The Brewers Used a LOOGY as a Starter
On Monday night in St. Louis, left-hander Dan Jennings received the start for the Milwaukee Brewers. It was notable because, of his 381 prior MLB games, he’d entered every one as a reliever. Outside of a few rehab games, Jennings’ previous 193 appearances in the minors had come in a relief capacity, as well. With the exception of 13 starts at Low-A back in 2008, in fact, Jennings had worked in relief for a decade solid.
Jennings’ appearance as a starter on Monday isn’t notable for what it says about Jennings, though. It’s notable for what it might reveal about the Milwaukee Brewers. Jennings didn’t pitch deep into the game. He didn’t exhibit an expanded repertoire. Instead, as he has on multiple occasions this year, he faced a single batter, recorded a lone out, and then departed the game in favor of a right-hander. Even with the use of an Opener spreading rapidly throughout the game, though, this was rare.
Chase Anderson had originally been on turn to start that night. After a solid 2017 campaign, Anderson has prevented runs well this year, too, recording a 3.93 ERA in 158.0 innings. He’s striking out fewer batters, though, walking more of them, and allowing more home runs. He’s been particularly vulnerable at Miller Park, which is friendly to hitters. As of Monday, he hadn’t pitched more than five innings in his last six starts and compiled only 7.2 innings over his previous two starts combined, putting up a 5.81 FIP and 4.57 ERA since the beginning of August. With one week to go and playoff ramifications attached to every game, Craig Counsell and the Brewers faced the choice of going with Anderson or doing something else. They chose something else.
The team could have opted to pitch Gio Gonzalez on regular rest, but that would have meant using Jhoulys Chacin on short rest the next day or relying on the bullpen to take care of that game. The team opted to get the bullpen game out of the way. The Cardinals’ lineup presented the team with an interesting opportunity. Over the 10 previous games, with 80 starting lineup spots to give out, the club had used left-handed batters in only 15 of them — including eight for leadoff hitter Matt Carpenter, one for Matt Adams, and seven for Kolten Wong, who had been out of the lineup the previous two days with an injury. That meant that, no matter how the Cardinals’ lineup looked on Monday, it was likely to include just one left-hander and that lefty would likely bat leadoff in the form of Matt Carpenter.
Dan Jennings came on to face Matt Carpenter, got him out, and then left the game. It’s not an unprecedented event. A search of the Play-Index at Baseball-Reference reveals 50 instances of a pitcher starting a game, throwing to one batter, recording one out, and leaving the game. An examination of all of the recent instances, however, indicates that all the short appearances were due either to rain or injury. Going back past 20 years or so, a vast majority of the occurrences were for starters who were on their normal schedule, so injury or rain delay seems like a likely occurrence. There were a few notable exceptions, however, where it seems like a one-out strategy was actually utilized on purpose.
- On August 20, 1912, Lefty Schegg of the Washington Senators made his first MLB appearance and one of just two in his career. He started the game and faced one batter, fellow left-hander Buddy Ryan, and retired him. Hall of Famer Walter Johnson came on and finished the game against Shoeless Joe Jackson and the rest of the Cleveland team. The Senators won, 4-2.
- On August 8, 1936, Joe Cronin managed the Red Sox against the Washington Senators. Jack Russell was primarily used as a reliever that season for the Senators and then the Red Sox, but he did start a few games for both clubs. In one such start for the Red Sox against his former team, the right-handed Russell faced righty Ben Chapman to start the game. He was relieved by aptly named Lefty Grove for the next batter, left-handed Buddy Lewis. Gomez was used as a starter nearly all year long and appeared to be on his regular turn to pitch. The Red Sox went on to lose 2-0.
Based on the box scores, it looks like both of those situations are similar to Craig Counsell’s use of Dan Jennings on Monday. Going over some of the other box scores, there are a few other games worth mentioning, including some unusual doubleheader strategies. It’s possible I’m missing some notable appearances, but here a few.
- On April 21, 1921, right-hander Suds Sutherland started the game but pitched to only one batter after Cleveland loaded their lineup with lefties. Of those lefties, two never actually batted and were removed before their turn while a third, leadoff hitter Charlie Jamieson, only batted to start the game before being removed.
- Similarly, on May 26, 1971, righty Al Santorini of the San Diego Padres started the first game of a doubleheader, but pitched to only one Houston Astros batter in a lefty-heavy lineup. Lefty Dave Roberts pitched the rest of the game, but two of the lefties in the lineup never actually came to bat. The Padres lost. Santorini started the second game of the doubleheader, and the Padres lost that one, too.
- In perhaps the strangest attempt to gain an advantage, on May 20, 1950, manager Leo Durocher started lefty-pitching, righty-batting Monty Kennedy in the first game of a doubleheader against righty Ewell Blackwell, but then immediately relieved Kennedy with lefty-pitching, lefty-batting Dave Koslo so that Koslo would have the platoon advantage when batting against Blackwell. The Giants did win 8-0, though Koslo went 0-4 with two strikeouts. Kennedy then went on to pitch the second leg of the doubleheader and the Giants won that game, too.
Relative to how other clubs have used Openers, Jennings’s case is a bit different. Scheduled starter Chase Anderson didn’t enter the game once Jennings left, and Freddy Peralta, who did come in, still had to face the heart of the Cardinals’ order in the first inning. There was also no expectation that Peralta would pitch enough innings to face the top of the order a second time through (but their third time up). He would be relieved before then.
In any case, the move mostly paid off. Carpenter grounded out to second in his first plate appearance, and though he doubled off Peralta in his second at-bat, he was forced to face Josh Hader in the fifth inning and struck out. The Cardinals eventually took the lead against Hader, but that doesn’t necessarily make it a bad strategy.
This situation might have some utility for the Brewers in the playoffs. If the team makes the Wild Card game and faces the Cardinals, then Matt Carpenter is likely to be at the top of the order. If they were to face the Rockies, lefty Charlie Blackmon is likely to lead off. While September’s expanded rosters allow for a greater margin of error, a one-game Wild Card matchup still permits more relievers, as teams need just one starter instead of four in later rounds. The Brewers conducted an interesting experiment somewhat necessitated by a lack of starting options, the unique circumstances of the September roster, and the Cardinals’ lineup.
It’s probably good for the Brewers to experiment with it, to see if they want to do it again on a bigger stage. Their team is likely a bit more comfortable with the strategy, though using it again with one of their regular starters might not be worth the disruption it creates, particularly if the leadoff hitter reaches bases and the scheduled starter is forced to begin his appearance from the stretch. The risk of having a team’s leadoff hitter moved to the second spot in the lineup also mitigates a potential benefit of using a reliever in this manner.
Craig Edwards can be found on twitter @craigjedwards.
While Matt Carpenter does make sense as the guy who bats leadoff given their other options, I think the Brewers just figured out a way to hurt a team when they put a guy who is clearly their best batter in to lead off the game.
Yet another reason for Cards fans to hope that Dexter Fowler returns.
The future is upon us…
The opener strategy keeps making more sense the more it gets deployed:
1. optimal innings distribution – using good relievers early makes sure they don’t run into issues of not pitching for a week
2. reduce top side of the lineup exposure for starters – using ‘higher quality’ relievers to handle the most difficult part of the lineup
3. take advantage of platoons
4. reduce starter overall workloads
5. increases the value of ‘two-pitch starters’ – guys who can go once through the lineup every 3 days, and can pitch more than one inning (potentially another undervalued segment of pitching talent)
Does this also…
…allow for maximum leverage application relative to WPA?
What else?
I think all of those points but the first are valid.
You wouldn’t want to use a god reliever just to use him. The reason he won’t pitch a week is due to lack of close games, so burning him early ‘for work’ isn’t a good idea in case that game was the one you need him late. Week-long breaks for relievers aren’t planned or foreseen.
Well, the first is valid too in that if the guy’s gone 3-4 days without pitching, you can now make sure he gets some work. And the first-inning leadoff hitter is actually somewhat a high ‘runs’ leverage hitter, what with the good hitters all coming up behind him. Very good chance you give up the run if he gets on.
Oh I agree with the potential leverage use. But using him purely because it’s been a few days is too narrow-focused.
No team should use an opener purely because he needs work after not pitching for a period of time. The only reason an opener should be used is if the team believes the benefits (increased odds of suppressing the best opposing hitters, getting more out of your following pitcher, etc.) outweigh the costs (using a good reliever early instead of late).
Except teams *do* do this. A closer that hasn’t pitched in four days gets used in a non-save situation just to get work in, even if it risks him throwing 30 pitches and being unavailable for the next day.
That’s exactly my point. STARTING the game purely to get work in runs the risk of needing him later in the game. But LATER in the game, if they need him they have him, and if they don’t need him, they’ll use him to get work in.
If you assume the game will have be close and you’d use your closer, then inning ONE is actually the most important inning, it’s got the leadoff man coming up first and the batters in order as picked by their manager.
Using the closer then may be the REASON you don’t need him later, or the only reason it is close enough to “need” him later. Not using him then may cause you to “need” him later.
That’s factually wrong. The first inning in a close game won’t have the higher leverage value…
I guess people don’t get it. There’s zero reason to use a reliever to start the game purely to get work in. If a matchup favors it, sure, but to get work in REGARDLESS of matchup is idiotic, as it burns him for later in the same game if you need him.
The most efficient use of a “closer” is in the highest leverage moments, given constraints of resource availability (current pitcher who is being removed and who is available in the bullpen). From a derivatives standpoint, a “high gamma” game where there’s a decent chance of one of the teams scoring runs quickly means that the leverage expectancy is low later in the game. Hence the use of a “high quality” opener similar to what we’ve seen with the Rays.
Not purely based on getting work in. See comment below.
I don’t care about the below comment. I’m replying to people who disagree with my comment, which specifically says ‘just to use him’ as in just to get work in, all else being equal.
Other benefits make it worth it! I agree. But I was addressing the original comment’s first point in isolation, which based on the reactions to my comment, readers didn’t realize.
Optimal distribution of innings/work over time for the pitcher is something that should be of concern and is likely a point of study at greater detail than our regular anecdotal information, where we “know” a guy can’t work 3 days in a row, for instance, and often the pitcher is less effective on Day 2. Distribution matters.
As baseball evolves on the opener strategy, one would suspect the shuttling of pitchers back and forth between the minors will become more and more common to offset the up tick in innings from the bullpen. (All the reason more to have your Triple A team close by). To expand on bulletin 5, I would add 2 pitch pitchers with option years will be in demand by analytically inclined teams this off-season.
As a baseball fan, this is within the confines of the law, but makes me feel dirty.
I think it’s funny, and frankly misguided, to term relievers as higher quality pitchers than starters. Relievers are largely, if not entirely, failed starters. Not entirely sure if you were making this point, but I’ve heard it made elsewhere.
While I totally see the merit of this strategy in a vacuum (or Wild Card Game) or over the course of the season, I still think the development of durable, effective starting pitching that is capable of pitching deep into games is necessary to win championships.
As we’ve seen bullpen usage expand in general, we’ve seen the warts of the “higher quality” relievers exposed in the playoffs, and managers haven’t adjusted. Even the best relievers aren’t capable of pitching 4-5 times in a given series, often for more than an inning at a time, and maintaining the same effectiveness. We can debate whether or not that is due to durability or opponent familiarity, but the reduced effectiveness in long series and deep playoff runs has been evident.
I definitely think there’s something to using “co-starters”, pairing together two of the “two-pitch starters” you’ve identified. I know this isn’t exactly breaking ground. The Astros employed the strategy during their run when their own bullpen proved ineffective, and we’ve seen the Dodgers employ at times, as well.
Higher quality is only in reference to the run expectancy. But understood on the point…need to frame that aspect more delicately.
Not arguing the efficacy of the durable, high quality starter either. The truth is a team plays about 1450 innings a year, and you want to manage those innings in the most efficient manner possible, and top tier starters help rebalance the shared workload of the rest of the staff.
When referring to run expectancy, you’re quite correct, and I hope you did not take my response as derisive.
As a relative traditionalist, I’m averse to the idea that openers can be deployed exclusively and successfully by championship teams, which wasn’t implied by you, but I have heard implied elsewhere. That said, I completely see the merit in deploying openers strategically throughout a season, and I think it will be important for teams to develop more multi-inning relievers and / or co-starters for the exact reason that you noted, as it will help maximize their players’ talent and supplement more traditional workhorses.
They’re all pieces of the puzzle in the evolution of the modern day pitching staff.
https://baseballanalyzed.com/2018/04/07/the-future-of-pitching/
I explained how it works here.
The way it works is by using multi inning relievers that are used based on matchups and righty/lefty splits. Closers don’t exist in this strategy and neither do starters. It benefits pitchers in every way, from their health to not having to go through a lineup more than once.
I don’t understand why the Cardinals don’t counter that move by bumping Carpenter down in the order. Lead off with Munoz or even Martinez. Bat Carp third. He’s been in a skid anyway it’s not like he could do any worse.
I guess the procedures of lineup cards matters here, right? When did Counsell officially have to announce his starter, and would it give Shildt enough time to make a change?
I’ve always figured they do so simultaneously, when they hand them to the home plate umpire? At which time your ‘starter’ now has to pitch to at least the 1 batter. So if Shildt (without announcing so beforehand, of course) switched Carpenter down, any Brewer starting LOOGY would now have to pitch to the one rightie.
Yes, I was wondering this as well. Whether or not there was a way to change your lineup after you see the starter, and then the other team can change the starter, and so on — there has to be somebody going last. But if they are simultaneous, how could the Brewers be sure that a lefty would bat lead-off? Or were they just going with the percentages?
In terms of this being an “experiment … to see if they want to do it again on a bigger stage”, Monday night was a darn big stage itself. It’s the last week of a pennant race in which one of the teams will be pretty much knocking out the other. Which looks like the Cardinals as of this moment.
Suds Sutherland? Where is notgraphs
My Jack Russell would certainly not have the patience to face more than 1 batter
I keep trying to delete this copy, but it refuses to.
Where I mildly disagree with Craig is “if the leadoff hitter reaches base and the scheduled starter is forced to begin his appearance from the stretch.” Can’t see where this is any kind of a big thing. The 35% of the time the leadoff gets on, the starter is now pitching from the stretch that handful of pitches in.
I think we should remember that this is a September game and the Brewers have something like 18 pitchers on their roster right now. This is not as likely to happen when rosters are not as large and really I think this is the kind of stuff that, although somewhat interesting, is what some might use to argue against the September roster expansion as it currently stands. Based on what I am seeing with 8 – 10 pitchers in a game I do think something should be done to keep the numbers in the bullpen down. I like the idea of a team declaring 25 (or 26) player rosters for a series. it gives you a couple extra guys in the bullpen and maybe an extra bench guy. I see no reason to have the game played like a Spring Training game when they actually mean something.
As far as using this in the WC game, John Smoltz has been saying this for a few years now whenever I see him announcing or on MLB Network. The WC game should be a bullpen game unless you have a Sale or Scherzer to throw out there. It’s an elimination game so it should be all hands on deck to start with and why throw your ace out there for 6 or 7 innings? It isn’t helping win this 1 ballgame and greatly hurts your chances of competing in the Divisional Series if you do win.
I think the WC game is going to be a bullpen game for the foreseeable future. Like I said, if you have a Sale or Scherzer or someone of that ilk, then yes, you are probably best served by having them go out and guarantee a W, but beyond that it’s going to be a stream of 1 or 2 inning relievers.
“Freddy Peralta, who did come in, still had to face the heart of the Cardinals’ order in the first inning. There was also no expectation that Peralta would pitch enough innings to face the top of the order a second time through (but their third time up).”
Huh?
If the Dodgers make the playoffs, there are opportunities for teams like the Brewers to really screw with Dave Roberts. As a slave to the platoon advantage and back to being limited to the playoff 25 man roster, a smart manager could nearly paralyze Roberts.
Piggyback starters.
One, say left-handed, starts and goes 3-4 innings.
He’s relieved by an opposite-handed pitcher, who goes 3-4 innings.
Relievers pitch the last few innings.
Almost impossible to platoon.
The Brewers didn’t use a LOOGY as a starter. They used to LOOGY as a LOOGY to start the game.
And, no, I don’t know why we drive on a parkway and park on a driveway.
This is really a by product of a September 40 man roster where using a pitcher for just one batter in the first inning is possible due to 15 additional players sitting on the bench.
Unlike making rules to ban the shift, this is a case where baseball should step in and allow teams to have a 40 man roster in September but make the manager designate which 25 players can be active each day.
If you are going to do that, make him include the four guys who pitched the most innings in the last 4 games, otherwise, you’re still giving him four extra roster spots.
Good point
In the article, Craig confuses Lefty Gomez with Lefty Grove. It was Grove who pitched in the August 8, 1936 game for the Red Sox.
Thanks. Lefty Gomez and Grove both pitched one batter, one out games and I wrote the wrong one down.