Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Wednesday, October 28, 2020

Are analytics to blame for Rays’ Kevin Cash pulling Blake Snell too early from Game 6 of 2020 World Series?

But in the top of the sixth inning, the Dodgers were about to get their third look at Snell, something that Cash wanted to avoid. Per MLB.com:

“The only motive was that the lineup the Dodgers feature is as potent as any team in the league. I felt Blake had done his job and then some. Mookie [Betts] coming around the third time through, I value that. I totally respect and understand the questions that come with [the decision]. Blake gave us every opportunity to win. He was outstanding. These are not easy decisions. ... I felt it was best after the guy got on base—Barnes hit the single—I didn’t want Mookie or [Corey] Seager seeing Blake a third time through. As much as people think that sometimes, there’s no set plan. This organization’s tremendous about giving the staff the trust to make in-game decisions to give us the best chance to win. I respect what unfolded today was pretty tough.”

Cash’s argument about no set plan doesn’t hold up when looking at the numbers. The top of the Dodgers’ lineup — Mookie Betts, Corey Seager and Justin Turner — were 0-for-6 with six strikeouts against Snell on Tuesday.

 

RoyalsRetro (AG#1F) Posted: October 28, 2020 at 08:53 AM | 154 comment(s) Login to Bookmark
  Tags: blake snell

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 1 of 2 pages  1 2 > 
   1. Jose Needs an Absurd Ukulele Concert Posted: October 28, 2020 at 09:07 AM (#5986045)
It's semantics probably but I would argue it's not "analytics" but "over-reliance on analytics." Blindly following the numbers wherever they take you is just as bad a thought process as "the good face." If I'm driving home and my GPS says the highway is traveling at 55 MPH but I look and I see it all backed up then I'm being dumb if I get on the highway.

Snell threw four (4) pitches in the 6th inning. Frankly if you are going to pull him that quickly then I'd be a bit more accepting of the decision if they'd just started the inning with Anderson (or some other reliever, using Anderson is another discussion worth having). I don't like sending a guy out to the mound with the plan on yanking him, you are putting the reliever in a position where he is inheriting a mess, let him start a clean inning.
   2. JJ1986 Posted: October 28, 2020 at 09:25 AM (#5986051)
I think bringing in Anderson is a much worse decision than pulling Snell.
   3. Howie Menckel Posted: October 28, 2020 at 09:33 AM (#5986053)
As much as people think that sometimes, there’s no set plan.


what a clown. Cash said BEFORE THE GAME that Snell would not face anyone a third time.

and in spite of his utter brilliance and his K K K K K K of the top 2 batters in the game, Snell didn't.

hey ass wipe Cash, that's the very definition of what a "set plan" is.

what a dope.
   4. Der-K's emotional investment is way up Posted: October 28, 2020 at 10:17 AM (#5986060)
2/jj1986 - why?
   5. Rally Posted: October 28, 2020 at 10:27 AM (#5986065)
I don't think they handled the problem of what happens when a plan meets reality.

You can make a case that a well rested Anderson is a better option than a tired Snell. But Snell had thrown only 73 pitches, and Anderson after last night pitched 14 2/3 innings in 20 team postseason games. That's a 119 innings per 162 game pace, and relievers have not done that since the 1980s. In 60 regular season games he pitched only 16 innings after 65 last season.

And this postseason you don't even have all the offdays to justify heavy reliever usage. The plan that works for a tired Snell and a fresh Anderson should not be used when Snell is fresh and Anderson is running on fumes.
   6. I am Ted F'ing Williams Posted: October 28, 2020 at 10:37 AM (#5986071)
Being a technocrat means never having to admit you screwed up royally. I would love to rip Cash for hours, but the fact is the Rays' front office wants their manager to be a technocrat more than an assessor of the situation at hand.

You can stupidly use analytics just as much as you can stupidly ignore them. The Dodgers have done the same stupid stuff the last few post-seasons, too, as have a few other teams. Fortunately for LA they ran into an opponent with a similar attitude/arrogance but less post-season experience.
   7. What did Billy Ripken have against ElRoy Face? Posted: October 28, 2020 at 10:54 AM (#5986074)
How many times do you think people have been burned by saying, "But *this* time is the exception to the rule that is well established by the data over a large sample!"? Because that's the argument you're making.

Look, I 100% get that pulling Snell is symbolic of the decline in aesthetic appeal - and, as noted above, whether Anderson was the right guy to bring in is an entirely different question. But let's at least be intellectually honest and consistent with decades of sabermetric principles.
   8. Never Give an Inge (Dave) Posted: October 28, 2020 at 10:54 AM (#5986075)

"It's called playing the percentages. It's what smart managers do to win ballgames."
   9. Adam Starblind Posted: October 28, 2020 at 11:03 AM (#5986079)
It's semantics probably but I would argue it's not "analytics" but "over-reliance on analytics."


I would say it's not understanding analytics. Don't analytics also say not to pull your ace starter who's thrown only 73 pitches through 6, allowed only 2 hits, and is striking everyone out in an elimination game? I mean, maybe the analytic guys need to write that one down. Maybe they didn't think they had to.

   10. RoyalsRetro (AG#1F) Posted: October 28, 2020 at 11:07 AM (#5986080)

Being a technocrat means never having to admit you screwed up royally. I would love to rip Cash for hours, but the fact is the Rays' front office wants their manager to be a technocrat more than an assessor of the situation at hand.


I think managers want this too. They want the decision out of their hands. This is just rigid bullpen roles redux. "This is my eighth inning guy, this is my closer, all I do is press buttons."
   11. bunyon Posted: October 28, 2020 at 11:15 AM (#5986082)
I'm with Adam in 9. No one (well, okay, there have been a few) has ever said, ALWAYS pull your pitcher after two times through the order. They've just pointed out that batters do a lot better the third time through. On average. But guys have thrown shutouts. No-hitters happen. If a stud starter is cruising and your option in the pen is a scrub with a 5 ERA, you stay with the stud. The numbers say, when in doubt, pull the starter.

If your argument is the starter should ALWAYS be pulled the third time through, well, batters also do better the second time through. So nine batter per pitcher, I guess.
   12. Pat Rapper's Delight (as quoted on MLB Network) Posted: October 28, 2020 at 11:17 AM (#5986084)
I think managers want this too. They want the decision out of their hands. This is just rigid bullpen roles redux. "This is my eighth inning guy, this is my closer, all I do is press buttons."

Absolutely. The managers love having that crutch so they don't actually have to -- you know -- manage. That way they can shift the burden of blame from themselves ("Were you on crack tonight?") to the players ("Anderson has been great all year. [All 20 innings of regular season and non-sucking playoff appearances] He's been our guy. We had confidence in him. [based on 20 innings] He just didn't make his pitches tonight.")
   13. Hot Wheeling American Posted: October 28, 2020 at 11:18 AM (#5986085)
It's semantics probably but I would argue it's not "analytics" but "over-reliance on analytics." Blindly following the numbers wherever they take you is just as bad a thought process as "the good face." If I'm driving home and my GPS says the highway is traveling at 55 MPH but I look and I see it all backed up then I'm being dumb if I get on the highway.

Michael Drives Into a Lake

I never loved the idea of this scene when it first aired (in 2007!) and even for a while later; seemed a bit much, even accounting for it just being a joke on a sitcom. But with what we've seen in the decade+ since, and how I've seen people drive relying completely on the gps in their phone...I've needed to revisit my thoughts on how realistic I find the scene.
   14. . Posted: October 28, 2020 at 11:20 AM (#5986087)
Treating someone like Blake Snell as just another random number generator in the 6th game of a World Series is dumb as ####. It is in fact kind of the end game of analytics, but in many ways analytics is dumb as ####, which makes its pretentions to "science" and "rationality" look even more comical.
   15. BDC Posted: October 28, 2020 at 11:25 AM (#5986091)
I dunno. Even based on results, they still just gave up three runs to a strong Dodger lineup - their best performance of the Series. The problem was more when the Rays came up to bat.
   16. What did Billy Ripken have against ElRoy Face? Posted: October 28, 2020 at 11:25 AM (#5986092)
Treating someone like Blake Snell as just another random number generator in the 6th game of a World Series is dumb as ####.
Snell was great in 2018. He was slightly above average in 2019 (104 ERA+), and quite good in 2020 (131 ERA+) - when managed very carefully, limited to 50 innings in his 11 starts.
   17. Lassus Posted: October 28, 2020 at 11:36 AM (#5986095)
Yes. And Alexandra Ocasio-Cortez.
   18. Jeff Francoeur's OPS Posted: October 28, 2020 at 11:39 AM (#5986099)
Snell shouldn't have been pulled at that juncture, but the bigger issue is the Rays offense. It's unlikely that you're going to beat the Dodgers with a first inning solo shot.
   19. Rally Posted: October 28, 2020 at 11:40 AM (#5986100)
I did not agree with pulling Snell. But I have to wonder, if the Rays had scored 4 runs last night, would the headline instead be:

Are analytics to blame for Dodgers’ Dave Roberts pulling Tony Gonsolin too early from Game 6 of 2020 World Series?

After all, the cat dude was pulled in the 2nd inning though he only allowed one run and the Dodgers won it with a bullpen game.

Doesn't matter which team wins, Analytics can never be credited, only blamed.
   20. BDC Posted: October 28, 2020 at 11:47 AM (#5986102)
I could have sworn Gonsolin went about seven innings, as slowly as he was working.
   21. My name is Votto, and I love to get blotto Posted: October 28, 2020 at 11:50 AM (#5986103)
I think #6 gets to the heart of the issue: the math will tell you probabilities, but they are not certainties. If the manager is just a figurehead for the GM, that's one approach; the art comes in by giving a manager the autonomy to know when to play the longer odds.

This is of course a much larger issue. I'm a masochist who occasionally listens to Sean Hannity, and he goes on and on about how the pollsters who gave Trump a 30% chance to win were wrong. That's roughly Paul Goldschmidt's career batting average, and he still gets plenty of hits.

   22. The Duke Posted: October 28, 2020 at 11:54 AM (#5986104)
The obvious answer to this is let Snell go batter to batter and see what happens. If Betts gets a hit (which would be two in a row ), cash is on much firmer ground in taking him out. The risk is that you put your reliever in a much worse situation.

Bottom line is that I doubt anyone in the Rays bullpen could have been considered the best pitcher compared to Snell at that point. Snell had only thrown 70 pitches.
   23. What did Billy Ripken have against ElRoy Face? Posted: October 28, 2020 at 11:57 AM (#5986107)
The obvious answer to this is let Snell go batter to batter and see what happens. If Betts gets a hit (which would be two in a row ), cash is on much firmer ground in taking him out.
Totally agree if you have more than a one-run lead, but there was zero margin for error.
   24. winnipegwhip Posted: October 28, 2020 at 12:10 PM (#5986113)
Grady Little says Hello.
   25. winnipegwhip Posted: October 28, 2020 at 12:13 PM (#5986114)
Stupidest pulling since 1980....Tretiak being pulled after 2 periods to put in Myshkin.
   26. Moses Taylor hashes out the rumpus Posted: October 28, 2020 at 12:24 PM (#5986121)
I think bringing in Anderson is a much worse decision than pulling Snell.

This. And it's not all Cash's fault unless the rest of the team agreed to stop hitting as a protest for pulling Snell.

Leaving a guy in too long and having them blow the game is much easier to put directly on the manager; pulling a guy too soon in a game the offense only scores once means he deserves only part of the blame.
   27. base ball chick Posted: October 28, 2020 at 12:27 PM (#5986124)
i guess the numbers say to never let a pitcher go more than twice through a lineup, regardless of his pitch count or his performance. some nights, pitchers are just ON, and snell was just that, ON. i saw the looks on the faces of the dodgers when snell got pulled and it was - YESSSSSSSSSSSSSSSS!!! now we can win. as soon as i saw cash come out i sez to mahself (well, actually, i yelled it at the tv) it's ovah, it's ALWAYS been ovah (there were also a lot of swear words directed at cash's brains and reproductive organs and use thereof, in there, tell you the truth, but gotta respeck the nanny around here).

in game 4 of the 05 WS, a #4 or #5 pitcher, brandon backe, threw EIGHT shutout innings against the white sox, scattering 4 or 5 hits, no walks. he was ON, but definitely not as ON as snell was last night.

the minute i saw cash pull snell, i KNEW the game was lost, just like the dodgers did. flashed back to 03 and grady little NOT pulling an obviously ineffective and tired pedro martinez. it was SOOO stupid, so incredibly STOOPIDDDDD. WHY WHY??? would you pull snell and send in someone who has sucked ALL of the WS? because the "numbers" say he's due? oh, oops, that's "gut"

i mean, numbers ARE, i guess, absolute, like when you add, subtract, multiply or divide, but situations are absolutely NOT guaranteed, regardless of odds. one thing i have learned in mah life is that the only nevers are - no raising the dead, no taxes except for the wealthy and powerful, and any person born with a male reproductive tract will never become pregnant.

i hate HATE this new TTO baseball, i don't care what the numbers say. and i know that cash's stupidity is not going to bring back anything like the old game, of say, 10 years ago

the only interesting game this WS was game 4, and even then, most of the runs were HR.

   28. Howie Menckel Posted: October 28, 2020 at 12:39 PM (#5986127)
Analytics can never be credited, only blamed.

Analytics can never be blamed, only credited.

ftfy
   29. Infinite Yost (Voxter) Posted: October 28, 2020 at 01:00 PM (#5986143)
I was a bit surprised by the decision myself, but I do think people oversell the result a little bit. It wasn't like the Rays bullpen came in and promptly melted down -- the Dodgers eked out a couple of runs on a double, a wild pitch, and a groundout. They didn't get a third run until the 8th, by which point there was basically zero chance that Snell was still going to be in the game, anyway.

I, too, miss the days of Randy Johnson and Roger Clemens. But they're over. Not that that means Snell couldn't have gone another few outs -- I just think this gets an outsized amount of attention because it represents how the game has changed so drastically of late.
   30. Buck Coats Posted: October 28, 2020 at 01:16 PM (#5986153)
I'll point out for all the talk that the Rays analytics forced them to pull Snell after 18 batters - In the 5 previous games of this Series, the Rays starter faced more than 18 batters 4 times. Glasnow faced 23, Snell himself faced 20, Morton 22, and Glasnow 23 again. So clearly the Rays are willing to stretch their guys a little bit past 18. They just chose not to last night.
   31. I am Ted F'ing Williams Posted: October 28, 2020 at 01:16 PM (#5986154)
Analytics is still a "what have you done for me lately" tool. What was done a few weeks ago is of much lesser importance in managing the current game than what was done the last few days and in the current game.

Blake Snell - what have you done for me lately? 5+ innings of great pitching tonight, not tired/low pitch count.
Nick Anderson - what have you done for me lately? Given up runs in EACH of previous SIX appearances.
Rays Offense - what have you done for me lately? One hit and 8 K's against Dodger relievers to that point tonight.

The decision: Expect the guy pitching well to collapse immediately and the reliever and offense that haven't produced recently to get better immediately.









   32. winnipegwhip Posted: October 28, 2020 at 01:19 PM (#5986158)
i guess the numbers say to never let a pitcher go more than twice through a lineup,


But the data in that argument is flawed when your data will not include the possibilities of successful outcomes.

Wasn't the first inning his third time through the same order in the last 6 days?
   33. What did Billy Ripken have against ElRoy Face? Posted: October 28, 2020 at 01:20 PM (#5986159)
I was a bit surprised by the decision myself, but I do think people oversell the result a little bit. It wasn't like the Rays bullpen came in and promptly melted down -- the Dodgers eked out a couple of runs on a double, a wild pitch, and a groundout. They didn't get a third run until the 8th, by which point there was basically zero chance that Snell was still going to be in the game, anyway.

I, too, miss the days of Randy Johnson and Roger Clemens. But they're over. Not that that means Snell couldn't have gone another few outs -- I just think this gets an outsized amount of attention because it represents how the game has changed so drastically of late.
Seconded.
   34. What did Billy Ripken have against ElRoy Face? Posted: October 28, 2020 at 01:21 PM (#5986162)
But the data in that argument is flawed when your data will not include the possibilities of successful outcomes.
The data include every starter who has been lights-out through 5 innings too.
   35. Ziggy: social distancing since 1980 Posted: October 28, 2020 at 01:21 PM (#5986163)
Here's how induction works (in a nutshell): you take a bunch of past situations that were relevantly similar to the present situation, and you suppose that the present situation will work out more-or-less like the past situations, on average, did. This works better when you have a larger bunch of relevantly similar past situations.

But one key here is that they've got to be relevantly similar. Snell had nine Ks and no walks, through 73 pitches. Times in which Joe McNotgood got smacked around the third time through the lineup are not relevantly similar to this case. Arguably, times in which Snell had been getting knocked around in his first two times through the lineup are not relevantly similar. The thing to do is take a bunch of situations in which a starter K'd a ton of guys without walking anyone and without throwing many pitches, and see how they do the third time through the lineup. (Storing dynamite might, on average, be quite safe. But if you are storing your dynamite next to a campfire, the experience of people who kept it in a fireproof warehouse shouldn't help you sleep at night.)

Basically you've got to find out if a pitcher pitching well in a given game is predictive of the pitcher continuing to pitch well in that game. My guess is that the answer is "yes", although of course it's an empirical question.

Once you figure out how to expect the performance of a pitcher who is pitching well to change over the course of the game, you take your initial expectations about Snell's performance, and update them accordingly. If your priors assign a high probability to Snell having a 7.00 ERA through the rest of the game, but conditional on him K'ing everyone and their brother, you can expect a pitcher to continue to strike out almost everyone (I don't know if that's true - it would depend on the history of pitchers who are pitching well), then you reduce the probability that you assign to Snell being bad for the rest of the game. (Probably by a determinate amount, I'm sure you can correlate strikeouts with expected runs scored.). And then you compare your updated expectation of Snell's performance with your expectation of your performance of your relievers, and choose to replace Snell, or not, accordingly. You want to conditionalize your expectations on whatever information you've got (including Snell pitching well), provided that doing so doesn't increase error bars beyond a range that you are comfortable with.

It doesn't seem to me that analytics are to blame here. Unless you want to count "drawing an inference from a sample that's not similar to the present situation" as "analytics". But when we say that analytics should play a larger role in teams' decision making, I don't think we're saying that they should commit more fallacies, so we probably don't want to count that as "analytics".

   36. The Duke Posted: October 28, 2020 at 01:34 PM (#5986170)
Analytics certainly aren’t to blame. Use of analytics are to blame. What about the Bench coach and pitching coach ? Did all three of them see a different game than we saw? There has to be some intellect put to use with these numbers.

Even in his post game comments, Cash is acting like the decision was completely out of his hands. Each situation needs to be evaluated within its context. The Matt Harvey situation looks the same but is much different: coming back from surgery, already into the eighth Inning, at 100+ pitches. Very different facts.

Too bad for Rays. Everyone says they didn’t score more but I don’t think they needed to if Snell goes 8
   37. Slivers of Maranville descends into chaos (SdeB) Posted: October 28, 2020 at 01:36 PM (#5986172)
I think we're edging closer to a Grady Little rehabilitation.
   38. villageidiom Posted: October 28, 2020 at 01:41 PM (#5986175)
As I said elsewhere, as we're talking about blindly following a set of practices that work generally well in the regular season, despite the different leverage associated with the season ending if it goes wrong, it is apt that the Rays were eliminated on a called third strike. In general, yes, wait for a pitch you can do something with; but in this instance you'll be waiting all winter.
   39. TJ Posted: October 28, 2020 at 01:41 PM (#5986176)
Can't recall who wrote this about Jim Bunning (Bill James? David Halberstam?)but they said Bunning's stuff was so good that opposing players were happy when he was pulled from a game thinking, "We don't care who comes in from the bullpen, he can't be as good as this guy- now we have a chance!"

That thought immediately came to mind when I saw the looks on the Dodgers faces when Snell got pulled...

As for the impact of analytics, I'm moving closer to the view that analytics help mediocre managers seem like they are good. Data that guys like Earl Weaver recognized as valuable that other managers did not gave his team an edge. Today every manager has the same data and tend to rely on it too heavily. Being a good manager today isn't about knowing when to follow the numbers, it is knowing when not to.

   40. Rally Posted: October 28, 2020 at 01:46 PM (#5986178)
I don't know..

Case 1: Pitcher has been dealing for 5 innings. Has thrown 73 pitches. Pulled after allowing a one out single.
Case 2: Pitcher was strong through first 7 innings. In the 8th he gets the first out. Then allows 4 straight hits, 3 for extra bases, is at 123 pitches for the game, and while he had a 3 run lead, the game is now tied before he is finally relieved.

Some might be able to find a middle ground to think one should be relieved and the other should keep pitching.
   41. SoSH U at work Posted: October 28, 2020 at 01:50 PM (#5986179)
I think we're edging closer to a Grady Little rehabilitation.


We shouldn't be. In that instance, the analytics (Pedro struggles past 90 pitches) lined up exactly with what we were seeing, a tiring pitcher who was lucky to get through the seventh.

Pitcher was strong through first 7 innings.


No, pitcher was strong through first six. Pitcher labored through the seventh, only managing to escape because he knew Alfonso Soriano couldn't lay off a slider in the opposite batter's box.
   42. base ball chick Posted: October 28, 2020 at 01:51 PM (#5986180)
Ziggy: social distancing since 1980 Posted: October 28, 2020 at 01:21 PM (#5986163)

Here's how induction works (in a nutshell): you take a bunch of past situations that were relevantly similar to the present situation, and you suppose that the present situation will work out more-or-less like the past situations, on average, did


- that sounds sensible. but that is NOT what happened

even the crappiest pitcher can, for some unknown reason, throw an ace game - see galarraga, humber, bud smith. or the kerry wood game - all the hitters saw him 3 times and they couldn't deal with the pitches any better the 3rd time around. you might could know all a pitcher's pitches, or even be sure you know what he's gonna throw you, but that doesn't mean you can hit that.
   43. Barry`s_Lazy_Boy Posted: October 28, 2020 at 02:12 PM (#5986183)
Snell's "2nd time thru order" stats were worse than "3rd time thru". They took him out 2 innings too late.

Anderson pitched longer than 1.0 only once in the regular season, which was his first appearance of the year on July 25.

In the playoffs, Anderson has pitched longer than 1.0 in 10 out of 13 games.

His ERA as his usage ramped up: regular season 0.55, WC 3.38, ALDS 1.93, ALCS 8.31, WS 9.00.
   44. Ron J Posted: October 28, 2020 at 02:22 PM (#5986187)
#39 It was a James comment on Bunning. Made at a time when we didn't have access to those kind of stats.

And Bunning's worst inning was the 9th. And was ... worse the 4th time through. Not a disaster, but .268/.317/.402 isn't exactly Jim Bunning. But would often be better than any option available in the bullpen.
   45. Steve Balboni's Personal Trainer Posted: October 28, 2020 at 02:27 PM (#5986188)
So here's something to think about with Snell:

During his CYA season in 2018, it was noted that he had very few innings pitched for the number of starts he made (31 GS, 180.2 IP, average of less than 6 IP per start). But the pattern really changed after hist start in July 12, 2018. That night, after a long rain delay to start the game in Minnesota, he came out and was pretty awful: 3 IP, 75 pitches, 5 hits, 3 ER, 3 BBs, only 2 Ks. After the game, he said he wasn't comfortable, was having trouble with his motion, etc.

Then, you have the All-Star break, where he had been snubbed, and then he misses time due to "left shoulder fatigue". When he comes back August 4th, here's the rest of his season (IP in each start):
4,5,5,6,6,6.2,5.1,7,5,6.2,5

It's 11 starts, a 1.17 ERA, 87/17 K/BB in 61.2 IP...and an average of 5.55 IP per start.

Then, in 2019, his average IP/game is even less:
6,7,6,6,3.1,3,6,5.2,6,6.2,6,4.1,6,3.1,.1,3.1,6,5,5,6,2,1.2,2.1
Average IP per start? 4.65.

In 2020? 2,3,3,5,5,5.2,5,5,5.1,5.1,5.2
Average IP per start? 4.54

He has started seven postseason games, all in 2019 and 2020. In those seven starts: 3.1,5.2,5,5,4,4.2,5.1. A total of 33 innings pitched.
Average IP per start? 4.71.


So, in his last 52 starts, he has pitched a total of 251.2 innings - an average of 4.83 IP per start.

The last time Snell finished 6 innings was 22 starts ago...July 21st, 2019.

The last time Snell finished 7 innings was April 2nd, 2019 - 40 starts ago.

Snell has gone into the 8th inning three times in his career. All three times he was pulled after 7.1 IP. The last time was July 7, 2018.

Of course, during most of these 45 starts, Snell has been outstanding. He has a Cy Young Award, and he has generally been excellent in the postseason. He appears to be fragile, and has had stoppages during his career for injuries, fatigue, etc.

This really comes down to whether he has been so good because his manager has been really in tune with when Snell is about to run out of gas, or whether the Rays have been so quick to pull him that they have voluntarily giving up quality innings from their ace for no real benefit.

In the postseason, he has a record of 2-3. The team has a record of 3-4 in his seven starts.

In his last 45 regular-season starts, which span from his return from arm fatigue in early August 2018 through 2020, his record is 19-10; the team's record in his starts is 29-16. That's a similar winning percentage whether he gets the decision or not.

Look, none of us know Blake Snell's arm better than the Rays do. The numbers show he is "babied" as much as any frontline starting pitcher in baseball, probably - but also that the team is not obviously suffering for doing so in the form of blown leads. So it probably isn't fair to say that what Cash did was definitely the wrong move.

But when it is the 6th inning of Game 6 of the World Series, and your best pitcher is having probably his best game since at least early 2019 (he was dominant), and he's thrown only 73 pitches...you let your Big Dog pitch.

It is aesthetically so unsatisfying. It makes it harder for the sport to develop stars or iconic moments. Baseball is supposed to be a marathon, and that is true of both the season (all these teams start 0-0, then we play 162 games, and then a small group of teams that have broken out to the front of the pack get to sprint the final mile (the playoffs) to see who wins the marathon), as well as the individual games (the horse who battles through 7,8, or even 9 innings, before handing it over to the closer who brings it home).

Stuff like this last night is like if the Lakers only played Lebron James 27 minutes in Game 6 of the NBA Finals or something (he played 41 minutes). That would be ludicrous. People want to see the best players in the biggest games, at the biggest moments. When Cash takes out Snell 73 pitches into a game, with Betts coming up next after going 0-for-2 with two Ks against Snell, the fans are deprived of seeing what would have been one of the most exciting moments of the whole Series - a former CYA winner at the top of his game vs an MVP at the top of his game, going at it. 3 1/2 hour "bullpenning" games, instead, isn't going to get it done...

   46. . Posted: October 28, 2020 at 02:49 PM (#5986194)
The existence of the aggregates does not mean that each particular situation should be managed to the aggregates. That doesn't logically follow in the least.
   47. Barry`s_Lazy_Boy Posted: October 28, 2020 at 02:53 PM (#5986197)
#45 is a great great post. The Rays do know Snell, and may be maximizing his productivity with his usage.

The funny thing is, they stuck to their guns on Snell usage, while doubling Anderson's usage in the playoffs and trashing his performance.



   48. Zach Posted: October 28, 2020 at 03:04 PM (#5986202)
Pedro was totally gassed in the Grady Little game. He had barely escaped the previous inning, and everybody watching the game knew it.

In the regular season, set roles and playing the percentages are good. But in an elimination game, you have to decide based on what you see.
   49. The Duke Posted: October 28, 2020 at 03:14 PM (#5986207)
45 is a great post and the conclusion is right. Even with all that info, you have to stick with him when he is so completely dominant. Even if he loses, you can’t feel bad about it - he was your best option.
   50. Howie Menckel Posted: October 28, 2020 at 04:05 PM (#5986223)
hen Cash takes out Snell 73 pitches into a game, with Betts coming up next after going 0-for-2 with two Ks against Snell, the fans are deprived of seeing what would have been one of the most exciting moments of the whole Series - a former CYA winner at the top of his game vs an MVP at the top of his game, going at it.


IMPORTANT: As Verducci noted on the broadcast, the worst MLB regular in terms of SLG PCT against LHP in 2020 was.... Mookie Betts at .218 (I was stunned, too).

so Snell is dealing, and Betts - rather impotent against any lefty - has whiffed and whiffed in this game.

Snell has only thrown 73 pitches.

so of course you remove the dealing lefty, and you replace him with a RHP whose "sell-by date" for 2020 expired two weeks ago. the result was unsurprisingly unappetizing.

sorry if I don't genuflect at the brilliant wisdom of the Rays in that spot.
   51. I am Ted F'ing Williams Posted: October 28, 2020 at 04:07 PM (#5986225)
The data about Snell in post #45 applies to EVERY Rays starting pitcher over the last few seasons, including Morton and Archer. It's an organizational philosophy about starting pitchers, and perhaps it's a very good one when you're playing 3-game series against some weaker competition. But it's pointless to follow it religiously through the World Series.

Billy Beane famously said "My sh*t doesn't work in the playoffs." Well, it's your job to figure out what does work through the playoffs.

   52. SoSH U at work Posted: October 28, 2020 at 04:13 PM (#5986230)
I'd be a lot more understanding of the Rays' actions here had this been a policy they stuck to religiousl. But only once in the playoffs did Snell not see at least one hitter a third time, and that was when he allowed the first two runners to reach in the fifth in Game 6 of the ALCS. And in every other game this postseason, he threw at least nine more pitches (in one instance, he threw 32 more pitches).
   53. Ron J Posted: October 28, 2020 at 04:36 PM (#5986247)
#50 That's a truly amazing example of stupidity. Both by Verducci and anybody else who thinks that Betts has suddenly become a bad hitter against LHP based on 55 AB (64 PA)
   54. Ziggy: social distancing since 1980 Posted: October 28, 2020 at 05:06 PM (#5986271)
sorry if I don't genuflect at the brilliant wisdom of the Rays in that spot.


I don't see much genuflecting going on. It was a #### stupid move, a fact that pretty much everybody realizes.
   55. Walt Davis Posted: October 28, 2020 at 05:15 PM (#5986272)
35 and 45 are quite good. On 45 ... you really want pitches or batters faced there, not innings, but I assume the conclusion is going to be pretty much the same. #30 is also interesting -- I don't know if it holds up in the regular season but it seems clear the Rays are more comfy letting Glasnow and Morton go through the top of the lineup a 3rd time than they are with Snell. As #45 demonstrates, that Snell decision isn't really based on his performance -- in his last 34 starts over 2019-20, he's faced a batter for a 3rd time in just 23 of them, totalling just 90 PA. He hasn't gotten particularly good results but it's a pointlessly small sample.

I will say that I've come around to the belief that if you're saying to yourself "if he gives up a hit here, then I'll pull him", then you should pull him now. You've already lost confidence that he's your best pitcher at this point. That should be even more true in the playoffs. Us Cub fans go back to G7 against Cleveland when Maddon pulled Hendricks in very similar circumstances -- b 5, Hendricks got the first 2 batters, walked Santana on 6 pitches, he's on just 63 pitches. The main difference is the Cubs led 5-1 at the time. Here comes the top of the order and Jason Kipnis (who was a good hitter then).

Maddon went with Lester of course. I still think it was the wrong decision but I can also understand it. I recall saying to myself the "if he gives up a hit here ..." ... but good LHB, followed by Lindor ... if you're thinking of pulling Hendricks, that's when you pull him, especially in favor of the LHP. Lester then didn't really get things done but he didn't do much wrong either really -- a squib hit that the C threw away putting runners on 2nd & 3rd then a very wild pitch that scored both runners. He then pitched scoreless into the 8th, getting the first two before a single ... and Chapman hoping he can get the 4-out save (he did not).

Along the lines of #35 I think from time to time about the standard base-out run expectancy table that we frequently pull out to debate decisions like IBB and sac bunts. But really, this table is pretty much useless information -- it's just an aggregated set of long-running outcomes. In essence it's "here's the probability distribution of what happens when an average pitcher faces an average hitter in this scenario." How does that help you make an actual in-game decision? What is the change in outcome if you IBB this batter? Well, the table is "assuming" your pitcher is average, the batter you're facing is average and the batter in the on-deck circle is average. Obviously, it almost never makes sense to IBB an average batter to get to an average batter.

So we're all clever enough not to do that in a real situation. We know we're facing Barry Bonds and Jeff Kent is on-deck so we plug in their expected outcomes for the batter. But now we've just thrown away almost all of the data that underlies the table -- we've plugged in Bonds' and Kent's small-sample results (maybe just this season's) for the average batter's. If we're taking anything from the table at that point, it's the assumption that a Bonds single is probably about as likely to score the runner from 2nd as any other single. At this point, we haven't taken into account our pitcher yet. We haven't taken into account whather the pitcher is LH or RH or whether we have a rested RH reliever ready for Kent (or a super-LOOGy already ready for Bonds). Every time we take something like that into account, we're throwing away more data from that table but we're never going to have enough data to reliably estimate "how will Jon Lester on 90 pitches do against Bonds and how will Papelbon on one day's rest do against Kent?"

Knowing when drilling down will produce enough signal to be worth the extra noise and when it won't is really, really hard. That's why everybody's so keen to turn it over to computers (and I don't mean baseball).
   56. The Yankee Clapper Posted: October 28, 2020 at 05:21 PM (#5986273)
Seems like the Rays players might have some grounds for believing a ‘Stab-In-The-Back‘ caused their defeat. Will that cause a rift with Cash, or the front office?
   57. What did Billy Ripken have against ElRoy Face? Posted: October 28, 2020 at 05:30 PM (#5986274)
Us Cub fans go back to G7 against Cleveland when Maddon pulled Hendricks in very similar circumstances -- b 5, Hendricks got the first 2 batters, walked Santana on 6 pitches, he's on just 63 pitches. The main difference is the Cubs led 5-1 at the time.
Plus, the Cubs' bullpen was nowhere near as strong/deep as the Rays' - thus Maddon believing his best option was a short-rested Lester.
   58. Walt Davis Posted: October 28, 2020 at 05:35 PM (#5986277)
Those Betts' numbers are kinda amazing, small sample or no. Previously he had K'd about 1 per 8.5-9 PA against LHP; this year over 20% -- that's the sort of dramatic change where it starts to become reasonable to think there's something wrong even with a small sample. (This is what saber nerds mean when they start to talk about after how many PAs does a change start to matter.) He had one XBH about every 9 PA; in this stretch just 1 in 64. His BABIP wasn't good but it wasn't atrocious -- it cost him a hit or two. This was a K and a power outage thing, the worst possible combination.

On the other hand, it would be a truly bizarre change. I don't know if there's any precedent for a good-excellent RHB to totally crater long-term against LHP (before suffering decline against RHP).

I wonder if it's partly a Green Monster thing. Maybe he handled LHP before by pulling them in the air and getting some cheap doubles and HRs and he hasn't adjusted yet. Some website must have platoon/home-road combined splits -- obviously they're not gonna be 1600 at home and 500 on the road, just curious if there's a big gap.
   59. Walt Davis Posted: October 28, 2020 at 05:38 PM (#5986279)
A stab in the back? Melodrama much.

Anyway, it's the Rays. None of the players will be around long enough for a rift to matter.
   60. Eric J can SABER all he wants to Posted: October 28, 2020 at 05:41 PM (#5986280)
On 45 ... you really want pitches or batters faced there, not innings, but I assume the conclusion is going to be pretty much the same.

Snell was given more than 73 pitches in all of his other playoff starts this year, going over 100 once. After his first four starts of the regular season (when the Rays seem to have been stretching him out, with pitch counts of 46-53-59-70), he got 73 or more pitches in every remaining regular season start as well, going over 100 three times (including two of his last three starts). 2019, five times under 73 pitches in 23 starts, most of those coming when he was getting knocked around (the one exception was his first start back from the IL). 2018, 73-plus pitches in all but two starts - again, first two back from a three week absence.

If Snell had been at 100 pitches, I wouldn't be particularly critical of the decision to pull him, because he doesn't have much history of going significantly beyond that (he's cleared 110 only twice in his career). But 73? If a pitcher of Snell's caliber gets through the lineup twice on 73 pitches, I'm giving him a few more.
   61. Eric J can SABER all he wants to Posted: October 28, 2020 at 05:45 PM (#5986282)
Maddon went with Lester of course.

Not to relitigate this situation again, but Maddon in 2016 was kind of doing the opposite of what Cash did here. Cash said going in that Snell wasn't going to face anyone a third time, and then pulled him exactly when he said he would. Maddon said going in that he didn't want to use Lester in the middle of an inning with runners on base, and then immediately used him in the middle of an inning with a runner on.

(To be clear, I think both decisions were dumb, just in different ways.)
   62. SoSH U at work Posted: October 28, 2020 at 05:48 PM (#5986284)
(To be clear, I think both decisions were dumb, just in different ways.)


And in Maddon's defense, there were so many dumb pitching decisions going around that series that it was really hard to keep track.

Maybe that isn't really a defense.
   63. Howie Menckel Posted: October 28, 2020 at 05:53 PM (#5986286)
Those Betts' numbers are kinda amazing, small sample or no. Previously he had K'd about 1 per 8.5-9 PA against LHP; this year over 20% -- that's the sort of dramatic change where it starts to become reasonable to think there's something wrong even with a small sample...... I wonder if it's partly a Green Monster thing. Maybe he handled LHP before by pulling them in the air and getting some cheap doubles and HRs and he hasn't adjusted yet.


that's a possibility, and clearly it would be stupidity to completely ignore the massive slide by Betts against LHP in 2020 - with a new team, new ballpark, new league. 64 PA ain't nothing.

Betts career is .931 home OPS and .861 road. his doubles are 136 home, 102 away - in spite of a modest 70 more PA on the road.

he entered the season with 229 doubles and 139 HR.

yet in 2020, it was 9 doubles and 16 HR after 5 straight seasons of 40+ doubles.

in just a 60-game season, you'd expect around 15 doubles for Betts.
   64. Eric J can SABER all he wants to Posted: October 28, 2020 at 05:59 PM (#5986288)
I mean, the doubles thing isn't necessarily the same as the lefties thing; that's just a Fenway effect in general. Red Sox home games in 2020 had 150 doubles (75 by both the Sox and their opponents); their road games had 86 (43 by both the Sox and their opponents). Which, by the way, made me triple-check that I hadn't just clicked the same split twice.

Willie Stargell had 55 triples in his career. 33 of them were in Forbes Field - which was the Pirates' home stadium for less than half of Stargell's career. He had just over 20% of his career PA in Forbes, but 60% of his triples were there.
   65. bunyon Posted: October 28, 2020 at 06:01 PM (#5986290)
Thinking about this listening to a sort of related meeting, "analytics" isn't a decision. It's just info that guides decision making. The manager is still the guy that walks to the mound and takes the pitcher out. That's the decision. Whatever info he used, it was the wrong decision both because a lot of people said so at the time and it resulted in losing the lead.
   66. Ron J Posted: October 28, 2020 at 06:03 PM (#5986291)
#58 Frank Thomas. Going into 1998 he'd been a monster (as you'd expect) against LHP. In 1998 he hit .226/.327/.365 against them. Partially recovered in 1999 (.253/.387/.453) and then in 2000 put up some OK numbers (.407/.549/.824)

Before then:
1994 .385/.504/.798
1995 .389/.524/.849
1996 .403/.544/.798
1997 .358/.435/.792

Single year platoon splits are just noisy. It's beyond stupid to find significance in platoon splits of a 1/3 of a season.
   67. SoSH U at work Posted: October 28, 2020 at 06:08 PM (#5986292)
in just a 60-game season, you'd expect around 15 doubles for Betts.


In the 18-game postseason, he hit 8. So he was close to being back on track.

And three of them were off lefties.
   68. The Honorable Ardo Posted: October 28, 2020 at 06:27 PM (#5986297)
Cash's comments bother me more than the actual decision. I'd much rather hear him say, "I screwed up. We had a plan to take out Snell at time t. I followed it, but I should've realized he was dealing and left him in." Instead, he's whining, "But THE PLAN!" which makes him sound like a doofus.
   69. Howie Menckel Posted: October 28, 2020 at 06:35 PM (#5986299)
well, the reasons the decision was so stupid were:

- Snell had 9 K, 0 BB and was utterly dominant;
- very low pitch count;
- had struck out the top 3 hitters a combined 6 times in 6 AB;
- Anderson was a batting practice coach of late.

you may or may not want to offer bonus points for Mookie's regular-season ineptitude against LHP in 2020 regular season and a switch to the sort of RHP he prefers. it's already a blowout, with a garbage-time TD or not having no impact on the result.
   70. BillWallace Posted: October 28, 2020 at 07:57 PM (#5986313)
Being a technocrat means never having to admit you screwed up royally. I would love to rip Cash for hours, but the fact is the Rays' front office wants their manager to be a technocrat more than an assessor of the situation at hand.

You can stupidly use analytics just as much as you can stupidly ignore them. The Dodgers have done the same stupid stuff the last few post-seasons, too, as have a few other teams. Fortunately for LA they ran into an opponent with a similar attitude/arrogance but less post-season experience.


Love this.

I once worked in a software development studio that was dead set on trying to measure everything and use data to inform every possible decision. We had a ton of data that should not have been used for anything beyond discussion and theorizing because it was hopelessly confounded by unmeasurable factors or otherwise compromised. The CEO was generally a very smart guy and a low level Silicon Valley darling in our industry but I could not convince him.

Blindly following analytics driven theories without fully understanding what their limitations are (limitations that the originators of the analytics would happily acknowledge) is just as much of a cop out as saying I don't need analytics because I know baseball.

   71. dejarouehg Posted: October 28, 2020 at 08:24 PM (#5986324)
As stated above, the move is indicative of what is wrong with baseball.

When Cash pulled Snell, I turned off the TV and went to bed. I was just rooting for Game 7, but that wasn't baseball to me, or a product that I choose to invest time or interest in.

Whatever your perspective on Sabermetrics impact, it's hard to argue that the product is now less entertaining.

   72. Barry`s_Lazy_Boy Posted: October 28, 2020 at 09:08 PM (#5986331)
Mookie's doubles were down because he hit more of them out of the park. HR/PA rate was a career high. HR/FB rate also a career high.

SLG was second highest in his career, after the MVP season.
   73. Howie Menckel Posted: October 28, 2020 at 09:09 PM (#5986332)
I once worked with a woman whose job it was to ensure that as many published stories as possible would average at least 500 clicks.

I remember talking to her about it. I explained that THESE 200-300 click stories here built up relationships that led to THIS 3,000-click story - so they are loss leaders that leave you ahead in the end.

but for her role, two 350-click stories and one 3,000-click story was a net loss on her personal ledger - which was all she cared about. so he did what she could to severely lessen the number of published stories.

that caused a variety of problems and led to a loss in total revenue. but she had a mission, dammit, and was hellbent on hitting her own mark.

the idea of reviewing clicks is not a bad one at all - it's how you react to what you learn. if someone couldn't produce anything of interest to more than 100 clicks, then you reassign them or you fire them. other employees could benefit if they see that the bulk of their time has been spent for little consequence, so they can adapt to that.

but some idiot gave this idiot an assignment - and not only didn't grasp the potential pitfalls, but also didn't supervise the results and find out that way.

the biggest problem is that the other employees - in this case, the players - can full well see the stupidity. it doesn't exactly lead to loyalty to the business, nor to enthusiasm.

Morton was told this would happen before he signed, so after his own ridiculously early exit, he tried to rationalize it. so did Snell last night.

but I doubt a single player on the team liked it, even at the time.

in this case, the Rays don't sign expensive free agents anyway. but if Snell is healthy in 2021 and they continue to pull this nonsense when his stuff is absolutely electric, I doubt if he is going to grin and bear it.
   74. Eric J can SABER all he wants to Posted: October 28, 2020 at 09:17 PM (#5986333)
Mookie's doubles were down because he hit more of them out of the park. HR/PA rate was a career high. HR/FB rate also a career high.

Also partly a park artifact. Betts with the Red Sox hit more HR in away games than home games, 74-65. As a Dodger this year, it was 11 home, 5 road.

Fenway is a very good hitter's park; it is not an especially good home run park, especially for right-handed hitters. (It has probably the smallest foul territory in MLB, and the Monster, as mentioned earlier, is great for doubles.)
   75. base ball chick Posted: October 28, 2020 at 10:25 PM (#5986348)
BillWallace Posted: October 28, 2020 at 07:57 PM (#5986313)

Being a technocrat means never having to admit you screwed up royally. I would love to rip Cash for hours, but the fact is the Rays' front office wants their manager to be a technocrat more than an assessor of the situation at hand.


- agree

Blindly following analytics driven theories without fully understanding what their limitations are (limitations that the originators of the analytics would happily acknowledge) is just as much of a cop out as saying I don't need analytics because I know baseball


- i would say it is more like understanding that there actually ARE limitations

dejarouehg Posted: October 28, 2020 at 08:24 PM (#5986324)
As stated above, the move is indicative of what is wrong with baseball.


- zackly right

When Cash pulled Snell, I turned off the TV and went to bed.


- you are a wise human being. if i had any sense whatsoevah i woulda done the same instead of screaming and swearing. it's not i was rooting for either team, but i HATE stupid. i also hate that this is the first WS i didn't watch with my mama (thank you NOT you stupid virus)



I was just rooting for Game 7, but that wasn't baseball to me, or a product that I choose to invest time or interest in.

Whatever your perspective on Sabermetrics impact, it's hard to argue that the product is now less entertaining


- yep. the games, except for game 4, were all waaay too long and very boring. too many Ks, too few MOB, too many HR. AND waaaay too long between [pitches and too **** many pitchers used
   76. Srul Itza Posted: October 28, 2020 at 11:05 PM (#5986361)
you might could know all a pitcher's pitches, or even be sure you know what he's gonna throw you, but that doesn't mean you can hit that.


We call that a Mariano situation.
   77. the Hugh Jorgan returns Posted: October 29, 2020 at 12:25 AM (#5986370)
When Cash pulled Snell, I turned off the TV and went to bed


I continued to follow, but immediately switched allegiances. Howie was literally 30 seconds after me, posting the same thing about the Rays needing to suffer for such a lame move. And it really was a lame move.

Most of those inning figures posted in 45 where Snell has gone 5 or 6 innings, he throws a lot of pitches, 80+ and in many cases close to 100. This was not the case here. Dude was in the 70's and cruising.

   78. McCoy Posted: October 29, 2020 at 08:34 AM (#5986386)
The Snell situation reminds me of the Carpenter brouhaha in 2011 for the NLDS where MGL was adamant that Carpenter should have been pulled in part because of facing batters a third time and because of the ability of relievers. Looks like Tango is considering that pulling the starter might not be obvious and it appears it's because of basically the same information we had back in 2011 that got pooh-poohed.
   79. kwarren Posted: October 29, 2020 at 08:49 AM (#5986388)
This issue that has infected baseball recently of starting pitchers continuing to be become more and mores less relevant, so many pitching changes in each and every game, and games getting continually longer with all that is involved in each pitching change could be largely corrected by limiting teams' active rosters to 23 players instead of 25. By limiting the roster this way, it would force teams to have less pitchers on their roster, and require more IP per appearance by both starters and relievers.

It would also be necessary to limit the number of times that pitchers can be sent and recalled to the minors, so that each team doesn't have five or six relievers being shuttled back and forth. It might also make sense to limit the number of pitchers a team can use in a season, maybe 20.

A typical pitching change requires a visit to the mound, a player jogging in from the bullpen, and then throwing eight more warm-up pitches all accompanied by an extra two minutes of exciting advertising. This often happens 13 or 14 times a game in the modern baseball game.
   80. Steve Balboni's Personal Trainer Posted: October 29, 2020 at 09:18 AM (#5986393)
As a lifelong Bill James fan, been posting on this site forever, etc., this may be provocative, but I'll say it: Stuff like this Snell situation makes me understand why a lot of people wanted Jack Morris in the Hall of Fame. I want the big moments, with the biggest names, competing against each other, with everything on the line. These are sports - but they are also entertainment. I watch baseball for pleasure, relaxation, and joy.

It has been 29 years since that wonderful Game 7 of the 1991 World Series. I was rooting for the Braves that night, because they had been such an unlikely contender.

The game was 10 innings, and lasted 3:23.
John Smoltz, young stud, future Hall of Famer vs Jack Morris, grizzled veteran, future Hall of Famer.
Smoltz pitches into the 8th inning before tiring. Morris pitches 10 innings, throws 126 pitches. There was zero f**king chance he was coming out of that game.
The drama of that game was unreal. 29 years later, it still gives me joy to think about it, or watch some of it on YouTube.
Imagine if that had been a "bullpenning" game. We wouldn't be talking about it today. In 29 years, are we going to be talking about this game? Will anybody remember the name of the pitcher who came in for Snell? In fact, in part because of decisions like this, will people be talking about Snell in 29 years? These are the moments when legends and narratives get built, and it helps feed the popularity of the sport. This kind of stuff does ZERO for building the popularity of baseball.

It also would have lasted 4+ hours.
   81. McCoy Posted: October 29, 2020 at 09:51 AM (#5986398)
A pivotal moment in a do or die World Series game? Yeah, we’ll be talking about it 29 years from now unless baseball goes all Brockmire.
   82. BDC Posted: October 29, 2020 at 09:57 AM (#5986401)
I've expressed nostalgia for ace starting pitchers too, and some of my best memories are Seaver v. Carlton and the like. But at this point it may be like nostalgia for one-platoon football.

I think it's still possible to enjoy baseball without prominent pitchers, but it means shifting interest almost completely onto position players. Pitching staffs are merging together into a kind of amorphous mass – kind of the way NFL defenses have tended, to continue the analogy. Some staffs, like some NFL defenses, are still measurably better than others, but only encyclopedic fans are likely to know how individual guys contribute to that. (And sure, here and there a really gifted pass rusher or cornerback is still a star, which will continue to be the case with a given starter or closer here or there.)

So I guess one way to keep enjoying the game is to focus on the offense and the fielders. Then your only worries are the crushingly slow plate appearances and the lack of balls in play :)
   83. BDC Posted: October 29, 2020 at 09:57 AM (#5986402)
Double post, I'm getting rusty at this. Time to bring in a relief poster.
   84. Rally Posted: October 29, 2020 at 09:59 AM (#5986404)
The Snell situation reminds me of the Carpenter brouhaha in 2011 for the NLDS where MGL was adamant that Carpenter should have been pulled in part because of facing batters a third time and because of the ability of relievers. Looks like Tango is considering that pulling the starter might not be obvious and it appears it's because of basically the same information we had back in 2011 that got pooh-poohed.


Do you have a link?

I assume this is the game where Carpenter went on to face the top of the Phillies lineup for the 4th time, finished the game with a 3 hit shutout, and Ryan Howard's time as a feared slugger came to a sad ending.

I wonder if that was the last playoff game where both starters pitched complete games.

#80, agree with you 100%.
   85. sunday silence (again) Posted: October 29, 2020 at 10:00 AM (#5986405)

On the other hand, it would be a truly bizarre change. I don't know if there's any precedent for a good-excellent RHB to totally crater long-term against LHP (before suffering decline against RHP).


Have you looked at his 2019 season? He shows the same massive reverse split 92 pts in slug. He doesnt have the high KOs though, about 12% vs LH and 20% vs RH. He also had a large reverse split in 2016, but no others.

It certainly seems there's enuf data pts there to form some sort conclusion. COuld it be LH are pitching him differently? But some primates, such as Snapper, dont believe reverse splits exist at all. (Don Zimmer says "Hi.")
   86. sunday silence (again) Posted: October 29, 2020 at 10:06 AM (#5986407)
interesting article from a few years back on hitters with reverse splits;

https://www.mlb.com/news/aaron-judge-leads-mlb-players-with-odd-splits-c262767352
   87. sunday silence (again) Posted: October 29, 2020 at 10:28 AM (#5986413)
this one says they were pitching him more inside in 2019, but the numbers dont really seem very different:

https://www.pitcherlist.com/going-deep-something-is-off-with-mookie-betts/
   88. Rally Posted: October 29, 2020 at 10:36 AM (#5986415)
   89. Rally Posted: October 29, 2020 at 10:39 AM (#5986418)
Unlike Carpenter in 2011, there is zero chance Snell could have made 1 run in the first inning stand up. Oh, he may have done what Anderson couldn’t and pitched out of the 6th, but soon after he would have been out of there. He just has never been trained to pitch deep into games.
   90. Ron J Posted: October 29, 2020 at 10:40 AM (#5986419)
#85 I know about his 2019 splits too. See the Frank Thomas splits I posted. A two year crater in a bigger sample size. Platoon splits are inherently extremely noisy. and 60 PAs tell you zip.
   91. snapper (history's 42nd greatest monster) Posted: October 29, 2020 at 10:52 AM (#5986422)

Blindly following analytics driven theories without fully understanding what their limitations are (limitations that the originators of the analytics would happily acknowledge) is just as much of a cop out as saying I don't need analytics because I know baseball.


It's even worse in my opinion, as long as you have a reasonable number of experienced people opining on the decision.

It's very hard for a group of 5-10 experienced people to all agree on something and have it be catastrophically wrong. They may be wrong on the margin, and you may choose the 45% side of a 55:45 decision, but they're not going to do something batshit crazy.

Poorly designed and understood analytics can and does tell you to do batshit crazy things.
   92. McCoy Posted: October 29, 2020 at 10:57 AM (#5986423)
   93. McCoy Posted: October 29, 2020 at 11:00 AM (#5986424)
There is also another thread from the blog that got split off from the one above that looked into in depth.
   94. Barry`s_Lazy_Boy Posted: October 29, 2020 at 11:02 AM (#5986425)
Over his career, Mookie hits RHP and LHP about the same. That is what I would expect going forward.

One of MGL's beliefs was that every hitter had the same platoon differential.

   95. DL from MN Posted: October 29, 2020 at 11:34 AM (#5986426)
he may have done what Anderson couldn’t and pitched out of the 6th, but soon after he would have been out of there


I agree with this. Snell probably could have pitched the 6th but he's done after that. He had about 20-25 pitches left.
   96. McCoy Posted: October 29, 2020 at 11:56 AM (#5986428)
   97. Rally Posted: October 29, 2020 at 11:56 AM (#5986429)
Poorly designed and understood analytics can and does tell you to do batshit crazy things.


While true, I don't think the Rays could possibly be described as using poorly designed analytics. The way they built such a good team without spending any $ is impressive, and especially so given that they are not the only quant-controlled team out there. That might have been the case back in 2002 when Moneyball was written. Not literally true, there were other teams doing analytics, Billy didn't pull DePodesta from his mother's basement, Paul was already working for a big league team (Indians). But there certainly were a lot of teams that had no use for analytics, and among the teams that did the analytics were not as big a part of the process.

But these Rays have built this team in an environment where every team is heavily into analytics. They are doing it on a 50-60 million payroll and competing with big market teams that are fully into analytics and combine that with 200 million payrolls.

They have assembled a better bullpen than the Yankees with castoffs and cheap minor league pickups while the Yankees are spending nearly 40 million on 3 relievers (Chapman, Britton, Ottavino). 40 million would cover most of the Rays payroll but those 3 Yankees weren't even as productive as the trio of Anderson, Fairbanks, and Castillo.

Then you've got Arozarena. I don't know what kind of analytics/scouting they used to find that guy. His minor league stats are very ordinary. But they did find him, got him in a trade from another organization that knows a little bit about analytics and player development. Maybe they got lucky, but they sure seem to have been getting lucky a lot. They are either using very good analytics, or they are simply the luckiest team in the game.
   98. Rally Posted: October 29, 2020 at 11:59 AM (#5986432)
One of MGL's beliefs was that every hitter had the same platoon differential.


He limited that belief to right handed hitters. He didn't dispute that some LHB had stronger platoon splits than others.
   99. snapper (history's 42nd greatest monster) Posted: October 29, 2020 at 12:08 PM (#5986434)
While true, I don't think the Rays could possibly be described as using poorly designed analytics.

OK, then they misunderstood/misapplied the analytics. There's no way to logically conclude that an exhausted Nick Anderson (who was dreadful in both the ALCS and WS), is a better option that a relatively fresh Blake Snell who has mowed down the three hitters you need to get out.

It's like a parody of someone who doesn't really understand analytics. "Oh, SP's do worse the third time through the lineup? Gotta pull him."

It's like PH for Mike Trout with a LHB to get the platoon advantage. Cur Mr. Burns' quote.
   100. Ziggy: social distancing since 1980 Posted: October 29, 2020 at 02:22 PM (#5986483)
Unlike Carpenter in 2011, there is zero chance Snell could have made 1 run in the first inning stand up. Oh, he may have done what Anderson couldn’t and pitched out of the 6th, but soon after he would have been out of there. He just has never been trained to pitch deep into games.


For all the gnashing of teeth that this decision prompted, it would no doubt have prompted even more teeth gnashing had it changed the outcome of the game. Yankee pitching completely shut down the Rays line up. We didn't know that that was how it was going to be at the time. But in retrospect, the Rays were going to lose this game either way.

and 60 PAs tell you zip.


It's almost like somebody has a law about this. I think he has red hair.
Page 1 of 2 pages  1 2 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Dynasty League Baseball

Support BBTF

donate

Thanks to
Randy Jones
for his generous support.

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogIs the Hall leaving out too many players?
(83 - 1:05pm, Dec 02)
Last: Rally

NewsblogNBA Post-Bubble offseason thread
(1979 - 1:03pm, Dec 02)
Last: Der-K's emotional investment is way up

NewsblogLet’s Fix MLB’s Salary Arbitration System: The Arbitrators
(7 - 12:54pm, Dec 02)
Last: Rally

NewsblogSt. Paul Saints to become Twins' top minor-league team; Class AA in Wichita
(8 - 12:42pm, Dec 02)
Last: RoyalsRetro (AG#1F)

NewsblogEmpty Stadium Sports Will Be Really Weird
(10679 - 12:41pm, Dec 02)
Last: Biscuit_pants

NewsblogBill James: The Biggest Problem With WAR
(107 - 11:45am, Dec 02)
Last: David Nieporent (now, with children)

NewsblogSources: Isiah Kiner-Falefa to replace mainstay Elvis Andrus as starting Rangers shortstop next season
(16 - 11:20am, Dec 02)
Last: Ron J

NewsblogA push to recognize the statistics of Black players from baseball’s era of apartheid
(43 - 11:10am, Dec 02)
Last: Mike Webber

NewsblogPrimer Dugout (and link of the day) 12-2-2020
(1 - 10:41am, Dec 02)
Last: Jefferson Manship (Dan Lee)

NewsblogOT - Soccer Thread - Winter Is Here
(216 - 10:37am, Dec 02)
Last: AuntBea odeurs de parfum de distance sociale

NewsblogTwins place Eddie Rosario on outright waivers
(7 - 9:55am, Dec 02)
Last: The Duke

Hall of MeritMost Meritorious Player: 2020 Discussion
(19 - 9:45am, Dec 02)
Last: DL from MN

Hall of Merit2021 Hall of Merit Ballot Discussion
(382 - 4:05am, Dec 02)
Last: Bleed the Freak

NewsblogOT - November* 2020 College Football thread
(209 - 8:27pm, Dec 01)
Last: Howie Menckel

Hall of MeritMost Meritorious Player: 1933 Discussion
(20 - 5:59pm, Dec 01)
Last: John (You Can Call Me Grandma) Murphy

Page rendered in 0.7948 seconds
48 querie(s) executed