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, October 20, 2018

The Brewers are becoming more and more positionless on defense | SI.com

Down is up and up is down.

Jim Furtado Posted: October 20, 2018 at 07:38 AM | 23 comment(s) Login to Bookmark
  Tags: brewers, strategy

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. Kiko Sakata Posted: October 20, 2018 at 12:11 PM (#5771702)
Interesting article. This and the "opener" - or even just "bullpenning" in general - have the potential to really screw up some player metrics. I have my own metric where I tie my comparisons to "positional averages".

One of the things they discuss here, for example, is switching players within the same infield based on the hitter. There's of course, the "classic" shift where the third baseman plays a variant of second base. This article is about the Brewers, but the Cubs at least once played Javy Baez and Tommy LaStella at second and third base, with the two of them switching positions based on the hitter (the goal being to maximize the number of balls hit to Javy / minimize the balls hit to Tommy). But comparing against positional averages (which is essentially what WAR does too), the baseline against which Baez and LaStella get compared would keep alternating - and, in the case of evaluating their HITTING relative to a 2B/3B, it alternates essentially depending on who the last batter was in the previous defensive half-inning, which doesn't necessarily make a whole lot of sense.

The problem is much worse, though, on the pitching side. Most metrics - including mine (and, I'm pretty sure, both versions of WAR) - rate starting pitchers against a lower baseline, because, on average, starting pitchers do a little worse in terms of outcomes than relievers, because they have to pace themselves and are more likely to face the "times-through-the-order" penalty. But, at least in my case, the "starting pitcher" is defined - quite logically - as the first pitcher of the game. But in, for example, Game 5 of the NLCS, the pacing/"times-through-the-order" penalty clearly affected Woodruff (5.1 IP, 22 batters faced), not Miley, who just faced one batter (and, one presumes, knew that was his role entering the game).

And, then, to the extent these are new strategies, should whatever adjustments we make to deal with these in evaluating 2018 (and 2019, etc.) be applied retroactively in evaluating history, when these roles were somewhat more set? Although, in some ways, I wonder if the starter/reliever roles were as set historically before the 1960s or 1970s when starting rotations were somewhat more fluid and a lot of teams' top relievers were just their starting pitchers on their off days.

Anyway, it's an interesting challenge. I'd started thinking through the issue with pitchers, but this article raises the interesting issue that it's worth thinking about the same thing with respect to position players as well.
   2. Howie Menckel Posted: October 20, 2018 at 12:52 PM (#5771707)
the only thing I like about the shift is that it confounds - or should confound - those who, with religious zeal, bought into the idea that defensive metrics had become as reliable as offensive ones.

this would come up in MVP discussions, for instance, here and elsewhere, as many in the audience seemed to see no need for discussion. the "BUT THE WAR SAYS!!!!" people.

if the WAR isn't over, it's at least tending toward peace in our time.
   3. . Posted: October 20, 2018 at 01:53 PM (#5771723)
Oh, goodie, baseball players don't have positions anymore!!!

Sensation!!!
   4. Greg K Posted: October 20, 2018 at 02:09 PM (#5771728)
Didn't Casey Stengel famously move guys all around the field?

I feel like strict positioning is more a feature of the modern game, and this is a bit of a throwback.

The most common configuration gives second base to Shaw and shortstop to Orlando Arcia, whose defense is sharp enough to keep him in the lineup despite his featherweight bat. But Schoop will often take over second against left-handed pitching, and sometimes, he’ll play at short. (He’d played just seven games at the position in six years in the major leagues before being traded to Milwaukee.) There’s also Hernán Pérez, who can grab either spot up the middle. Most often, they’ll play Aguilar—Shaw—Arcia—Moustakas. But it’s not uncommon to see Aguilar—Schoop—Perez—Moustakas, or Aguilar—Shaw—Schoop—Moustakas, or Aguilar—Schoop—Perez—Shaw, or ... you get the picture.


This sounds like a return to the way baseball used to be played, rather than the 21st century model of having one guy at every position, a back up catcher, and 18 bullpen arms.
   5. caspian88 Posted: October 20, 2018 at 02:56 PM (#5771734)
The 1956 New York Yankees had a number of players play regularly at multiple positions:

- Gil McDougald started 88 games at shortstop, 24 at second base, and 4 at third base.
- Billy Martin started 101 games at second base and 11 at third base.
- Jerry Coleman started 27 games at second base, 18 at shortstop, and 6 at third base.
- Elston Howard started 50 games in left field, 19 as catcher, and 1 in right field.
- Joe Collins started 25 games at first base, 20 in left field, 16 in right field, and 1 in center field.

Of course, a lot of this was probably because the Yankees were carrying ten or so pitchers at a time (eight pitched in the World Series), and a bunch of these guys (and other part-timers and platoon halves) just didn't hit.

Their World Series opponents, the Dodgers, had some similar action:

- Jackie Robinson started 71 games at third base, 22 at second base, 9 at first base, and 2 in left field.
- Jim Gilliam started 98 games at second base, 49 in left field, and 5 in right field.
- Gil Hodges started 121 games at first base, 22 in left field, 8 in right field, and also caught two innings in relief.
- Pee Wee Reese started 134 games at shortstop and 9 at third base.

However, the backups were mostly single-position players (Walker, Nelson, Neal, Fernandez). Gino Cimoli played all three outfield positions, but Hodges Junior Gilliam were really the Dodgers' primary backup outfielders. Sandy Amoros was strictly platooned (only 14 plate appearances against left-handed pitchers all season).

Then again, the 1956 Dodgers saw only 651 plate appearances against left-handed pitchers all season, in 49 games. What was up with that?
   6. vortex of dissipation Posted: October 20, 2018 at 03:00 PM (#5771737)
Didn't Casey Stengel famously move guys all around the field?


He certainly did. Taking the 1957 Yankees at random, this is what his games played for his infielders looks like (games played at 2B-SS-3B):

Bobby Richardson - 93 games at 2B
Andy Carey - 81 games at 3b

then the fun begins...

Gil McDougald - 21-121-7
Jerry Coleman - 45-4-21
Tony Kubek - 1-41-38 plus 41 in LF and 22 in CF
Billy Martin - 13-0-26
Jerry Lumpe - 0-6-30
   7. Kiko Sakata Posted: October 20, 2018 at 03:21 PM (#5771740)
Then again, the 1956 Dodgers saw only 651 plate appearances against left-handed pitchers all season, in 49 games. What was up with that?


All of their big hitters, except for Duke Snider, were right-handed hitters: Campanella, Hodges, Robinson, Furillo, even Pee Wee Reese (Gilliam was a switch hitter; earlier in the 1950s, their third baseman was righty Billy Cox). I think Ebbetts Field might have also favored right-handed hitting, but I'm less sure about that.

Pitching rotations tended to be more fluid back then, so it was easier for teams to avoid unfavorable matchups. Warren Spahn, for example (certainly the best left-handed starter of that era), started 665 games in his career, all in the NL, most when there were only 8 teams (i.e., 7 possible opponents). He started against the Dodgers only 66 times, versus 74 (Cubs) to 125 (Cardinals) starts against the other "original 8" teams. And he started in Ebbetts Field only 18 times in his career (vs., for example, 39 starts at the Polo Grounds; he started at least 34 times in all of the other road ballparks in use in the 1950s). In 1956, Spahn pitched 281.1 innings; ZERO of them in Ebbetts Field!
   8. Howie Menckel Posted: October 20, 2018 at 03:25 PM (#5771742)
the 1956 Dodgers saw only 651 plate appearances against left-handed pitchers all season, in 49 games. What was up with that?

that was part of a league-wide bid to get Duke Snider into the HOF.
:)

career vs RHP: .949 OPS in 7067 PA
career vs LHP: .743 OPS in 1170 PA

1956 vs RHP: 1.036 OPS in 583 PA (42 HR)
1956 vs LHP: .677 in 69 PA (1 HR)

more seriously, look at the rest of the lineup
EDIT: ah, mostly Coke-d by Kiko
   9. Kiko Sakata Posted: October 20, 2018 at 03:41 PM (#5771745)
Per #8, Duke Snider:

career vs RHP: .949 OPS in 7067 PA
career vs LHP: .743 OPS in 1170 PA


versus Eddie Mathews, also a left-handed batter, roughly contemporary, didn't play for a team that was quite so righty-heavy:

career vs. RHP: .927 in 7469 PA
career vs. LHP: .741 in 2282 PA

Basically, Mathews and Snider were about the same as hitters, but Snider got the platoon advantage 85.6% of the time. Mathews only had the platoon advantage 76.6% of the time. Result: Snider's career OPS was .919 vs. .885 for Mathews (park effects and the fact that Mathews played a bit more into the low-scoring 1960s gives him a career edge in OPS+ over Snider 143-140)
   10. Ray (CTL) Posted: October 20, 2018 at 03:44 PM (#5771747)
Interesting discussion. I actually think long term the pitchers will be "roleless." Basically just having pitchers go short or long with no regard for a plan other than game situation and recent-game usage.

I could see us there within 5-10 years, actually. This is moving fast.
   11. . Posted: October 20, 2018 at 03:54 PM (#5771748)
Guys playing games at multiple positions isn't what "positionless" is all about.
   12. cardsfanboy Posted: October 20, 2018 at 04:07 PM (#5771752)
Basically, Mathews and Snider were about the same as hitters, but Snider got the platoon advantage 85.6% of the time. Mathews only had the platoon advantage 76.6% of the time. Result: Snider's career OPS was .919 vs. .885 for Mathews (park effects and the fact that Mathews played a bit more into the low-scoring 1960s gives him a career edge in OPS+ over Snider 143-140)


Mathews also had about 2000 more pa than Snider, and yet both of them are what you would have to call 'clear' hofers, and yet it took Mathews 5 years to get in, and Snider 11...(Mathews has to be the biggest embarrassment for the writers though, the guy has 96.6 war and got 32% on his first ballot. )
   13. caspian88 Posted: October 20, 2018 at 04:24 PM (#5771755)
Eddie Mathews also had 512 home runs (tied for 9th all time with Ernie Banks in 1974), 1453 RBI, 1509 R, and 12 All Star appearances.

I guess the .271 batting average was a real problem for BBWAA members in the mid-1970's. Ernie Banks and his .274 batting average didn't have the same problem just a couple years later, but Banks was a two-time MVP and Mr. Cub.

Mathews wasn't elected until the year after Banks.
   14. BDC Posted: October 20, 2018 at 04:35 PM (#5771756)
I guess the .271 batting average was a real problem for BBWAA members in the mid-1970's


I think so. When he retired, Mathews was 6th in career HR, 21st in career RBI, 10th in career walks, and (using a 5000-PA minimum, ten years of full-time play) … tied for 272nd in batting average with Elbie Fletcher, Max Bishop, Billy Werber, and Burt Shotton, among others.

When Mathews was inducted in the late '70s, the only (non-pitcher) HOFers with lower batting averages were Tinker & Evers, Bobby Wallace, Rabbit Maranville, and Ray Schalk.
   15. cardsfanboy Posted: October 20, 2018 at 04:59 PM (#5771757)
The thing about Mathews to me is that he had over 10,000 pa... there are 28 players not in the hof with over 10,000 pa... among that list of 28 is a isht ton of deserving hofers(base on merit) Rose, Bonds, Jeter, Arod, Beltre, Palmeiro, Pujols, Beltran, Dahlen, Ichiro... along with another group of borderlines Nettles, Dwight Evans, Sheffield, Abreu, Damon, Ortiz, McGriff, etc... ultimately though is Mathews is so clearly beyond the line, that the .271 average is a disgraceful way to look at his candidacy, and insulting to people who are supposed to know baseball that he didn't make it first ballot...
   16. Rough Carrigan Posted: October 20, 2018 at 06:00 PM (#5771765)
#10. If you look at the stats of pitchers from the 1920's through the 1960's, you see a lot of good pitchers with 38 games pitched including 30 starts and 8 relief appearances.
   17. Howie Menckel Posted: October 20, 2018 at 07:26 PM (#5771774)
another commonality, sort of:

Mathews was a beast thru age 31, then settled for OPS+s of 112, 126, 108, 111, and a 98 in just 57 PA for the Tigers at age 36. oddly, that year Eddie was 0 for 5 - all PH appearances - in the first 3 weeks of the season. a herniated disk basically cost him the entire summer, though he returned in September and was a member of the Tigers' World Series squad. (he played for the Astros for much of 1967, which was weird.)

Snider's last full-time season ended just days after he turned 31 - and just before the Dodgers hightailed it to LA. but in a twist, Duke kept his OPS+ around his career level for five years, with PAs of 365, 435, 285, 266, and 196. then came a 115 in 415 PA for the Mets (!). Duke, too, closed his career in a second weird uniform - the SF Giants (!), for a 75 OPS+ in 189 PA at age 36 (until he turned 37 a week before the regular season ended).

they each built up such momentum for HOF in their 20s that they didn't have to be big stars in their 30s.
   18. Walt Davis Posted: October 20, 2018 at 07:44 PM (#5771779)
The "positionlessness" of the Brewers seems to be vastly overstated. It mainly amounts to this:

1. The SS can't hit but can field so, like many teams before them, they sometimes play a guy who (maybe) can hit but isn't as good a fielder there.

2. The innovative part is "we shift a lot so we've seen Shaw play second base a lot and we decided he could handle it ... and our current 2B weren't hitting." With that decision, it cleared the way to acquiring Moose.

The question was/is what effect would this have on their defense. When the trades went down, the media, nerds, most here assumed Schoop was going to get a lot of time at SS which seemed questionable. But Schoop got only 11 starts there in two months, 8 of those in Aug, none since Sept 18 and none in the playoffs. They've pretty obviously decided that didn't work. Unless somebody can show that there are heaps of times he's the "2B" but lines up at "real SS" (not just a step on the other side of 2B in a RHB shift).

So the primary thing is that they discovered that Travis Shaw is Don Money (and not Bill Madlock/Ron Santo).

And as Kiko points out, Maddon's done this with the Cubs for the last 4 years. We don't shift as much (near the league bottom) so I suppose we have less crazy configurations but:

Javy 2B/SS/3B
Zo 2B/LF/RF
Bryant 3B/LF/RF
Happ LF/CF/RF and tiny bits of 2B/3B

And this year, Bote at 2B and 3B; previous years Contreras at C/LF and a bit of 1B; and they turned Schwarber into a LF. Definitely affected by Bryant's injuries but the Cub leading GS at each position:

C: Contreras 123
1B: Rizzo 148 (4 others with GS)
2B: Javy 75 (Zo 42, Murphy 32, 2 others with GS)
3B: Byrant 79 (Bote 42, 3 others with GS)
SS: Russell 110 (Javy 52, Bote 1)
LF: Schwarber 115 (4 others)
CF: Almora 94 (Happ 49, Heyward 20)
RF: Heyward 91 (Zo 52, 2 others)

The Cubs only had 15 position players who started a game all year -- the original 12 plus Caratini (replacing Gimenez), Bote (mostly for Bryant's injury) and Murphy (Aug pickup). The extra guys who started at 1B, 3B, LF weren't Sept call-ups or emergency call-ups, they were (except for Bote) established MLers (or in the case of Caratini, our semi-intended backup C).

Or put differently, the following Cubs got a substantial number of their starts at what would not be considered their main position -- plus Zo who I'm not sure he has a main position at this point.

Caratini -- 31 at C, 9 at 1B
Javy -- 75/52/18 at 2B/SS/3B
Zo -- 52/42/10/3 at RF/2B/LF/1B
Happ -- 49/27/12/12 at CF/LF/RF/3B
Bryant -- 79/10/8/1 at 3B/LF/RF/1B (not quite 20%)
Heyward -- 91/20 (also not 20%)

This isn't just swapping corner OFs or having a back-up guy float through the IF. Zo, Happ and Bryant each received double-digit starts in both the IF and OF. Javy, a "full-time" starter by Cubs standards, has more starts at SS/3B than 2B over the last two years.

And those are just starts. With Joe's managering, the Cubs shift guys constantly throughout a close game. Javy appeared at a position 192 times. Zobrist 171, Happ 170 (not incl P, in 128 games in the field). So about 25-30% of the times they played, those guys played 2+ positions in that game (not including PH,DH,P).

But I doubt any of that is all that uncommon -- that many guys splitting IF/OF might still be. With the Cubs, about all you could be sure of was that Rizzo would be at 1B and, if he played, Russell at SS and Contreras at C. For the Brewers I'm not sure it was much more than Cain in CF, the Cs at C, Arcia (when he played) at SS and eventually Aguilar at 1B.

For the Dodgers, I don't think they ever had a set OF, Turner was at 3B when he played, the Cs at catcher and that was about it. The Dodgers started 8 different guys at 2B including 13 for Muncy (!) and 6 for Barnes (!) Muncy also grabbed 30 starts at 3B and 58 at 1B. C'mon, Max Muncy is a much better example of "positionless" plus "punting defense for offense" than anything the Brewers did. In CF, their leading starters were Bellinger (50), Hernandez (42) and Taylor (32) -- in terms of how they were developed in the minors, that's a "1B," a "2B," and a "SS" playing supposedly the most defensively important OF position. And per TZ and DRS, that trio were a bit above average with both metrics also agreeing that the worst CF they used was the "natural" CF Joc Pederson.
   19. caspian88 Posted: October 20, 2018 at 07:51 PM (#5771780)
Of course Maddon has been moving Zobrist around for decades, it seems.
   20. Howie Menckel Posted: October 20, 2018 at 08:02 PM (#5771783)
The Dodgers started 8 different guys at 2B including 13 for Muncy (!) and 6 for Barnes (!)

Barnes also started 7 other games at 2B in 2015-17, for 13 out of his 114 MLB starts (the others being at C). he has 161 career minor league starts at 2B, vs 254 at C.

and yeah, while Craig Biggio just shrugs, it seems weird to the rest of us
:)
   21. Walt Davis Posted: October 20, 2018 at 08:05 PM (#5771784)
On one of Kiko's other points -- clearly if "openers" become a thing, then start/relief splits (as currently defined) become problematic. They also pose some problem for official scorers and it might be best to wait to see how they decide to handle it. We might have to shift towards splits based on number of batters faced -- <8, <18, 18+ or something ... possibly also some inning thing in there. That also obviously impacts on things like quality starts and obviously pitcher wins (why the scorers may have to change the rules on that).

As to WAR, until there is genuine "bullpenning" nearly all the time, there will still need to be some "role" component. Ryan Yarborough maybe gained some small advantage relative to a "real starter" but he still was expected to go 4-5 innings most of the time so he still had to pace himself. That also suggest a batters faced criterion of some sort. There's a disadvantage with regard to "theory" in that using BF means that the "role" is determined after the performance ends while you'd rather have it determined before but the practical impacts of that might be minor.

As to applying whatever criterion is settled on to past eras -- I wouldn't bother. It hardly seems worth the effort plus there's no reason to think that we won't see further changes to usage that would make the new criteria obsolete and you'd have to do it all over again. Have a role component, apply the best criteria you can cook up for each given "era" to determine that role component, let those criteria shift. But it's baseball stats and somebody will do it eventually which is fine even it if shows it makes no real difference.
   22. Walt Davis Posted: October 20, 2018 at 08:50 PM (#5771805)
Well, those bozos couldn't even get Santo over 5% and still mostly refused to vote for him when given a second chance ... and then the VCs couldn't be bothered to elect him either so it's not really a surprise that Mathews got limited love. The voters really struggled with 3B until Brett, Schmidt, Boggs.

Part of that was the "early" collapse thing. It seemed like career length wasn't viewed as PA or even years played necessarily, it was how old were you when you retired. Certainly in the 90s and I suspect even the 00s, I'd read writers referring to Santo's "short" career. It was 9400 PA, >2200 games, >2100 starts at 3B -- that ain't short. He was highly durable so in his case it was "just" 15 seasons, the last one barely and he was done before 35. But his 13 full seasons averaged 156 games (so 20131 total).

Mathews was obviously vastly superior to Santo but he had much the same problem. He got a good bit of AS love and two #2 MVP finishes but he never drove in 100 after 28, he hit 30 HRs only once after 29. He had only 16 points of black ink and most of that was in stuff that didn't matter at the time -- he led the league in HR twice and that's it.

He received a big boost when Banks came on the ballot, apparently the first time a lot of the writers noticed the 512 in Mathews' HR column, and a slightly bigger boost to election after Banks was elected. He had originally entered the ballot the same year as Mantle which didn't bode well for his 1st ballot chances but it didn't excuse his relatively low debut. It had taken a LOT of ballots for Ralph Kiner but it's still embarrassing to the BBWAA that on the two ballots they shared, Kiner received way more votes. (Gil Hodges too but you had the Mets managerial gig and his death in 1972 factored into his support.)

But y'know, they'd just finished never getting Johnny Mize to even 45% in his 15 years the ballot before Mathews so at least they were being consistent. And they would soon be embarrassing themselves with Santo and Allen not making 5%, and it took Killer 4 ballots and they've continued this fine tradition with 4 ballots for Piazza and 7 for Bagwell. It's fortunate that Banks entering the ballot helped Mathews (as Eck helped Gossage and Sutter) rather than hurting him (as the 300-game winners coming on killed Bunning).
   23. Greg Pope Posted: October 21, 2018 at 10:26 AM (#5772019)
There's a disadvantage with regard to "theory" in that using BF means that the "role" is determined after the performance ends while you'd rather have it determined before but the practical impacts of that might be minor.

This is what leaped to my mind regarding this. For example, last night the Brewers brought Hader in the third. We don't know how much he knew about how he was going to be used. Did Counsell say "You're going 3, so pace yourself" or maybe "You're giving me 35 pitches tonight". Or even "Throw as hard as you can, I'll take you out when you're tired".

Should the baseline be batters faced? Number of pitches? We know that 25 pitches in 1 inning is more stressful than 26 over 2 innings, so you may need an end-of-inning factor in there.

Will it make any difference?

You must be Registered and Logged In to post comments.

 

 

<< Back to main

News

All News | Prime News

Old-School Newsstand


BBTF Partner

Dynasty League Baseball

Support BBTF

donate

Thanks to
BDC
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogRed Sox owe $12 million in luxury tax, showing why they'd want to shed payroll
(9 - 4:18am, Dec 17)
Last: Scott Ross

NewsblogWhy gambling used to scare baseball and why it doesn’t anymore
(60 - 2:56am, Dec 17)
Last: yest

NewsblogMLB: Mets to sign catcher Wilson Ramos
(16 - 12:51am, Dec 17)
Last: Howie Menckel

NewsblogOT - Catch-All Pop Culture Extravaganza (December 2018)
(629 - 12:36am, Dec 17)
Last: Howie Menckel

NewsblogTHE HALL OF FAME VALUE STANDARD (Bill James rank 25 worst players in HOF & 25 best not in the HOF)
(95 - 11:34pm, Dec 16)
Last: Sunday silence

NewsblogNY Post: Lenny Dykstra reveals his new life as an amateur Torah scholar
(4 - 11:25pm, Dec 16)
Last: Sunday silence

Hall of Merit2019 Hall of Merit Ballot Discussion
(354 - 9:53pm, Dec 16)
Last: Howie Menckel

NewsblogOT: Soccer Thread (The Berhalter Thread?)
(206 - 9:46pm, Dec 16)
Last: Mefisto

Hall of Merit2019 Hall of Merit Ballot
(16 - 9:34pm, Dec 16)
Last: Kiko Sakata

NewsblogJeurys Familia finalizes 3-year deal to return as likely setup man
(17 - 9:32pm, Dec 16)
Last: Russlan thinks deGrom is da bomb

NewsblogOT - NBA Thread (2018-19 season kickoff edition)
(3709 - 9:20pm, Dec 16)
Last: tshipman

NewsblogThibs' Hall of Fame Tracker
(502 - 7:18pm, Dec 16)
Last: SoSH U at work

Newsblog2019 BBWAA HALL OF FAME BALLOT
(130 - 2:19pm, Dec 16)
Last: snapper (history's 42nd greatest monster)

NewsblogCardinals prospect Griffin Roberts suspended 50 games for a drug of abuse
(20 - 1:57pm, Dec 16)
Last: snapper (history's 42nd greatest monster)

NewsblogCampaign to make Negro League bobbleheads
(4 - 1:31pm, Dec 16)
Last: Sunday silence

Page rendered in 0.3724 seconds
46 querie(s) executed