes some sense. But it also wasn't this godly pitch. He wasn't missing bats with it. Stop with this "it's a 90 movement pitch" nonsense. Another issue: He wasn't throwing strikes. Here's a chart of his out-of-zone % baseballsavant.mlb.com/leaderboard/custom?year=2020&type=pitcher&filter=&sort=5&sortDir=desc&min=50&selections=z_swing_miss_percent,oz_swing_percent,oz_swing_miss_percent,out_zone_percent,&chart=false&x=z_swing_miss_percent&y=z_swing_miss_percent&r=no&chartType=beeswarm So he threw 58% of his pitches outside the strike zone, 33rd-highest out of 469 pitchers who faced 50 hitters. His OOZ swing percentage was just 26.6%, 259th-highest on the list. So it's not like he's getting a bunch of chases outside of the zone. I don't care how much a pitch moves, if you can't throw it for a strike, hitters are going to lay off of it. I don't know how to get strike percentages on pitch types - please advise if you do! - but presumably, we can apply his overall failure to throw strikes to one of the pitches he most relied on. Here's another chart, showing that his xERA was more than 3 runs higher than his ERA, making him the third-luckiest pitcher in baseball this season based on that metric: baseballsavant.mlb.com/leaderboard/expected_statistics?type=pitcher&year=2020&position=&team=&min=50&sort=15&sortDir=ascHere's another, showing he was the 9th-luckiest using xwOBA: baseballsavant.mlb.com/leaderboard/expected_statistics?type=pitcher&year=2020&position=&team=&min=50&sort=12&sortDir=ascYou're tying yourself up in knots over the limited data regarding movement on his sinker and ignoring the fact he's throwing too many balls and the outcomes on his sinker aren't particularly great. There was a lot of good and a lot of bad. But here's the thing - this idea that the entire scouting community missed that he had a scale-breaking sinker or something is nuts. Your problem is your assumption that your model is correct and relevant. Have you done anything to show that positive outcomes result from what your model pops out? Here's the thing - I think what you're looking at is useful in showing why he might have success with his pitch set. It's interesting to show why his sinker has been successful. But if his sinker isn't being thrown for strikes, if MLB hitters are able to identify when he's throwing it, they're going to start laying off and destroying the four-seam, which as you say, he'd need to throw with some regularity to make the sinker play up. I'm not trying to be glass half-empty, but if you need to throw a bad pitch a third of a time to make one of your other two pitches good, I don't see that as a good thing necessarily.
|
|
|
Post by philsbosoxfan on Nov 19, 2020 19:46:52 GMT -5
Put me in the corner that's a huge fan of tunneling. The quality of the pitches in the tunnel is likely to have an impact on the tunnel itself. I've been on the Houck bandwagon since the Pitching Ninja videos and Eric's work re-enforces that view.
|
|
|