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

Saturday, November 11, 2017

Mets pitchers Syndergaard de Grom | Newsday

One way to stop the insanity of really long games and help offense would be to limit teams to 11 pitchers on the roster.

The Mets intend to rethink the way they use a pitching staff that disintegrated during a 92-loss season, sources have told Newsday. It’s a philosophical shift that will shape their decisions as general manager Sandy Alderson begins a critical offseason reboot.

With the exception of Noah Syndergaard and Jacob deGrom, Mets starters may be shielded from facing lineups more than twice in a game, mirroring an industry-wide trend, according to a source. The adjustment comes after a season in which team officials watched many of the Mets’ starters fade badly as they pitched deeper into games.

Fewer innings on a nightly basis from the starters means a greater emphasis on the bullpen, and sources said the Mets will look to establish a norm of carrying eight relief pitchers, one more than the standard seven.

Jim Furtado Posted: November 11, 2017 at 07:02 AM | 25 comment(s) Login to Bookmark
  Tags: mets

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.

   1. PreservedFish Posted: November 11, 2017 at 09:24 AM (#5574459)
I couldn't agree more, Jim.

But even if you have 14 pitchers, you cannot realistically limit your starters to 4-5 innings and also limit your relievers to short outings. Maddon and Roberts both tried that repeatedly and got burned repeatedly, and that's in the postseason, when everyone is willing to go above and beyond.

I do wonder if the Astros' successful use of piggybacked starters in the postseason - guys like Peacock, Morton and McCullers all going 3+ innings in relief - will become the new vogue. Maybe that 6th starter type that doesn't have a real third pitch can still dial it up for 2+ innings at a time.
   2. snapper (history's 42nd greatest monster) Posted: November 11, 2017 at 11:28 AM (#5574480)
One way to stop the insanity of really long games and help offense would be to limit teams to 11 pitchers on the roster.

This. Yesterday please.

But even if you have 14 pitchers, you cannot realistically limit your starters to 4-5 innings and also limit your relievers to short outings.

Correct. You can use your 7 RP one inning at a time, but then your SPs need to go 6. If your SPs only go 4, several of your RPs need to average 2+ IP per outing.
   3. cercopithecus aethiops Posted: November 11, 2017 at 01:09 PM (#5574490)
sources said the Mets will look to establish a norm of carrying eight relief pitchers, one more than the standard seven


Yeah, this is backwards. It could work with eight "starters" and five relievers, but not the other way around. Even then, at least two of the starters need to be real starters who can go through a lineup three times instead of two.
   4. cmd600 Posted: November 11, 2017 at 01:12 PM (#5574491)
Correct. You can use your 7 RP one inning at a time, but then your SPs need to go 6. If your SPs only go 4, several of your RPs need to average 2+ IP per outing.


But you aren't going to be limited to just seven relievers. The Indians, who had about as stable a bullpen as you can get, while throwing the second fewest bullpen innings, used 13 relievers last year, and 11 relief appearances from three of their starters.

help offense


9.3 runs/game last year. Does offense need help or do we just want to shape it differently? And more offense, especially if it comes from stringing hits together rather than a home run is going to lengthen games.
   5. snapper (history's 42nd greatest monster) Posted: November 11, 2017 at 01:21 PM (#5574493)
But you aren't going to be limited to just seven relievers. The Indians, who had about as stable a bullpen as you can get, while throwing the second fewest bullpen innings, used 13 relievers last year, and 11 relief appearances from three of their starters.

You can't just send RPs up and down to the minors at will. Most of your good RPs won't be likely to have options left. You can play around with your bottom 2 or 3 guys, but your top 5 RPs are going to be on the roster all year long, and most guys can't give you more than 60-70 IP anymore.

So that's ~300 IP from your top 5 RPs. If your top 2 SPs give you 175 IP each. That's 650 IP. Teams need 1450. If you're other 3 SPs only average 125 IP, you're 425 IPs short.

That's 425 IP from your 6th and worse RPs. It just doesn't work.
   6. cmd600 Posted: November 11, 2017 at 01:35 PM (#5574496)
Yeah, this is backwards. It could work with eight "starters" and five relievers, but not the other way around. Even then, at least two of the starters need to be real starter


You would need something just shy of 19 outs from your top two (66 games), and 13 from your bottom three (96 games), to reach the Reds' last in the league 820 starters' IP from 2017. You are going to need to be both good and healthy, so maybe the Mets aren't the team that can pull this off, but this is feasible.
   7. The Yankee Clapper Posted: November 11, 2017 at 01:51 PM (#5574502)
But you aren't going to be limited to just seven relievers.

Lots of teams can't find 7 decent relief pitchers, so good luck finding 8, or more if you want to yo-yo 1 or 2 back & forth to the minors to increase the availability of fresh arms. That also leaves a rather thin bench which has to prioritize playing multiple positions over being good in a more limited role. Could be tough to manage all the day-to-day injuries, as well as late inning strategy, with such limited resources. Perhaps it still makes some sense to avoid having a weak starting staff go 3 times through the order, but there's no magic bullet here - trying to minimize that weakness exposes others. The better option is to improve the starters, but easier said than done.
   8. cmd600 Posted: November 11, 2017 at 01:59 PM (#5574504)
You can't just send RPs up and down to the minors at will. Most of your good RPs won't be likely to have options left.


Sticking with the Indians, they opened the season with five guys in the bullpen out of options, and a sixth in Allen who was never going to get optioned. Despite just one rotating spot in the pen, they still used 13 relievers. It's not that difficult, especially with the 10 day DL now.

That's 425 IP from your 6th and worse RPs. It just doesn't work.


You were already getting about 300 IP from those guys anyway. But yes, you are going to need to be more aggressive with your top guys. 190 each from the top two starters, 70 from the top five relievers leaves you at 345 IP from those guys. The question becomes can you find sixth, seventh, and eighth relievers who are better than your bottom three starters the third time through the order?
   9. Pat Rapper's Delight (as quoted on MLB Network) Posted: November 11, 2017 at 02:01 PM (#5574506)
You are going to need to be both good and healthy, so maybe the Mets aren't the team that can pull this off, but this is feasible.

Good luck to any GM whose prevailing roster construction strategy depends on both high effectiveness and high availability from all of the top 13 pitchers in the system.
   10. cmd600 Posted: November 11, 2017 at 02:09 PM (#5574508)
Lots of teams can't find 7 decent relief pitchers


Whatever "decent" is, is it the baseline? We're looking to improve upon a team's worst starters the third time through the order. 2017 saw an .800 OPS against starting pitchers a third time through the order, and that's obviously with the better pitchers getting most of those PAs. Your sixth through eighth guys can allow an .800 OPS and still be an improvement, you don't even need to get "decent".
   11. cmd600 Posted: November 11, 2017 at 02:12 PM (#5574510)
Good luck to any GM whose prevailing roster construction strategy depends on both high effectiveness and high availability from all of the top 13 pitchers in the system.


I pondered not saying that because of this kind of response. Because the reality is, it's true everywhere. Conventional, avoid-TTOP, four man rotation, or any other strategy you want from the pitching staff, the most important thing is still how healthy your best pitchers are.

We're trying to scrape up one to two percent advantages here and there, not find a magic pill to turn bad rotations into great ones.
   12. yo la tengo Posted: November 11, 2017 at 04:42 PM (#5574550)
Is it at all sustainable to buy some innings from starters on their 'throw' days between starts? Especially if you are looking to limit their innings when they do start? Still seems a bit crazy...
   13. snapper (history's 42nd greatest monster) Posted: November 11, 2017 at 06:23 PM (#5574565)
Whatever "decent" is, is it the baseline? We're looking to improve upon a team's worst starters the third time through the order. 2017 saw an .800 OPS against starting pitchers a third time through the order, and that's obviously with the better pitchers getting most of those PAs. Your sixth through eighth guys can allow an .800 OPS and still be an improvement, you don't even need to get "decent".

How about managers actually manage? Make decisions based on the pitcher's stuff that day, the score, the game conditions, etc.
   14. Pat Rapper's Delight (as quoted on MLB Network) Posted: November 11, 2017 at 07:00 PM (#5574570)
How about managers actually manage? Make decisions based on the pitcher's stuff that day, the score, the game conditions, etc.

But that puts the manager on the hot seat in case too many bad decisions start piling up. The advantage (from a self-preservation perspective) is that such tightly-slotted iron-clad roles allow the manager to say, "That's his role. He just didn't perform today." on days that The Plan blows up.
   15. Walt Davis Posted: November 11, 2017 at 07:50 PM (#5574579)
First, 8 relievers is already standard usage these days.

I'm not sure who had options on the Indians but their top 6 relievers by IP gave them 383 innings. The other 7 guys combined for just 91 give or take. As you noted, they got relief innings from some guys who were mainly starters. Anyway, their 489 relief innings were the FEWEST in the AL. They aren't a blueprint for how to take innings away from starters.

This "movement" is, for now, overblown. Relief innings sat steadily at 490-500 for quite a few years. Over the last 2-3 years, it has pushed out to 550 this year, requiring the addition of the 8th reliever. Maybe it will stabilize there, maybe it will keep growing, maybe it will fall back. But an extra 50-60 relief IP is losing just 1 to 1.5 outs off of the average start. It's certainly at its lowest level ever but the average start is still 23.6 batters faced.

In 2017, there were only 16 guys who started every game they pitched and averaged 20 or fewer batters. The most IP thrown by any one of these was 55.1 by the terrible Brett Anderson. The best pitcher here is Syndergaard who got hurt. Of course a lot of those guys would have gotten moved to the pen so I did a split search, possibly correctly. Now we've 56 pitchers but Anderson's 13 starts and 55 innings still leads the way. Only 14 pitchers who average <=20 BF in their starts made even 5 starts. Noah, Richards and Delgado were hurt, the others have ERAs between 5 and 10.

If I did the search correctly, the Dodgers led the majors with 46 starts of 20 or fewer batters. I know they had some injuries but that seems silly given their rotation. The median teams were the Cubs-Pirates with 27-28. The Giants, oddly, had the fewest with 10 followed by the more likely suspects in the Nats and DBacks. In fairness to the Dodgers, they also led easily with 25 wins in such starts. The median number of wins in such starts was 10 (Cubs had 14, Pirates has 12) ... 314-481 overall so this is surely dominated by bad starts.

Barring a radical change in reliever usage and/or roster sizes, there is no way any team can get remotely close to avoiding 3rd time effects. And mainly what we've done over the last few years is save the starter from facing the 6th and 7th batters in the lineup. It is still roughly the case that the guy has to go at least 5 innings. What happens at that point is a function of the score, where the lineup is, whether he had to be PH for ... and even then, he's probably going out to at least start the 6th but now is more likely to get pulled at the first sign of trouble.

An average starter inning is about 4.3 BF. If restricted to 18 batters, that gets you one batter into the 5th inning. Note, that some pitchers are better/worse doesn't matter a whole lot. Poor starters probably average 4.4 to 4.5 BF/inning (note that's a WHIP of at least 1.5 ... HBPs and DPs) which gets you only through 4; very good pitchers might be at 4.1 which gets you two batters into the fifth.

The typical game is about 38-40 batters (that's pretty consistent through MLB history if I remember right), just over 38 in 2017. There was no difference in 2017 between BF/IP for starters vs. relievers (4.288 vs 4.275) so if starters are only facing 18, relievers have to face 20. That would be a 38% increase in total reliever workload. A 16-man staff just might pull it off. Of course whether relievers 9-11 are really going to be better than your starters on their 3rd time through ...

Yeah, but we would only limit the crap starters to 18 BF per game. OK but, at the moment, there's not a big difference. The average starter is at 23.5 BF but there were only 40 starters with 20+ starts who averaged at least 25. There are another 22 who were between 24 and 25. So, realistically, your top 2 starters might average 25 each. If they make 30 starts each (lucky you!), that's 420 batters you've saved your pen or about 100 innings (pretty nice). Now you're down to a 14-15 man staff.

And now you figure that you don't care how crappy your relievers are in low-leverage innings. Starter does his 18 batters but is getting rocked. So this is when you use relievers 7-11. Of course getting rocked for 18 batters means we may not even be out of the 3rd inning yet and your lousy relievers have at least an average of 25-26 batters to get through to finish this game. And they aren't very good relievers so it's probably 27-28. That's taking you all 5 of those relievers to get through.

There are other possibilities besides roster expansion. Maybe you have 8 "starting" pitchers and you can handle a 4-game rotation of

4 guys who average 18 per start
4 guys who average 9 in relief
5 guys who cover the other 11-13 per game.

A typical relief slot (assuming 8) averaged 293 BF in 2017. Unfortunately, that usage requires those 5 guys to average about 390 each. To maintain current short reliever usage, you've got to use 7 guys there for a 15-man staff. Note those 4 long relievers are averaging 360 BF which is also a quite heavy load. The 4 starters are facing 720 while currently averaging 766 so it's not a huge break for them, especially given less rest between starts. (you could potentially alternate the 4 starters and long relievers so they averaged 27 batters per 2 appearances but that's pushing much more onto your #6-8 "starters").

How about 2 starters who average 22 and 2 who average 18 along with the 4 relievers averaging 9. Now you could retain current usage with just 6 guys in short relief, 14 overall. Even 5 isn't out of the question as that would be an increase of about 30 batters each.

If everybody stays healthy, the last model is almost feasible, at least in the aggregate. But you've cut a days' rest off of your starters and your 5 short relievers are almost required to cover 6 innings every 2 games so you get a pattern like this:

ABC
DEA
BCD
EAB
CDE

So three appearances every 5 games, 96-100 appearances on the year covering about 75 innings. Oh yeah, look at all those mid-inning pitching changes. :-)

Leading back to the irony of the intro. Pushing back to 11 pitchers from 13 is going to increase individual reliever workloads which is likely to lead to more mid-inning pitching changes -- push the starter until he gets into trouble in the 6th, mid-inning change, reliever pitches to 3 batters to get out of the inning, first 2 batters of the next inning, mid-inning change, maybe this guy finishes the 7th and gets through the entire 8th ... or maybe a mid-inning change for a LOOGY or a mid-inning change for the closer to go 4 outs.
   16. cercopithecus aethiops Posted: November 11, 2017 at 08:15 PM (#5574584)
I was thinking something more like 2 guys who average 23 and 6 who alternate between starts of 18 and relief appearances of 10 - 12. IOW, tandem starters for the #3 - #5 slots in the rotation. Five one-inning relievers should be able to cover the rest, no? Question is whether your #6 - #8 starters facing 10 batters are better than your #6 - #8 relievers facing 4. Answer is probably no.
   17. snapper (history's 42nd greatest monster) Posted: November 12, 2017 at 10:44 AM (#5574644)

But that puts the manager on the hot seat in case too many bad decisions start piling up. The advantage (from a self-preservation perspective) is that such tightly-slotted iron-clad roles allow the manager to say, "That's his role. He just didn't perform today." on days that The Plan blows up.


Are you trying to win games or protect the manager?

Given how little power the field manager has relative to the front office these days, I see no reason why he would or should be allowed to adopt a system that only serves to cover his ass.
   18. cercopithecus aethiops Posted: November 12, 2017 at 12:55 PM (#5574673)
Are you trying to win games or protect the manager?


Current bullpen usage didn't evolve to protect managers. It evolved to maximize availability and effectiveness. Future evolution of starter usage will be driven by the same types of considerations. Protecting managers is a side effect -- feature or bug is in the eye of the beholder.

Given how little power the field manager has relative to the front office these days, I see no reason why he would or should be allowed to adopt a system that only serves to cover his ass.


Seems to me that the impetus for these changes is coming from the FO much more than the field. Especially WRT TFA.

   19. snapper (history's 42nd greatest monster) Posted: November 12, 2017 at 01:27 PM (#5574681)
Current bullpen usage didn't evolve to protect managers. It evolved to maximize availability and effectiveness.

I disagree. It evolved to eliminate decision making.

There is no way using three or four relief pitchers, for exactly one inning each, in exactly the same order, to protect any lead from 1-3 runs, regardless of platoon advantage, when the heart of the opponents order is up, the number of pitches they throw, how heavily they've been worked recently, or their effectiveness that day, optimizes either effectiveness or availability.
   20. BDC Posted: November 12, 2017 at 01:38 PM (#5574689)
There is no way using three or four relief pitchers, for exactly one inning each, in exactly the same order, to protect any lead from 1-3 runs, regardless of platoon advantage, when the heart of the opponents order is up, the number of pitches they throw, how heavily they've been worked recently, or their effectiveness that day, optimizes either effectiveness or availability

That's true, but what you describe has almost never been the case. Aside from the 2014-15 Royals, you'd be hard-pressed to name a team that even ran three relief pitchers out there in well-defined 7th-8th-9th roles, and even they showed more variation than we remember. Bullpen usage is still largely improvisational and can be fiercely complicated by all kinds of factors.

There are a few elements of current pitching usage which seem to eliminate decision making: leaving an ineffective starter in for 90-100 pitches almost no matter what the score (or pulling an effective starter after 90-100, for that matter); always going to the closer in the ninth if a save situation and never if not.

But current bullpen doctrine seems to me driven mainly by one thing: never using a single reliever for so long that he can't be used tomorrow. (Another way of putting cercopith's point.) There are exceptions – sometimes somebody's got to go 3 or more innings in a blowout – but basically the idea is to have as many options tomorrow night as possible. You could argue that that increases the complexity of decision-making.

Just my impression, though – I'd happily hear counter-arguments.
   21. cmd600 Posted: November 12, 2017 at 02:43 PM (#5574706)
How about managers actually manage? Make decisions based on the pitcher's stuff that day, the score, the game conditions, etc.


Umm, thats exactly what we've been doing, and letting managers actually manage has led to an .800 OPS against. "The Plan" that protects managers is to wait until trouble happens, then pull the guy. Trying to predict trouble before it happens, which is the goal here, is what will put guys on the hot seat.

for exactly one inning each, in exactly the same order, to protect any lead from 1-3 runs, regardless of platoon advantage, when the heart of the opponents order is up, the number of pitches they throw, how heavily they've been worked recently, or their effectiveness that day, optimizes either effectiveness or availability.


Where did someone suggest to ignore all this stuff?
   22. PreservedFish Posted: November 12, 2017 at 02:48 PM (#5574710)
It evolved to maximize availability and effectiveness.

I disagree. It evolved to eliminate decision making.


False dichotomy. Both elements are strong influences on today's bullpen strategy.
   23. cmd600 Posted: November 12, 2017 at 02:51 PM (#5574712)
I'm not sure who had options on the Indians but their top 6 relievers by IP gave them 383 innings. The other 7 guys combined for just 91 give or take. As you noted, they got relief innings from some guys who were mainly starters. Anyway, their 489 relief innings were the FEWEST in the AL. They aren't a blueprint for how to take innings away from starters.


Bringing up the Indians was not an argument of "just run your bullpen like this", but as I thought was pretty clear, evidence that its not difficult to turn seven bullpen roster spots into a lot more pitchers. I noted their very few IP, I noted their lack of options, and I noted that they had good relievers that generally didn't get hurt or need a trip to AAA to sort things out. They still found themselves using 13 relievers on the season, not anywhere close to just seven.
   24. cercopithecus aethiops Posted: November 12, 2017 at 03:24 PM (#5574723)
But current bullpen doctrine seems to me driven mainly by one thing: never using a single reliever for so long that he can't be used tomorrow. (Another way of putting cercopith's point.)


Yes. Exactly. Thank you.

And I can't really see any argument that relievers as a group have not become more effective since the advent of one inning per game usage.
   25. BDC Posted: November 12, 2017 at 04:10 PM (#5574733)
Of course I may be blinkered by following a team with a terrible bullpen. The Rangers' most effective reliever, for some time now, has been a guy named Alex Claudio. Claudio is listed as 6'3" and 180 lbs, both being overstatements. He is a left-hander, a sidearmer, and he throws about 45 MPH from a tiptoe start on the 1B side of the rubber. For all these reasons, even though B-Ref lists him as their 2017 "Closer," he has never been simply handed ninth-inning save duties. He doesn't fit the stereotype.

Claudio's 2017 game logs are a mess in terms of strategy. He threw anywhere from 0.1 to 4 innings, on anywhere from 0 to 6 days' rest, in wins, in losses, saving some games, setting up others, doing emergency late relief into a third inning when his manager (Jeff Banister) trusted nobody else at all. Because he throws so softly, he is able to throw a lot and come back soon, though he's not exactly Mike Marshall. Banister may be using him idiotically, but he doesn't use him formulaically, and Claudio has been pretty good (70 games, 83 innings, ERA 2.50 last year).

You must be Registered and Logged In to post comments.

 

 

<< Back to main

News

All News | Prime News

Old-School Newsstand


BBTF Partner

Support BBTF

donate

Thanks to
1k5v3L
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogRyan Thibs has his HOF Ballot Tracker Up and Running!
(37 - 12:55am, Nov 24)
Last: cardsfanboy

NewsblogOT - November* 2017 College Football thread
(198 - 12:21am, Nov 24)
Last: Howie Menckel

NewsblogHall makes room for an "idiot" on the ballot
(25 - 11:48pm, Nov 23)
Last: Chris Fluit

NewsblogOTP 20 November 2017: Sheriff’s official suspended 10 days over Cubs World Series sneak-in
(824 - 11:06pm, Nov 23)
Last: Lassus

NewsblogFavorite Thanksgiving recipes from MLB | MLB.com
(4 - 10:55pm, Nov 23)
Last: Lassus

NewsblogJoe Morgan Wants To Keep Steroid Cheats Out Of Baseball Hall Of Fame | MLB | NESN.com
(181 - 7:55pm, Nov 23)
Last: Bug Selig

Gonfalon CubsLooking to next year
(204 - 7:01pm, Nov 23)
Last: Meatwad in mourning

NewsblogFinding the best spot for Giancarlo Stanton
(12 - 4:18pm, Nov 23)
Last: Walt Davis

NewsblogSources: MLB hammers Atlanta Braves, declaring 12 minor league players free agents
(95 - 3:55pm, Nov 23)
Last: You Know Nothing JT Snow (YR)

NewsblogA Discussion of WAR Wherein I Ardently Attempt to Avoid any WAR-Related Puns | Sports-Reference.com
(60 - 3:27pm, Nov 23)
Last: Blanks for Nothing, Larvell

NewsblogPrimer Dugout (and link of the day) 11-22-2017
(11 - 11:01am, Nov 23)
Last: PreservedFish

NewsblogProspectus Feature: Bill James vs. The Noise - Baseball Prospectus
(13 - 9:39am, Nov 23)
Last: Russ

NewsblogOT - NBA 2017-2018 Tip-off Thread
(1516 - 9:02am, Nov 23)
Last: don't ask 57i66135; he wants to hang them all

NewsblogChambliss Ready to Part With Bat and Ball That Sent Yankees to World Series
(33 - 5:44am, Nov 23)
Last: Omineca Greg

NewsblogShohei Ohtani MLB posting deal reached | MLB.com
(4 - 12:30am, Nov 23)
Last: John Northey

Page rendered in 0.5837 seconds
47 querie(s) executed