How the “Opener” Spread to the Dodgers
PITTSBURGH — Dodgers reliever Scott Alexander had just finished his lunch and was walking down the 16th Street Mall in Denver last Friday when he received a text from Dodgers manager Dave Roberts.
“‘Hey, you’re getting your start today,”’ the text read, as Alexander remembers it. “‘One or two innings.’”
Alexander had not regularly started professional baseball games since he was in Rookie ball with the Royals in 2010.
The left-hander had watched with curiosity last month as Rays reliever Sergio Romo started back-to-back games for the Rays, ushering in a new label, “the opener,” and a new game strategy. And on that Friday at Coors Field, the movement spread to the Dodgers and the NL West, as Alexander pitched the first inning of an 11-8 win over the Rockies. The Dodgers employed the strategy again yesterday in Pittsburgh when Daniel Hudson started against the Pirates.
After learning of what the Rays were doing with Romo, Alexander approached Dodgers bullpen coach Mark Prior in the bullpen during a May 28 game at Dodger Stadium. There Alexander “half-jokingly” broached the idea with Prior, saying he would be open to “opening” for the Dodgers.
Alexander knew the Dodgers were hurting, with Clayton Kershaw (back) soon to be back on the disabled list and the combination of Rich Hill (back), Kenta Maeda (hip strain), and Hyun-Jin Ryu (groin) already housed there. He knew the Dodgers were short on starters. He knew they were leaning more and more on inexperienced arms, that they might want to shield their young starters from having to work through a lineup three or more times. He knew the Dodgers prided themselves on finding solutions, even if they were uncommon ones.
Number of starting pitchers used by Dodgers:
2018: 11 (so far, including Daniel Hudson today)
2017: 10
2016: 15— Bill Shaikin (@BillShaikin) June 7, 2018
“I had never seen it before,” Alexander said of Romo starting. “I told [Prior] that I could do it if they needed me to. I knew that we were maybe thinking about experimenting and we were calling up some guys with the injuries… I didn’t think too much about it.”
Alexander may have not thought too much about his proposal, but Prior followed up their conversation by relaying it to other Dodgers personnel.
“I talked to [Roberts],” Prior told FanGraphs. “I said ‘Hey, [Alexander] is definitely open-minded.’”
Prior said it was brought up again during a coaches’ meeting.
“We brought it up there,” Prior said. “He’s a ground-ball pitcher, and we’re in Colorado… Everyone kind of kicked it around. ‘OK. Sounds good. Let’s go.’”
That the Dodgers are copy-catting a strategy more or less introduced by the Rays a couple weeks ago shows how quickly an idea can spread in Major League Baseball. It’s certainly not an isolated case. Consider how substantially hitting philosophy and the emphasis on launch angle has changed since Statcast started measuring it. Consider how quickly shifts proliferated before peaking last season. Teams are not only turning to their bullpens more often but are now moving away from traditional roles.
But it’s how this idea spread to Los Angeles that is of interest to this author.
“We weren’t thinking about it,” said Prior of the club considering using an “opener” up until that point. “We’ve always thought about [Alexander] as more than just a lefty specialist. We’ve thought about him going multiple innings because he can be efficient. It just so happened that we needed someone that day, and given Colorado’s lineup with the lefties at the top, it made sense to get him through the fourth or fifth hitter and then go to someone else.”
The spread of the opener to Los Angeles was not a top-down directive from an analytically minded front office; rather, it was player-driven. It was an idea that trickled from the bottom-up. What does that say about how communication is evolving within major-league teams?
“I kind of view it as more of a collaboration or partnership,” said Prior of interacting with players. “My specific job is to make them feel good, feel confident, to try and help them. I don’t think it’s the [role] of me, a coach, to tell a player what to do. That’s definitely changing in a lot of places, specifically here. ‘Let’s talk about it. Let’s come together and figure out a solution.’ With Scotty, we have open and honest communication where he feels comfortable enough to say that.”
The other lesson here is how Romo and Alexander adapted. Neither appears to have been affected by pitching in an unusual role or unfamiliar inning. Hudson opened with a scoreless frame on Thursday at PNC Park.
Alexander did alter some of his routine.
“I walked into the clubhouse and acted like I was telling Siri to play my starter playlist,” Alexander said. “I had fun with it.”
Some of his routine was the same. He went out for BP and stretched. But he warmed up as the Dodgers batted in the top of the first.
“As a player, you do what you’re asked to do. It’s our job. In our situation, to help a team out… we don’t really second guess it. ‘OK, what do I have to do to get ready?’” Alexander said. “It wasn’t a real start. I wasn’t over-hyped. If anything, it was less adrenaline because I am used to pitching in the seventh or eighth, or extra innings, in close games. Pitching in the first inning, it almost felt like no one was paying attention… It felt more like a spring training.”
Alexander pitched into the second. He faced six batters, recorded four outs, allowed two hits and a run. Dennis Santana followed with 3.2 innings as the Dodgers employed six pitchers to cobble together a win.
The box-score evidence:
“[Romo opening] was different,” Alexander said. “I don’t think a lot of teams are going to start to doing it. In our case, we had Santana making his debut [last Friday] in a really difficult environment to pitch in… Sometimes you have to do different things in order to keep yourself in the game. It’s obviously not stupid, but it’s different.”
It is different. It is probably not stupid, at least in some situations. And “opening” has now spread to the Dodgers. We’ll have to wait and see how widespread the practice becomes, but how the idea is spreading might be as interesting as the idea itself.
A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.
The question really isn’t “can the opener do it”, he’s got an easier job than normal relief. It’s can the inning eater who follows him do it?
Santana is used to starting, which lets him know when he’ll come in and always start a clean inning. Having an opener changes his role in ways that make it slightly harder, OTOH, he got a win for 3.2 innings, which is not something that happens for starters, so it’s not all bad.
(Usual added comment: WAR for starters and relievers is increasingly broken, you need an adjustment for appearance length, not based on who “started” the game.)
I thought WAR was adjusted for the leverage, not simply for whether it is a starter or reliever?
No, WAR isn’t a situational stat. Fangraphs’ version is based on FIP.
WAR is not leverage adjusted; it is a context neutral statistic.
The expected performance levels that are replacement level ARE different, as its easier to pitch well in relief.
There is a leverage factor for relief pitcher WAR: https://www.fangraphs.com/library/war/calculating-war-pitchers/
I wouldn’t say it’s broken. WAR is based on an adjusted FIP and scaled to innings pitched. If you want to look at situational stats, that’s what WPA is for.
edit: didn’t know they used different replacement levels for starters vs relievers. You could do something different based on length, but until this is a widely used strategy I’m not sure it would be worth it.
Pitcher WAR is based on FIP and FIP does factor in appearance length (it’s the denominator for the entire calculation, other than the offset designed to make FIP look more like ERA). On the various Fangraphs pages you can split out starters and relievers (because “games started” is a stat that is tracked) but that’s no different than splitting out batting stats by position for multi-position players, or pinch-hitting vs starting, etc.
However, despite what a couple of people have said, TKDC is correct and pitcher WAR does incorporate leverage — for relievers. From that first link above:
The leverage is “devalued” to some extent because relievers aren’t entirely responsible for the leverage (since they often inherit it from a meltdown by a previous pitcher): see this post by Dave Cameron (from a long time ago — and note the argument about “WAR doesn’t value relievers correctly” goes back pretty much to the beginning of WAR)
Starters get a lower replacement FIP, a starter and reliever can produce identical results over an identical number of innings in a game with an average leverage, and the reliever gets (far) less WAR because the replacement levels differ. The replacement levels differ because it’s assumed that short appearances are easier than long ones.
Santana got less WAR for this than he’d have gotten if he started, even ignoring leverage, because he’s now a “reliever”, the opener got more WAR because as a “starter” he faces a lower replacement level.
I’d argue that an “opener” facing the opposition’s top of the lineup right out of the gate is at least as difficult as almost any other situation a reliever might encounter.
On paper it’s a lot harder. Speaking from experience though , the reduced pressure compared to late and close games does a lot for an athlete.
makes perfect sense, since the first inning, statistically, is THE single-highest scoring inning.
With more evidence it will be interesting to see how this strategy compares with rolling out the team’s 5th starter. I could see the ‘opener’ strategy working out better than using your 5th “best” starter.
Absent a particularly large platoon advantage (like with the Angels all RH top of the lineup) I don’t really see how this makes a team better.
The SP face weaker hitters than he would have if he started and pitched to 20-24 batters, but the RP faces tougher hitters than he would have coming in after the SP.
Seems like a wash.
I think the 1st inning platoon advantage is the key factor.
Also, if more teams do this to the Angels, maybe Ohtani DH’s and bats second every non-pitching day (at least we can hope.)
I think you are overlooking the fact that the average reliever is better than the average starter (about 10% better by ERA or OPS+). As a result you are guaranteeing that your better pitcher faces the better part of the opposing lineup.
Additionally, your starter now has to face the top of the lineup fewer times. Previously, if your starter faced 21 batters, they would have to face the top of the lineup (1-3 hitters) three times. However, if your relief pitcher gets the first 6 outs then your “starter” gets to start with the 7-9 hitters and, if they faced 21 batters would be getting the 7-9 hitters three times instead of the 1-3 hitters. This provides all sorts of advantages including
– Your 3rd time through the order penalty applies to weaker hitters
– Your starter can probably last longer because he is facing weaker hitters
Aside: If openers became prevalent enough, what stat would we give them, a la “hold” or a “save”? Would it simply be considered a hold? After all, the opener would be a bridge to, presumably, the “starter”—someone with a historically defined role. Or would we come up with a completely new stat, and if so, what would we call it? An “Open”? Furthermore, what would the requirements of an open be? One inning of transfer of either a lead or a tie?
Maybe Game Started?
I know the stat that will get used, even if it’s not useful. “Team W-L in games started”
Less stat-inclined institutions love that for everything from pitchers to backup goalies to overachieving quarterbacks.
Curious as to whether a team must announce its opener prior to setting lineups, and, if so, whether the use of the opener would cause the other team to alter its batting order for better first-inning match-ups..
Managers already adjust the lineup depending on the starter so they could react to an Opener similarly but I expect most opposing managers won’t change lineups .
Does the opener mitigate the risk of a starter going 3x through the lineup? The “opener” maybe faces hitters 1-4, so the “starter” (“seconder”?) Faces hitters 5-9 three times but 1-4 only twice.
In the NL you might put the pitcher in the 3 hole (home games) and pinch hit in the bottom of the first. Double switch the pinch hitter later so the second pitcher is slotted at the usual 9th. So you then also gain the advantage of an extra professional at bat. Unless my logistics are off.
Yes. This is the biggest value-add of the ‘Opener’.
No, you’re spot on. Been suggesting this for quite some time as one of the hidden advantages of starting a reliever. It’s only a matter of time.
Double switching then loses a bench bat, so I doubt that teams would do that often.
Teams carry what? 5 bench bats. Forget using your bench catcher. The likelihood of needing all 4 useable bats likely does not outweigh the benefit of essentially using a DH.
1 place where this would be good is if you have a guy who normally starts and isn’t pitching that day but is a real good hitter- a guy like a Madbum or Scherzer even- and have him be used as the PH in the 1st inning.
Yes, the hitters who face him three times are notas good, so the third time through the order penalty is mitigated because it is incurred against less talented hitters.
The benefit is lost if he faces 27 batters. He just faces them in a different order.
I was at the Dodger game yesterday, and this particular one was injury related. Dennis Santana was the announced starter and went with Yasmani Grandal down to the bullpen about 30 minutes before game time. Then about 10 minutes before first pitch the public address announcer said that there has been a change and Daniel Hudson would be the starter. Apparently Santana pulled a lat muscle during his warm-ups.
This came 17 hours after Caleb Ferguson in his first ever start only managed to get 4 outs for the Dodgers.
yeah i dont understand why a clear injury situation is being portrayed as a use of “the opener” here.
You can take the man out of the rays organization but you can’t take the rays out of the man.
The Rays really figured out something here and other teams will emulate them more and more. Too bad they wasted it on a wasted season, and by the time they could use this to any important advantage everyone will be doing it.
The Rockies have been particularly bad at allowing runs in the first inning this year.
They might want to take a serious look at this.
I find it amazing that utilizing the “Opener” begins the same way that I would imagine a threesome starts, “Hey, listen, I’m not saying I would definitely do it, but I am open to the idea…”
Why is this so crazy?
I think one problem with the opener we saw Wednesday with Tampa. They used Venters and well he goes single,walk,double,k, walk,double. Then they bring in a guy who is used to starting in a situation with guys on 2nd and 3rd, and he winds up giving a hit to the pitcher and all of a sudden you have 5 runs allowed.
So if an opener is off, it really makes the situation that much worse. They have a guy following him who isn’t used to coming in such a difficult situation.
If he’s “off” that day, he’s just going to give up those runs later in the game. Extrapolating one game’s results for an entire strategy is not a good way to evalute whether that strategy is beneficial.
But the problem is if he’s off later in the game, manager is going to get him out of there quicker, and more importantly a guy coming in who is used to coming in rough situations is coming in. I think we’d need to look at not only the openers stats but also the guy coming in directly after him in looking at the effectiveness.
Great job, Travis!