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.
Recent Posts
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 30, 2024 0:37:12 GMT -5
this is the quality product mlb umpiring gives us in a tense game leading off the ninth inning Declines in Win Probability, key plays
Dalbec fans to end 7th: 14% (versus drawing a walk)
Story grounds out to end 8th: 21%% (ditto)
Umpire turns Casas leadoff walk to out in 9th: 22%
Crawford homers (versus making an out): 19%
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 28, 2024 18:54:22 GMT -5
Best: WS Champs. Worst (assuming ordinary injury rates): I spend all winter pointing out that the last-place Sox would have made the postseason in two or more other divisions. Nobody cares.
I've been a Sox fan since 1962 and an analytically informed one since 1971, when I read Percentage Baseball and pledged my soul to OBP and SA. And the current Sox rotation has a very strong chance at being the best I've ever seen.
(Yeah, that sounds crazy. So did my prediction for the '21 team, 92 wins and the 1st WC. I can't give that kind of prediction this year because I have not followed the rest of the division at all.)
Here's a rundown, a short, pointed version of the analysis I've been doing all winter.
My stat (as always) is xwOBA allowed, the best single stat in all of baseball: SO, BB, HB, exit velocity, and launch angle. Rankings are among the top 150 pitchers in PA.
12th. Rank of Bryan Bello of all starting pitchers from June 1 on ... if you ignore the three terrible outings he had pitching at day, and on 4 days rest to boot.
And you should. Not all outcomes on the playing field are predictive. Identifying and eliminating the ones that aren't gives you hugely better projections. (The success of that methodology was a decent chunk of why John Henry hired me personally.) In this case, we can expect Bello to find an effective between-starts regimen for that combination, just as he he found them last year for day games and for four days rest (to a statistically significant degree). Oh, and this rankings does include 4 meh to bad outings in the two problem situations, while he was in the process of figuring them out.)
8th. Rank of Nick Pivetta of all pitchers, beginning at the end of June when he introduced his sweeper and rebooted his cutter. He was actually worse as a reliever, so I included that.
2nd. Rank of Kutter Crawford of all starting pitchers on the season, when there wasn't a runner in scoring position with first base open. He was 10th worst when there was, and 10th best overall. I assume they will have him fix whatever he was doing wrong in that situation.
This is admittedly rather misleading given that they rarely let him go past five innings. But there's a lot of headroom between 2nd best in MLB and #2 starter, which seems like a reasonable expectation.
2nd. Rank of Tanner Houck among all starters, over innings 1 through 3, at the time he was injured. He was 25th worst from innings 4 on.
Even though there were numerous stories about his being unable to do any off-season work last winter, and how he had worked his ass off this winter to go deeper in games .... the entire planet seems to think he has a reliever's repertoire. Christopher Gasper in today's Globe: Houck needs a third pitch pitch to stick as a starter. It would have taken him maybe two minutes to get this pitch breakdown (same data set that has him at #2):
(MLB average, all pitches, .322)
.159 / .110. Splitter. (I'm including wOBA because that's what someone actually watching the games would see.) .226 / .183. Slider .226 / .193. Cutter. Also his #3 pitch by usage (slider was 1, sinker 2).
.272 / .202. Sinker
1st. Rank of Garret Whitock in innings 1 through 4 by a wide margin in his first three starts in 2022. He pretty clearly strained something when they send him out for a 5th inning in his 3rd start, and he hit 60 pitches for the first time in more or less forever. (Yes, he qualifies, ranking 150th in PA at that point in the season. He would have qualified more easily with IP, but Statcast doesn't have that.)
I'm going to post this now and add bit more about Whitlock later ... his decline in innings 1 through 4 in his remaining six starts, versus the first three, was easily significant statistically. I think he's fully healthy for the first time since them.
(I'm also going to double-check my negative rankings, which I did through memory!)
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 21, 2024 15:35:26 GMT -5
If they do give some time to Rafaela at 2B, it will also be interesting to see who gets the most CF time. Lot's of valid options there, just none as good as him. I know Cora had mentioned a platoon at 2B. Typically that’s a L/R thing, but depending on 1) how long Grissom is out, and 2) how comfortable Rafaela is both at 2B and switching between infield and outfield, there’s a case to be made for a de facto defensive platoon. Sox 2024 Starting Pitchers: Groundball to Fly-ball Ratio in 2023 Bello - 1.94 Houck - 1.92 Whitlock - 1.20 Pivetta - .87 Crawford - .71 No pitcher will be upset with having Rafaela’s defense in center. But the number of plays there will be different on days when it’s say Kutter Crawford out there (a heavy fly ball pitcher), as opposed to someone like Bello (where Rafaela might often get the ball in center merely after it’s squeaked by the second baseman) - and I don’t think it’s insane to speculate Rafaela would be the best non-Story defensive 2B on the roster. (With that said, we’re also talking about a 23 year old rookie here just trying to crack the team - asking him to be constantly moving around while he’s adapting to the big leagues is a big ask.) Ratios can get you in trouble. Crawford last year had a 98 where 100 = average number of balls hit into the air. Bello (starting in June), Pivetta (once he introduced his sweeper) and Houck were 85, 84, and 86 respectively. Whitlock was 101 in 2021-22. So there's likely 2 dead-average starters and a trio of GB types.
The numbers for grounders, 137 Bello, 126 Houck, 102 Whitlock, 79 Crawford, 77 Pivetta.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 13, 2024 20:51:38 GMT -5
I'll echo a couple of the previous critiques - I guess my concern is that I'm not sure how you reached the conclusion that stamina was the variable here. I think it's a reasonable hypothesis, but there seem to be other factors. Your control group, as it were, is the guys who continue to succeed from the fourth inning onward, meaning that you're kind of pre-selecting your sample of guys who are going to have both strong stamina and stuff. Like, Spencer Strider holds his stuff well, but his slider is also baseball's best pitch. If the theory was correct, then high-stamina players who are more average-ish would hold their results better into the games, but you've filtered that group out of the sample. Like, Spencer Strider holds his stuff well, but his slider is also baseball's best pitch.
As opposed to Houck?.
Looking at innings 1 to 3 for both, Houck when he got injured versus Strider's whole season ...
Houck trailed Strider in xwOBA just .226 to .202, and (for what it's worth, but it's what most folks are looking at), topped him in wOBA, .183 to .223. Houck had a bit more than twice the gloveside run (11.5" to 5.6") and 3.1" more of downward movement (15.8" to 12.7", relative to the average FB). Overall, 40% more movement. Strider had a bit more velo, 85.7 to 84.0 (but Houck has more velo relative to his FB, and that may be the more important factor).. Further comparison is tougher, given that they used very different pitch-use strategies, and pitch usage of course affects results; for instance, it's well established that the more you throw a pitch, the less effective it becomes (the recent continued increase in SO is a function of everyone throwing their FB less often). Strider threw his slider 27% of the time vs. LHB and 41% vs. RHB; in contrast, Houck threw his slider … 28% vs. LHD and 42% vs. RHB. But what’s surrounding this coincidence (?) is very different, especially vs. LHB (57% for Strider, 51% for Houck). Strider went 62-27-19, 4Seam / Slider / Change, while Houck was 29-28-18-16-9, Cutter/ Slider / Splitter / 4Seam / Sinker. Lefties facing Strider have to look out for the FB, and in fact they did well in innings 1-3, .346 xwOBA, .347 wOBA. I think that with Houck, more hitters are guessing slider, and that’s going to inflate his xwOBA at least a bit. Strider was a pure 2-pitch guy vs. RHP, 58-41, and had a huge split --.331 / .345 with the FB and .162 / .171 with the slider. Again, I don’t think a RHB can afford to guess slider when he has to be hunting for a hittable fastball, which we know existed. Houck was 49-42-5-2-2, Sinker / Slider / 4Seam / Cutter / Splitter, which is to say, Strider's approach but with one in six FD replaced with one of three other pitches. (And it worked: 5 PA, .238 / .178.)
Strider is using his FB to make the slider more effective, which is a very common strategy. Houck is kind of on his own with the pitch, with more guesses available for the hitters. I won't dispute the claim that Strider's slider is baseball's best pitch in context, but I think there’s a good argument that Houck’s slider is better than Strider’s in a vacuum. That 40% extra movement – which puts him with the best-breaking slides on the planet – stands out in a way that Strider can’t match. Just imagine what slider results Houck could get if he had a 97 mph FB.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 13, 2024 3:55:33 GMT -5
I'll echo a couple of the previous critiques - I guess my concern is that I'm not sure how you reached the conclusion that stamina was the variable here. I think it's a reasonable hypothesis, but there seem to be other factors. Your control group, as it were, is the guys who continue to succeed from the fourth inning onward, meaning that you're kind of pre-selecting your sample of guys who are going to have both strong stamina and stuff. Like, Spencer Strider holds his stuff well, but his slider is also baseball's best pitch. If the theory was correct, then high-stamina players who are more average-ish would hold their results better into the games, but you've filtered that group out of the sample. I guess my concern is that I'm not sure how you reached the conclusion that stamina was the variable here.
I'm trying and failing to come up with a snappy answer here, given that the fact that pitchers tire as they pitch, to the point that they need to be removed from the game, is about as well established as the fact that there are four "bases" rather than three or five.
We do not need to look for another factor. The only question here is whether the only other factor that has been proposed -- that smaller pitch repertoires also contribute to the quality decay -- is correct or not.
So, I did quite a bit more number-crunching.
First, I calculated the collective drop-off of all my 74 qualifying guys. It's .032 worth of xwOBA. I then normalized all the individual numbers to that. Merrill Kelley, for instance, was .001 worse starting in the 4th inning, but compared to the average pitcher, he's .030 good (the other .001 is a rounding fluke).
I then decided to add guys to the database, and figured everyone that was .030 or more, good or bad, was a nice compromise, since adding the data by hand is a pain. It turns out that in both the good and bad sets, there is nobody between .030 and .026, so that was an obvious place to draw a line. So I now have data for 19 bad guys and 18 good ones.
Finally, I weighted the contribution of each pitcher to the aggregate performance of their group. That is, Houck's (normalized) .101 means that he makes about twice the contribution to the aggregate performance of the bad group that Max Scherxer does, at .049.
So I can derive the average pitch profile of the good and bad groups. These are the percentage of the time the average guy in the group throws his fave pitch, his sec fave, and so on.
GOOD 45 26 14 09 04 01 BAD 40 27 16 10 06 02
So the bad gang throws their fave pitch 11% less often than the good gang, and distributes the difference more or less equally across the reamonder.
The repertoire-size hypothesis predicts the opposite.
Given that the pitchers included have an average weight of .050 and the pitches not yet included are .015 ... you're just never going to get there.
It does make sense that pitches with fatigue try to compensate by mixing their pitchers more.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 12, 2024 18:03:00 GMT -5
Maybe I am being dense, but I am missing what the connection is between pitch use and stamina. Is the idea that a tiring pitcher will go to their weaker offerings more often? Why? As for Houck, he looks about as good as any pitcher in the world in the first 3 innings or so. If he just had a regular gig of pitching 3 innings every four days he could rack up 120 IP, which is already 2/3rds of a starting pitcher's load. He could be super valuable in that role, and a real weapon in the playoffs. There's no connection. The standard take on the times-around-order affect is that pitchers start the game going to their best pitch or two and later in the game they need a deeper repertoire to succeed, as the hitters have acclimated themselves to the pitches they've already seen. I'm calling that nonsensical, in the age of video. Houck spent the 22-23 winter recovering from surgery and unable to keep up his stamina, let alone work on it.
As for Houck, he looks about as good as any pitcher in the world in the first 3 innings or so.
And that is how good his stuff is.
Look, imagine that the only thing you knew about Houck was the following:
1) Of the 150 starting pitchers who had the most PA in innings 1 to 3, he ranked 2nd in xwOBA allowed, with .244.
2) He had this pitch breakdown in these innings: Pitch Pitches PA xwOBA Slider 201 50 .226 Sinker 166 39 .272 Cutter 91 30 .226 4-Seam 62 16 .348 Split 59 16 .159 3) Among the same 150 guys he ranked 125th from inning 4 on, with .376.
How would you explain #3? That's not a reliever's pitch profile. I can't think of anything other than running out of the proverbial gas.
And if you formed that hypothesis and then looked for evidence, you'd say, nailed it! You'd discover that Houck missed the entire preceding winter recovering from surgery, that he felt certain that this was the cause of his problems, and that he had proudly worked his butt off this winter.
You may recall that when he came up I discovered that his slider movement was among the best in MLB and the sinker wasn't far behind. The cutter and splitter seem to be terrific pitches as well. I do wonder whether he should reduce the 4-seamer to an occasional keep-em-honest pitch, which is to say the new regime will either do that or (better, of course)) transform it into a solid pitch. It doesn't have a lot of movement.
BTW, if you saw his My Story, you know his makeup is impressive.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 12, 2024 15:50:59 GMT -5
Also with many pitchers trying to max out each pitch these days it’s not surprising they don’t have the stamina to sustain performance for longer outings Yes. It no longer makes sense to pace yourself given that quality and depth of bullpens.
What I have encountered again and again is tales of young pitchers learning that getting in shape to pitch 7 of even 6 innings at the MLB level takes a lot more work than they ever imagined.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 12, 2024 15:46:07 GMT -5
I don't doubt that this sort of thing happens, but the idea that hitters as a rule get a measurable edge from seeing a pitcher a second and third time -- as opposed to getting an edge because the pitcher is continually tiring -- just make sense no me.
There's a lot less evidence here than you think. But one thing that counters your lightly supported theory is that familiarity with relief pitchers is starting to be understood as an issue in the playoffs. And of course on a longer time scale we are all familiar with the relative success the Sox had with Mariano Rivera in the days of yore when it seems the Sox-Yanks were playing tense playoff style games every other week. Just because you want to "prove" a specific Sox starter is better than non Eric Van's think he is doesn't really mean you can toss all examples of familiarity effecting the pitcher-batter matchup. Career familiarity, at least in the era before ubiquitous video to study, was very real. This is hugely different from the within-game version.
I did these splits for the Sox. A career familiarity split was at the heart of my convincing Jed Hoyer to convince Tito to bench David Ortiz against Mike Maroth have Doug Mirabelli (!) DH instead (May 3, 2005). Maroth fit the profile of guys Mirabelli could hit, Mirabelli had awful familiarity splits (pitchers have or had the edge for the first 8 PA, career) but he had already gone yard against Maroth, IIRC the 4th time he saw him. Mirabelli hit a GS with the Sox trailing 2-1 and that was it for the scoring.
Hideki Okajima, with his straight over-the-top delivery, had a huge such split. Some of you may remember Barry Bonds facing him for the first time and taking strike three right down the middle! The only time I got into trouble with the Sox was when I told Zack Scott (who had replaced Jed as the speaker-of-consultants-to Francona, and in his first stab at the job wasn't very good at talking him into anything) that he could not pitch Okajima in a high-leverage situation in this series ... whereupon Tito did exactly that and we immediately lost, a very tough loss at that. I was so upset that I fired off a "why am I doing all this work if my advice is being ignored?" complaint and I heard from (IIRC) Jed to the tune of "this is not helping."
It would be very interesting to know if this effect is still in place now that we have video and even machines that can simulate a specific pitcher.
As I said, the concept that in every game all the hitters get an edge in PA 2 versus 1, and 3 versus 2, is completely different.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Mar 11, 2024 23:41:49 GMT -5
I know I spent most of the winter trying to convince folks that the ability to go deep into games was largely a function of stamina (versus fatigue) rather than depth of arsenal (pitch variety). I was wrong, and I apologize.
It's entirely about stamina. Let me first give you the reasons this makes sense, and then the facts that back it up.
First, in modern baseball players have film of their opposition that they study before a game. There's not a heck of a lot of room left for new discoveries or insights. You're basically looking for guys who couldn't sleep much at night after a huge arguing with their wife, or chose not to (for opposite reasons) with their girlfriend.
But the big thing that everyone overlooks is that the supposed acclimation to the opponent is a double sided coin. The pitchers are not only watching the hitters and looking for differences from the scouting report, they are also monitoring themselves.
I don't doubt that this sort of thing happens, but the idea that hitters as a rule get a measurable edge from seeing a pitcher a second and third time -- as opposed to getting an edge because the pitcher is continually tiring -- just make sense no me.
Some Facts.
The impetus for this study was of course Tanner Houck. Hocuk last year in his first 8 starts had had a 0.38 ERA in innings 1 through 3 and a 12.50 thereafter. In his last 5 starts before his injury this became 3.00 and 6.30, so be might have started pacing himself. Overall, 1.38 vs. 10.05. That was a .244 xwOBA versus .376.
So I did the following:
-- Looked at all the starting pitchers up to two days after Houck's injury (to grab further gus in his rotation)
-- Took the 150 who faced the most hitters -- Selected the 84 who were above average through innings 3 -- Removed 8 guys whose PA after inning three were less than 70% of those in the first 3 innings. (The remaining pitchers averaged 93%.)
-- Identified the 10 pitchers who fared best in innings 4+ relative to 1-3, and the 10 worst.
We can name the groups after their champions: Strider-Types vs. Houck-Types
I then put their pitch use percentages into the spreadsheet, in order from most favorite to least used. For instance, Zach Grienke was 26 20 17 16 14 7, while Joe Ryan was 57 27 11 5. (These are totals across all innings.)
So, how did the struggling Houck-Types pitch, compared to the killer Strider-Types?
The usage level of #2 and #3 pitches were all over the map. You might have expect that. But there were clear trends for favorite pitch use, and for total 4th to sixth pitches, total.
And this is what you see:
The Houck-Types threw their fave pitches 16% less than the Strider-Types (41% vs. 48%), and their 4 to 6 pitches 56% more (19% versus 12%).
(Yes, this is backwards from what people believe.)
Put another way: in this sample, there's little difference between these two types, except that once in 14 pitches the guys who flourish going deep in games go to their favorite pitch while the guys who struggle doing so use one of the extra ones.
Ideally you'd use the full season and grab everyone who was more than 0.5 standard deviations better or worse than average. Given that in this sample, the reverse case (#1 pitches) just misses being statistically significant (p = .08), there's no way the full study is going to indicate a real effect in the supposed direction.
----
My favorite Times Around Order Split:
.673 / .673 / .611 (OPS)
Pitches: 43% sinker, 42% 4-seamer, 15% slider.
Justin Masterson, 2011.
That winter I did a study of this question, and the only thing I could find was that ace pitchers almost always had a usable 4th pitch.
It's all stamina, folks. And given that Houck, Crawford, and Whitlock have all devoted themselves to improving theirs this year, that's good news.
`
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 29, 2024 19:14:25 GMT -5
He had 5 starts in September. The next to last was a day game on 4 days rest, a combo he had dealt with just twice, which means he hadn't the time to develop a between-starts routine as he had earlier done for the separate challenges of 4 days rest and day games. I'll run the numbers at the end and you'll agree that they were meaningless as a skill measure.
In is other 4 starts, he threw his change, sinker, and slider a total of 300 times resulting in 75 PA. (Actual numbers, not rounded!)
He allowed a .201 xwOBA and .242 wOBA. That's note a "fade," folks; it's a strong CY candidate's numbers.
Alas, he threw his four-seamer 89 times for 19 PA and had a .433 xwOBA and .696 wOBA.
Oops.
How did it fair before September? I'm excluding the 8 games where he struggled at day or on 4 days rest.
This is xwOBA, wOBA, starts, pitches, PA
.388 / .347, 4, 98, 21 May
.245 / .131, 5, 131, 31 June
.365 / .578, 6, 98. 22 July / August
Worst case is he junks the 4-seamer, but look at that June: 131 pitches and a .131 wOBA allowed.
-----
His June 20 start, xwOBA, wOBA, pitches, PA:
.768 / .801, 23, 8 sinker .453 / .615, 21, 4 slider .244 / .577, 20, 5 change .474 / .474, 18, 3 4-seam
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 27, 2024 23:30:59 GMT -5
Jim Callis: "My favorite part of the spring is watching top prospects try to grab a starting role in the big leagues for the first time. Here are five I'm monitoring closely" ... Rafaela is the 5th guy.
I think the decision here is very easy. If he makes such a large advance offensively that he would be upset to be sent down, he's your opening day CF. Otherwise, back to Woo. Getting a look at Abreu as an everyday player with the opposition having a scouting report on him doesn't suck.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 27, 2024 23:04:23 GMT -5
I use Fubo as my main source for TV. $80/mo which includes any channel I could think of including NESN I just junked the cable portion of my Comcast / Xfinity with the intent of switching to Fubo a few days before the Oscars. I believe they have the Sox in 4K, and I have the gear to use it. It'll save me $60 a month, too.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 27, 2024 22:33:24 GMT -5
He threw a splitter last year didn’t he? Might’ve been classified more as a changeup last year. I was reading an article where it was coming from himself that he’s trying to add a split He threw his splitter 163 times last year, 44 of them ending a PA, with a .273 xwOBA.
That ranked 87th among the 212 pitchers who had 40+ PA on their splitter or change -- 59th percentile. Great pitch in the big real of things, solid above-average pitch relative to other off-speed pitches.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 26, 2024 21:18:24 GMT -5
Final numbers! (Including the bulk outing.)
Kutter Crawford faced 63 hitters with runners on and 1B base open, and had a .416 xwOBA, which ranked as the 7th worst performance among the 150 SP who faced the most hitters.
(He also ranked 40th worst in his 8 PA with runners on the corners, with a .381 xwOBA. The average sample size was 14, but you would still expect some correlation with overall pitching ... and in fact there was hone at all (slightly negative, actually). That was the situation toughest to handle with the new SB rules, and the lack of correlation can be attributed to pitchers taking widely different approaches. In any case, it didn't affect his numbers tangibly.)
Without these two problem areas he had a .261 xwOBA (in 375 PA), which ranked second, and he still ranks second with the 8 runners on the corners included.
Honest question: Does anyone have an educated guess about much of this is noise and how much is actually attributable to a difference in approach when 1B is open? Like, after how many hitters faced would we expect real differences in runner-on scenarios to stabilize? Or, if we were to examine a veteran SP's 10-year career, would we see substantial (say, .100 or more) year-to-year variations in xwOBA for, e.g., 1B open or RISP. I'd love it to be real, obviously, because it would suggest that Crawford's biggest problem is simply one of approach, but I don't know how to think about what seems like a really small sample. Significance is a function both of sample size and effect size.
If you’re otherwise the second best starting pitcher in baseball facing 383 hitters but the seventh worst (of 150) facing 63 hitters in a situation where we know pitches often alter their approach – that’s a mind-boggling effect size, together with a likely explanation.
I’d have to have the xwOBA of every plate appearance of both types to put a number on the odds, and Statcast doesn't have that data handy. (I can work around it and might do so later).
The totals PA, by the way, were enough to not need an adjustment for small sample size.
Year to year pitching performances differ for real. A single season has some chance of being one data point ... but last last both Pivetta and and Bello became different pitchers after the first two months.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 26, 2024 0:48:45 GMT -5
Final numbers! (Including the bulk outing.)
Kutter Crawford faced 63 hitters with runners on and 1B base open, and had a .416 xwOBA, which ranked as the 7th worst performance among the 150 SP who faced the most hitters.
(He also ranked 40th worst in his 8 PA with runners on the corners, with a .381 xwOBA. The average sample size was 14, but you would still expect some correlation with overall pitching ... and in fact there was hone at all (slightly negative, actually). That was the situation toughest to handle with the new SB rules, and the lack of correlation can be attributed to pitchers taking widely different approaches. In any case, it didn't affect his numbers tangibly.)
Without these two problem areas he had a .261 xwOBA (in 375 PA), which ranked second, and he still ranks second with the 8 runners on the corners included.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 23, 2024 4:49:04 GMT -5
OK, this is a very big step forward.
Crawford actually didn't have a RISP problem. He had a first base open problem.
Recall that he had a .262 xwOBA allowed without RISP, including his tremendous bulk game.
With runners on 1st and 2nd, 1st and 3rd, or the bases full, he had a .258 -- without including the bulk game.
He had 73 PA with a man on 2B, 3B, or both, and gave up a .418 xwOBA.
And of course pitchers change their approach with men on but first base open ... especially with good hitters up.
I also found decently strong suggestion that stamina was a factor, including stamina within innings, but that's a small factor in comparison.
He's worked all winter to add stamina. All he needs is for someone to tell him to ignore the fact that first base is open.
What I think was happening ... OK, no need to come right at him, let's see if I can get him to chase outside the zone ... damn, that didn't work, but I don't want to walk him.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 17, 2024 18:31:57 GMT -5
A bit more, with updated numbers from Statcast ... I checked the top 30 starting pitchers by xwOBA (of the 150 that had the most PA) for bulk starts ... Crawford was the only one who even pitched in relief, and his 7-inning bulk outing was the only one by any of the 30.
Crawford ranked 10th in xwOBA, with a .285. Here's some new numbers, all including the bulk outing ... With the bases empty he faced 282 hitters and had (again) a .285 xwOBA, which ranked 18th. With a runner just on first, he faced 87 hitters and had a .184 xwOBA. The leaderboard: .182 Taril Skubal (.008 added for regression to mean in small sample size, 62)
.184 Crawford .229 Zack Wheeler (142 PA) .236 Cristopher Sanchez (.004 added for 75 PA; see the geekabe note at the end) .251 Bobby Miller (87 PA)
After that the rankings are closely clustered. What Skubal and Crawford did was insane. And Skubol led MLB in xwOBA with .246 (.235 before regression). Here's the top 5 pitchers for non-RISP PA : Name Non-RISP RISP Ratio Tarik Skubold .226 .348 1.54 (adjusted for sample size) Kutter Crawford .262 .365 1.39 Zack Elfin .264 .301 1.14 Zack Wheeler .264 .340 1.29 Freddy Peralta .268 .357 1.33
Wow. This is somewhat of a thing. The MLB average for RISP is .329. While Crawford was hit hardest, Peralta was not far behind. Eflin was the only guy who was better than average.
All of these made the top 15 xwOBA (= ace-level performance) out of the 150 that faced the most hitters. I did the other 10 xwOBA aces and the worst case was Max Scherzer, a .333 with RISP and a ratio 1.22.
Some thoughts:
If this were just one or two guys, and the effect size smaller, you'd say "luck." But 4 guys whose median performance translates to 2nd best in MLB with no RISP and 28th worst with RISO--super ace vs. average 5th starer? Not happening.
Splits by bases empty / runners on are common and can be ascribed to pitching better or worse out of the stretch. Those splits should be be looked at.
I still suspect that this is at least partly a psychological thing. If a pitcher has RISP rarely, does he tend to put too emphasis on them? Overthinking it, and/or trying to be too fine, getting behind in the count and then forcing things?
Zack Wheeler's ERA off-year looks like it may have been entirely a product of bad RISP outcomes. Comparing his 2023 to the few previous years might be very helpful.
It blows my mind that you can be the second best stating pitcher in MLB without runners in scoring position and the 19th worst (out of 150) with.
The next step is to do all 150 starting pitchers, broken down multiple ways. Easier than it sounds ... maybe next week.
Next: Houck.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 7, 2024 4:18:09 GMT -5
See my previous post ...
But the larger issue is this. The Red Sox literally have an infinitely better set of information on Dalbec's great AAA year than we do, since we have only the rawest data, and know nothing about what he was working on or might have modified or changed.
What we do know is that Sox need a backup 1B / 3B who hits righty, so that the rare off days for Devers and Casas can be against lefty pitchers with a big platoon split. They had an interesting 3B in Urias who could have probably learned 1B and they traded him. They chose not to resign their very good veteran (and clubhouse model). That left one guy for the job. Period.
And since you suddenly seem to be a fan of small samples, the first thing he did upon his his September recall was outhit Devers and Casas for a week. So it's fairly evident that he still has the upside he always has; the difference is an apparent major improvement in staying this last year.
That Dalbec is the backup 1B / 3B is as clear as the starters at those positions.
So one might ask rhetorically, how does this post fit in a thread named "Predicting the 2024 Opening Say Roster?"
I am more than half serious about this: someone should start a thread called "Let's Start Bitching Now!" Those of us with a low tolerance for pointless complaints will be grateful.
Doesn’t the fact that the Sox have had all this data on him and have kept him in AAA kind of go against your own point? No.
They kept him in AAA because he wasn't better than Justin Turner, who filled his potential role by moving from DH. And there were no injuries at either 3B or 1B until late in the season. As soon as one of those finally went on the IL, they called him up immediately. (And he outhit Devers and everyone else his first week!)
(They also called him up on 4/10 when Duvall got hurt, swapping him out for Duran after a week; on 5/7 when Arroyo got hurt, for the 6 days before they got Pablo Reyes; and from 6/23 to 6/27 when Verdugo was on the bereavement list. Three brief recalls to fill positions he didn't even play.)
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 6, 2024 21:30:30 GMT -5
If Rafaela's defense can even be top 5 in terms of outfielded value over the stretch of his prime he could wRC in the mid 70s-80s and still be a 2-3 WAR guy. Overall if the plan is to have Rafaela as the every day CF out the gate this year I wouldn't hate it. Mix and match with Duran, abreu, O'Neill and refsnyder in the corners. JBJ and Michael A. Taylor: both had career wRC+'s of 82, both have been right about at 2 WAR/600 PAs.
When JBJ had that run of 3 years in a row from 2017-2019 with a 90 wRC+ he was at 2.8 WAR/600.
I'd say the benchmarks are: if Rafaela has wRC+'s in the 70s, he's a bench player; in the 80s, he's a starter; in the 90s, he's above average; 100+, he's a star.
I've been meaning to post the following nugget for probably 10 days now -- but in the prediction thread.
The Davenport Translation (to MLB) for Rafaela's full year was an EqA of .260 -- MLB average (it's scaled to match BA). He would project to be a bit better this year.
If the Sox reached a similar conclusion, it would go a long way to explaining why they haven't resigned Duvall.
Remember that he was dramatically better in AAA than AA, and that only happens if you were largely cold in AA and hot in AAA. So what we saw in his call-up would just be being cold, rather than some mysterious deficit at hitting in MLB.
My sense here is that when he's hot he punishes mistakes and when he's cold he misses them. His offensive tools are not those of someone who gets hits off of good pitches. But if he can have a sufficient ratio of hot to cold and post a 105 or 110 wRC+, as his performance last year suggests, yes, that a star.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 4, 2024 0:49:28 GMT -5
Among other things Refsnyder might be their backup first baseman, as things stand. Well, he has averaged 3.5 innings at 1B over the last 6 seasons. But that's zero over the last three years and 7 innings per year the previous 3, so maybe not.
(Replies to multiple posts are not working right ... apologies for all of these!)
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 4, 2024 0:26:26 GMT -5
Dalbec has about 950 mlb PA. For two brief periods, one about 95 PA and one about 120, he has hit like Jimmie Foxx. The rest of the time, he has hit like a poor backup SS. That doesn't include going 0-12 with 5k and zero BB in the postseason, when we were a few hits away from the World Series. Isn't it time to give up hope? See my previous post ...
But the larger issue is this. The Red Sox literally have an infinitely better set of information on Dalbec's great AAA year than we do, since we have only the rawest data, and know nothing about what he was working on or might have modified or changed.
What we do know is that Sox need a backup 1B / 3B who hits righty, so that the rare off days for Devers and Casas can be against lefty pitchers with a big platoon split. They had an interesting 3B in Urias who could have probably learned 1B and they traded him. They chose not to resign their very good veteran (and clubhouse model). That left one guy for the job. Period.
And since you suddenly seem to be a fan of small samples, the first thing he did upon his his September recall was outhit Devers and Casas for a week. So it's fairly evident that he still has the upside he always has; the difference is an apparent major improvement in staying this last year.
That Dalbec is the backup 1B / 3B is as clear as the starters at those positions.
So one might ask rhetorically, how does this post fit in a thread named "Predicting the 2024 Opening Say Roster?"
I am more than half serious about this: someone should start a thread called "Let's Start Bitching Now!" Those of us with a low tolerance for pointless complaints will be grateful.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 3, 2024 23:50:12 GMT -5
Dalbec in AAA was .269 / .381 / .557. Clay Davenport's MLB Translation system pays extra attention to high K rates. and correctly raised a red flag on Will Middlebrooks when everyone else was psyched about him.
He has Dalbec's AAA numbers translating as .256 / .331 / .485 (in an average league offense). That's a really good guy coming off the bench.
Do you know if that model has been updated to account for the strike zone changes or rule changes? I’d find it incredibly surprising to see Dalbec run a .330 OBP when he’s a career .297 guy and that number has fallen year after year (smaller sample last year). He’s a fine up and down guy and if he played better defense would be a really good bench player, but i think he should be traded or starting in worcester. The numbers in every league are adjured for league offense.
Re the incredible, he had a .301 career OBP in AAA (176 PA) before last year, when he had a .381. What you're saying is "I’d find it incredibly surprising if a player did something to improve himself."
In this case, it's entirely about the hot to cold ratio. And despite JBJ's failure to improve his ratio, there is precedence for guys fixing it. In 2006 I did a long analysis for the Sox on Carlos Pena that showed he was one of these guys -- red hot or ice cold -- and added my take on what he was doing wrong to try and bust out of his slumps. The Sox picked him up but couldn't find a space for him in the lineup or the next 40-man ... and he went from a 111 OPS + to 145 (three years before vs. three years after).
See my next reply for more.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Feb 3, 2024 22:38:37 GMT -5
Refsnyder posted a .308/.428/.400 vs LHP last year with an elite 15.9% walk rate and more walks than strikeouts. That’s very valuable imo, this team needs at-bats grinders. That was over 39 starts totaling 145 PA. The Sox had 40 games where they faced a lefty SP.
He also started 9 games vs. RHP and had 98 PA against them. He hit .159 / .261 / .195, but the OBP is misleading because he had 5 GDP's. His effective OBP was .196.
As I mentioned, the trade of Verdugo is likely to reduce his starts vs. LHP to 20, platooning with Duran on the road (more would require injuries to LHB). The only guy he'd pinch-hit for is McGuire.
Finally--and this is the killer--Dalbec vs. LHP has numbers nearly as good the last three years, and maybe just as goodpt better (better xwOBA, .366 to .357). He takes Ref's role as lefty-killer, in addition to backing up 1B and 3B.
A team that needs a full-time platoon partner in an OF spot (as we no longer do) would gladly pay him $1.8M.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Jan 31, 2024 22:54:37 GMT -5
Some good thoughts in here. Giving any sort of meaningful ABs to Dalbec is terrifying to me, though. I would much rather give some run to Abreu in RF, who looked the part last year, and use a guy like Valdez as the sometimes-DH LHH bench bat. Dropping Dalbec from the roster also has the advantage of making Refsnyder less redundant, although a Duvall signing would once again cannibalize a chunk of his role. The obvious counterpoint is, "If not Dalbec, who is going to back up 1B?" To that I'd reply, "Refsnyder, or Duvall if he gets signed". If Casas actually gets injured, Dalbec should of course get called up, but otherwise I'd fine with one of those two swapping in to give Casas a breather. Dalbec in AAA was .269 / .381 / .557. Clay Davenport's MLB Translation system pays extra attention to high K rates. and correctly raised a red flag on Will Middlebrooks when everyone else was psyched about him.
He has Dalbec's AAA numbers translating as .256 / .331 / .485 (in an average league offense). That's a really good guy coming off the bench.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,933
|
Post by ericmvan on Jan 30, 2024 23:10:46 GMT -5
My take:
1) Dalbec is (at the least) the backup 1B / 3B (plus occasional RF and emergency SS). I endorse that: let's find out of he can stay in that role.
2) With the swapping-out of Verdudo for a RH bat TBD, Refsnsyfer barely has a role to play. Both Yoshida and Duran had great numbers versus LHP in Fenway, so that gives Ref 20 - 25 starts platooning with Duran on the road. I might keep him around through spring training, but it's hard to justify him for the 26. And that's because ...
3) You really want a LH bat with some talent on the bench, since there are are 3 RHP for every lefty. And in Abreu we have a guy who has already demonstrated that he's a solid backup outfielder. He can also start against RHP with a big platoon spit.
4) How many different ways is re-signing Adam Duball a smart move?
* He gives you a guy who can start the season hitting 5th, thus removing any pressure on O'Neill to hit there, if he latter has not had a great spring. In the long run they can share the 5 and 7 spot based on matchups.
* That allows you to hit Yoshida 6th, where he'd be great [1], and not waste him. The 7 hole is for a lower OBP / higher SA type while 9 is opposite, and they have a great 9 guy in Grissom. If Rafela struggles you have to pop Grissom up to 7, but then there's very little hitting in front of Casas or Devers in the 2 hole. A 9 hitter who can get on base at an average clip or better is a big plus in today's lineups. In short, adding Duvall transforms a week 7 through 9 to a strong one.
* If Duvall gets hurt, you just call up Rafaela and give him another taste of the show, with no pressure on him. It turns a negative into a positive!
* If they're not in contention at the deadline, you can trade him and call up Rafaela for good.
RHB, TBD, Find out what it means to me
Scenarios ...
Stand Pat for Now. The expectation or hope is that Rafaela will be a bottom of the order asset sometime around June 1. In the interim, Dalbec is the most frequent 9th man in the lineup. You might get a better handle on his offense and you definitely will get a better handle on Yoshida's defense in LF. When Rafaela arrives, Duran goes to LF, Yoshida goes to primary DH, Dalbec goes to the bench, and if all the outfielders are healthy, the little-used Refsnyder is dealt.
Major drawback: at this point in the season you're not looking to give Casas or Devers the occasional DH half-rest (in fact, it would probably piss them off). That means Dalbec is a near-regular at DH, and most hitters do not hit as well as normal when they DH. So maybe the only thing you learn about Dalbec is that he's one of those guys.
In short, it's a formula for a slow start on offense (and bad vibes to overcome for Dalbec, who above all needs the opposite). How much of the fan base tunes out?
Rafaela Starts the Season in CF.
No. No. Let's imagine that they have a secret reason to believe that the kid who put up a 74 wRC+ in 82 PA last fall will put up 100+ out of the gate this year. How much is gained by that if you're right? Why not wait a few weeks to confirm your idea works in AAA? The downside is that he starts to doubt himself and you wreck him for good. The reason why there's an edit is that I forgot to point this out, because it seems so obviously wrong. (Note that tremendous ST could change that, if there was a specific new thing he's doing ... JBJ's rookie year was not that.)
Duvall Imminent / Playing Salary Chicken. The only FA I could find that might be an interesting alternative was Aaron Hicks, and he just signed withe Angels, the only other team that was reported to be engaged with Duvall. It's possible that ownership is OK with Duvall at the right price and that they made him an offer a while ago that he rejected. Does his agent give in now? If I were him, I propose to split the difference the two offers ... but do the Sox insist on their number, since there's no other bidder? Hey, everyone can lose!
Same, But With am Ownership-imposed Salary Cap ... that they are struggling to meet. Trading Jansen and handing the closer job to Martin would be the obvious route. That actually wouldn't suck if they could pick up a (secretly?) high-upside replacement.
Having said that ... do I really believe that ownership could be so blindingly idiotic as to impose a salary cap so extreme that at present it makes Duvall impossible? I don't have an answer to that.
Hicks, BTW, cost the Angels the minimum, as the Yankees pay the remainder.
[1] Rationale for this claim is in theory owned by the Red Sox.
|
|
|