“If we swapped pitching staffs, we would probably be doing something similar to what they’re doing and they would be doing something similar to what we’re doing,” Dodgers general manager Farhan Zaidi told the Los Angeles Times, comparing the way the two NLCS teams handled pitching plans in the series.
Had the Brewers won the game and/or won the series, perhaps more people would have complained. Perhaps baseball would have even instituted some rule requiring a starting pitcher to face a certain number of batters or record a certain number of outs, unless he leaves the game because of injury.
Perhaps teams will just stop listing probable starters, although that seems unlikely. While listing probables has been seen mostly as a courtesy to opponents, it can also help spur ticket sales for games involving aces. It also matters greatly to gamblers and to those playing daily fantasy games, some of which are now in commercial partnerships with MLB and individual teams.
Teams will continue to try for favorable batter-pitcher matchups, which in essence is what the Brewers were attempting to do. But the strategic deception probably only works in a playoff series, and only with the right combination of pitchers and opponents. It’s not like the Dodgers were going to have Kershaw face the opening batter of a game a day or two before he is really scheduled to start.
The Dodgers wouldn’t do that with Kershaw. The Brewers wouldn’t have done it if they had a Kershaw.
“Pick the five best starters in baseball,” Brewers GM David Stearns told the Los Angeles Times. “[Jacob] deGrom, [Justin] Verlander, [Max] Scherzer—if we had those three guys at the top of our rotation, I think [Counsell] would be really comfortable. I don’t think he’d have a problem letting those guys run out there.”
And not to face just one batter.
38. You Can Start, But You Can’t Finish
LONG BEFORE THE OPENER CAME TO BE, THE role of a starting pitcher had already changed. Where once a starter went to the mound hoping to be able go the distance and win a game, in recent years starters have been taught to get the team to the later innings with a chance to win and then hand it over to the bullpen.
Zack Greinke of the Arizona Diamondbacks had a 1.66 ERA for the Los Angeles Dodgers in 2015. Jacob deGrom of the New York Mets had a 1.70 ERA in 2018. Greinke and deGrom each made 32 starts; each completed just one of those starts.
Jack Morris, who went into the Hall of Fame in 2018, never had fewer than four complete games in any season where he started at least 25 games. He ended his career with 175 complete games.
At the end of the 2018 season, no active pitcher had more than 38. No pitcher under 30 had more than 15.
That’s 15 for a career. Morris completed 20 games in 1983 alone—and he didn’t even lead the league (Ron Guidry had 21).
Warren Spahn had twice as many complete games as Morris (375). Spahn had 12 seasons with at least 20 complete games. Fergie Jenkins and Robin Roberts each had eight.
In their day, top starting pitchers never thought about coming out of games. It was their job to pitch nine innings, or even more if the game went extra innings. Jim Kaat tells the story of a 1964 game he pitched for the Minnesota Twins in Boston.
Kaat had a 12–2 lead when he gave up three runs in the seventh inning. At some point, Twins manager Sam Mele was going to take him out of the game.
“I told him to give the guys in the bullpen a rest,” Kaat said.
He gave up three more runs in the ninth, but the bullpen did get a rest—and Kaat got a 15–9, complete-game win. That was in July. A month later, Bob Sadowski gave up nine runs in a complete-game win for the Milwaukee Braves. In the 54 years since, only one pitcher has given up as many as nine runs in a complete game he won—Burt Hooton in 1974 with the Chicago Cubs.
If Kaat wasn’t coming out of a game when he allowed nine runs, he certainly wasn’t coming out in the eighth or ninth inning when he had a chance at a shutout. It did happen, but just nine times in his 625 career starts. Compare that to the 180 times he had a shutout going and went all nine.
The game changed for starting pitchers, and the changes began long before analytics took hold in front offices. It’s been more than 30 years—since Fernando Valenzuela in 1986—that pitcher had 20 complete games in a season. Only one pitcher this century has gotten even halfway to 20 in a season—James Shields of the 2011 Rays had 11. In 2017, no one even went eight innings 10 times.
The game has changed, and it’s unlikely to change back anytime soon. Forget a complete game, which even if no one reaches base requires a pitcher to go through the entire batting order three times. In modern baseball, you have to be a special pitcher having a good game to even get the chance to face the opponent’s best hitters a third time.
And even most aces don’t get a chance to go a fourth time through.
In the entire 2017 postseason—that’s 38 games, 76 starting pitchers—only two pitchers were allowed to face an opposing batter a fourth time in the same game. Justin Verlander did it for the Houston Astros in Game 2 of the ALCS. Jake Arrieta did it for the Chicago Cubs in Game 4 of the NLCS (and he only faced one batter a fourth time). Far more often, the starting pitcher was gone before the middle of the opposing order batted a third time, and sometimes well before that.
It was no different in October 2018, when there were 39 games and still just two pitchers who made it to the fourth time through the order (Verlander in Game 5 of the ALCS, Walker Buehler of the Dodgers in Game 3 of the NLCS). And even in those cases they were done after facing the leadoff hitter a fourth time.
It didn’t matter whether the pitcher was cruising. It didn’t matter if his pitch count was low. CC Sabathia had thrown just 77 pitches for the Yankees in Game 2 of the American League Division Series against the Indians. He had retired 13 of the last 14 batters he faced.
But it was his third time through the order. Yankees manager Joe Girardi couldn’t wait to get to the bullpen, to give those hitters a different look. Sabathia was out, in a game the Yankees led by five runs.
“That made no sense,” Jimmy Rollins said on TBS a few days later.
It made sense to the guys who make decisions by the numbers, because the numbers say hitters have a better chance the more times they see the same pitcher in a game. In 2018, batters had a .784 OPS in their third plate appearance against a starting pitcher, compared to .700 the first time through and .720 in their first appearance against a reliever.
The numbers don’t lie. But they also don’t apply to every situation.
“Stop focusing on the computer,” Gary Sheffield said on TBS, agreeing with Rollins. “Let your eyes tell you what you see.”
It’s more than just eyes. Managers can also look at a pitcher’s history. They have numbers that show not only whether a pitcher is losing velocity but also whether his arm angle is changing (a sign of fatigue). The decision doesn’t need to be one-size-fits-all.
Most often, though, it’s close to that. The third-time-through rule has become as established as the 100-pitch rule. If a starter is nearing 100 pitches, it’s time to think about a change. If the lineup is turning over and the leadoff hitter is coming to the plate for a third time, it’s time to think about a change—if the manager hasn’t made one already.
Look at what happened in Game 2 of the 2017 World Series. Dodgers starter Rich Hill had allowed just three hits in the first four innings. It was a 1–0 game. Hill’s spot in the lineup wasn’t coming up. But he’d already gone through the Astros’ batting order two times. Despite the success, Dodgers manager Dave Roberts was concerned. Hill was out of there.
Removing his starter so early—and unnecessarily—meant Roberts ran through almost his entire bullpen by the ninth inning. It almost worked, but when closer Kenley Jansen gave up a game-tying home run in the ninth, Roberts had only Josh Fields, Tony Cingrani, and Brandon McCarthy remaining. It was hardly a surprise to see the Astros score two runs off F
ields and two off McCarthy to win Game 2 and even up the Series.
When you’re asking your bullpen to cover that many outs in one night, you’re counting on way too many people to have a good night. Sometimes it’s unavoidable because your starter didn’t have it. But Roberts made the same move on a night his starter did have it.
Roberts took plenty of heat that night, but he was doing what he had done all postseason and really all season. He was doing, presumably, what the Dodgers’ analytically inclined front office wanted him to do. Dodgers starters faced a batter a third time in the same game fewer times than starters on any other team in baseball in 2017. Four times in their first 10 postseason games, Roberts pulled his starter after he had gone through the order exactly two times (and in another game he pulled him one batter into the third time through). The only starter allowed to go deep in the game was Clayton Kershaw.
Until Game 2 of the World Series, the strategy worked. The Dodgers bullpen didn’t allow a single run. The Dodgers swept the Diamondbacks in the Division Series and won the NLCS over the Cubs in five games.
“The way we’ve done things all year long, I know our players understand it, believe in it,” Roberts said a day after Game 2. “I know I believe in it.”
They were willing to run the risk of what happened in Game 2. It was an unnecessary risk.
Even in today’s game, the smartest managers know to trust their eyes and the smartest teams let them do it. No team relies more on analytics than the Houston Astros, and general manger Jeff Luhnow knows all the numbers about the third time through the order. But in Game 2 of the 2017 ALCS, the Astros let ace Justin Verlander throw 124 pitches in nine innings. Verlander went the distance and faced 32 batters, meaning manager A.J. Hinch allowed him to face the middle of the Yankees order four times.
The last starter to face that many batters in an LCS game: Verlander, six years before in a win over the Texas Rangers.
“Those rules of thumb about times through the order, they’re exactly that, rules of thumb,” Luhnow said a couple days later. “There are some guys who have trouble the second time through. There are some that have trouble the first time through. In aggregate, I think the third time through is when most starting pitchers probably have trouble. It makes sense, logically, but when you have a pitcher like [Verlander] whose stuff seems to tick up later in the game, those rules don’t apply.”
He has to trust his manager, and the manager has to trust his eyes. Luhnow is willing to do that with Verlander, and more importantly with Hinch.
It used to be it was like that with every pitcher and every manager. Managers knew they had to watch some pitchers more carefully as the game went on, but they weren’t nearly as quick with the hook. Just a decade ago, the 2005 White Sox had three starting pitchers with 25 more more games where they went through the order three full times. Twelve years later, in 2017, not one starting pitcher in all of baseball had 25 games like that.
Part of the reason is an increasing concern about pitch counts, to be sure. Part of the reason is that teams work harder on improving their bullpens. But a lot of it comes down to numbers. Most starting pitchers don’t do as well the third time through the order. Chris Sale is one of the best pitchers in baseball, but batters went from a .282 slugging percentage the first time they saw him to a .408 slugging percentage if they saw him for a third time in the same game in 2017. Even as good as Kershaw is, opponents had a .297 slugging percentage the second time through, and a .438 slugging percentage the third time that same season.
It’s so easy to talk yourself into bringing in a fresh arm, but there are plenty of times you’re just doing your opponent a favor. Take Game 2 of the 2014 National League Division Series between the Giants and Nationals. Nats starter Jordan Zimmermann was one out from evening the series at a game apiece. The Giants had done absolutely nothing against him. He walks a guy with two out, and Nationals manager Matt Williams goes to his bullpen.
You don’t have to guess what the Giants were thinking. They told us, after they tied the game off Drew Storen and won it in 18 innings.
“[Zimmermann] is one of the best pitchers in baseball,” Giants pitcher Tim Hudson told Jayson Stark of ESPN.com. “So obviously, when you don’t face him, you’re not exactly pissed.”
Not every starter is Zimmermann in October 2014. It’s not ridiculous at all to pay more attention when your average starter gets deeper into a game.
In his 2016 book Ahead of the Curve: Inside the Baseball Revolution, MLB Network host Brian Kenny made the argument for what he called “bullpenning,” redistributing innings in a way that hitters would almost never see the same pitcher three times. Kenny quoted 2014 MLB numbers showing hitters with a .755 OPS the third time they faced a starter, as opposed to a .682 OPS in their first plate appearance against a reliever.
But even Kenny says his ideal staff would include one “ace” who throws 220 innings (almost seven innings a game, if you assume 33 starts) and another starter who throws 200. Not a single team in the major leagues had two pitchers doing that in 2017. Only three teams had a pair of 200-inning pitchers. Not surprisingly, all three (the Red Sox, Nationals, and Indians) won their division.
In 2018, only one pitcher in all of baseball—Max Scherzer of the Nationals—threw 220 innings. Of the three teams with two or more pitchers who topped 200 innings, only the Arizona Diamondbacks missed the playoffs.
Kenny’s premise is that you want to limit the innings thrown by your lesser starting pitchers. He advocates for using the opener, as the Tampa Bay Rays did in many games in 2018.
The trend in today’s game has been to limit innings and exposure for almost every starting pitcher. That means more innings for the bullpen to soak up, and that means carrying more pitchers. Teams now routinely have more pitchers than position players on their 25-man roster.
One problem with that and with limiting innings thrown by your starters: it means more innings thrown by pitchers who wouldn’t otherwise be in the major leagues. It also means if you do ask your starter to go deeper, he may not even know how to do it.
When starters routinely worked deeper into games, they learned to go through the early innings without using all their pitches. When they faced a batter for the third time, he wasn’t seeing the exact same sequence of pitches he saw the first and second times up. Now starters know they’re probably gone from the game before that guy comes up a third time. Why save anything for an at-bat they’ll never see?
It was never that way in the past. Morris, who pitched in the major leagues for 18 years before retiring after the 1994 season, said he went to the mound hoping he could throw only fastballs to the first nine batters in a game.
“If I could do that, then they still hadn’t seen my slider, forkball, or changeup,” he said. “I had something different to throw them later, with the game on the line.”
It showed. In 1983, when Morris had those 20 complete games and threw a league-high 2932/3 innings, opposing hitters had a lower OPS when they faced him for a third time in a game (.684) than when they faced him the first time (.706). Morris wasn’t giving manager Sparky Anderson a reason to routinely take him out of a game at that point.
In the best-known game of Morris’ career, he went through the opponents’ order four full times and faced two batters a fifth time. It was October 27, 1991, and it was Game 7 of the World Series. Morris told Minnesota Twins manager Tom Kelly he wasn’t coming out of the game, and fortunately for the Twins, Kelly listened to him. Morris threw 10 shutout innings in one of the most iconic pitching performances ever, with the Twins winning 1–0 on Gene Larkin’s single off Atlanta Braves reliever Alejandro Pena.
Morris threw 126 pitches that night at the Metrodome. Looking back, he said he could have gone even longer, if needed.
“I could have gone 15,” he said. “It was the last game of the season. There was no need to hold anything back.”
; It’s been more than 40 years since a major league pitcher threw 15 innings in a game. No one has ever done it in the postseason (Babe Ruth threw 14 for the Red Sox in the 1916 World Series).
Morris never threw more than 11 innings in a game, but he did throw those 175 complete games.
“The funnest part about pitching was walking off the field with my teammates,” he said. “The game’s over and I was involved in all of it, not just part of it.”
Pitching in a different era, Verlander had just 24 career complete games through the 2018 season. But he also had 53 other games where he finished the eighth inning. Verlander’s idol as a kid was Nolan Ryan, who completed 222 of his 773 career starts (including 26 of 39 with the 1973 Angels).
“Every time I take the mound I have the mentality of trying to go nine,” Verlander said during the 2017 postseason.
Few pitchers today would say the same thing. As well as deGrom pitched in 2018, he seemed to have no problem at all when Mets manager Mickey Callaway pulled him after eight shutout innings in his final start of the season, even though he was working on a two-hit shutout of the National League East champion Atlanta Braves. DeGrom told reporters he was happy because his goal had been to get the 10 strikeouts he needed to reach 1,000 for his career, and he did it with the final out in the eighth inning.
Some old-school scouts watching the game cringed at hearing that, but it’s not deGrom’s fault. In today’s game, when the third-time through is already considered the danger zone, complete games are just not a realistic goal.
39. Bullpen by Gabe
WITH STARTING PITCHERS LEAVING GAMES earlier, even with the lead, bullpen management becomes a bigger issue than ever. For much of the past three decades, that meant finding a closer who would normally pitch just the ninth inning and then building a bridge from starter to closer by finding one or two set-up men who would normally pitch the seventh and eighth innings when you have a lead.
Unwritten Page 16