While managing the Kansas City Royals to consecutive World Series trips in 2014 and 2015, Ned Yost seemed to have discovered the perfect October blueprint: Have your starters scratch out five or six innings and then turn the game over to three or four dominating relief pitchers.
In 2014, it was Kelvin Herrera, Wade Davis and Greg Holland. In 2015, it was Luke Hochevar, Ryan Madson, Herrera and Davis. All were right-handed, which made Yost's job even easier. He didn't have to worry about matchups, where the opposing team was in the order, or wait for his starter to get into trouble. The Royals went 22-9 over the two postseasons, including a title in 2015, and those five relievers combined to go 9-0 with 11 saves and a 1.18 ERA. The Royals' starters had a 4.56 ERA.
With relievers impacting the game more than ever and the days of Bob Gibson and Sandy Koufax throwing complete games in the World Series long in the past, this was going to be the formula for the future. After all, it was certainly easier in theory to find three or four superior relievers than four starters who could go deep into a game through multiple playoff series.
Except a funny thing happened along the way to Copycat Land: It hasn't happened.
The past four World Series champions all relied on starting pitchers coming out of the bullpen throughout their playoff runs, and the past four postseasons have seen more pitchers both start and relieve than ever before.
It began with the Astros in 2017, although not by design. A.J. Hinch had to improvise when his top two relievers, Ken Giles and Chris Devenski, both struggled in the postseason. By the middle of the American League Championship Series against the Yankees, Hinch had lost confidence in Giles. In Game 7, Charlie Morton started and pitched five scoreless innings while Lance McCullers Jr., who had started Game 4, threw the final four innings -- closing out the 4-0 victory with 24 consecutive curveballs.
After Giles blew a save in Game 2 of the World Series against the Dodgers, the Astros used just two pitchers in a 5-3 victory in Game 3: McCullers for 5.1 innings and Brad Peacock, who had started in the division series, for the final 3.2. In Game 7, McCullers started and pitched 2.1 innings, Peacock threw two and Morton the final four.
Morton batted in the ninth inning of that game with the Astros leading 5-1.
"The way he was pitching I wasn't taking him out," Hinch said after the game, even though Morton was pitching on three days of rest after going 6.1 innings in Game 4. His message to Morton: "Whatever you do, don't swing. Take every pitch. You're going to finish the game."
With the Red Sox in 2018, Alex Cora had a calculated plan to use his starters in relief. Rick Porcello and Chris Sale earned eighth-inning holds in the AL Division Series against the Yankees. In the ALCS, Porcello and Nathan Eovaldi both started and pitched in relief. In the World Series, it was Sale -- and not closer Craig Kimbrel -- who got the final three outs. Over the entire postseason, all five Red Sox starters would both start and relieve.
In 2019, Nationals skipper Dave Martinez had four stellar starters in Max Scherzer, Stephen Strasburg, Patrick Corbin and Anibal Sanchez, but only two trustworthy relievers in Daniel Hudson and Sean Doolittle. Scherzer and Strasburg both made relief appearances, while Corbin would end up making three starts and five relief appearances, including three scoreless innings to get the win in Game 7 of the World Series.
Dave Roberts did a little bit of everything with the Dodgers in 2020, including bullpen games. Julio Urias, Dustin May and Tony Gonsolin appeared as both starters and relievers. Roberts would ride the hot hand like Hinch did in 2017, as it was Urias who pitched the final nine outs of Game 7 of the National League Championship Series and the final seven outs of the clinching Game 6 of the World Series.
Of course, using starters in relief isn't a new strategy -- think Randy Johnson coming out of the bullpen for both the Mariners and Diamondbacks -- but those were more often emergency, do-or-die situations as with Johnson or Madison Bumgarner in 2014. Joe Torre, who could rely on Mariano Rivera, rarely did it with the Yankees. Bobby Cox rarely did it with the Braves; Tom Glavine never appeared in relief in a playoff game, and Greg Maddux and John Smoltz (as a starter) did so only twice.
Usage of this approach has increased in recent postseasons, however. Twelve pitchers appeared as both starters and relievers in the 2020 postseason, matching the totals from 2017 and 2019 for the most in a single postseason since the division-series round was added in 1995. Compare to 2006, when no pitchers appeared in both roles, or 2008, when only one did. Even in 2016, only four pitchers did it and two of those -- Jon Lester and Trevor Bauer -- did so only in that wild World Series Game 7. The average number of pitchers doing this each postseason through the years:
1995-1999: 5.8
2000-2009: 4.3
2010-2016: 4.6
2017-2020: 11.3
While managers often attempt to script out their pitcher usage for a nine-inning game and have specific matchups they want to get to, you never know how a game will unfold. Managing a postseason game -- let alone an entire month of high-tension baseball -- still requires a fair amount of managing from the gut.
That's what Roberts did with Urias in using him to close out two series.
"I didn't know [how long he would pitch]," Roberts said after the Dodgers clinched the title in Game 6 of last year's World Series. "I just liked him right there to get out of that spot and I felt as long as he's throwing the baseball well and feels good, we're going to keep running him."
That contrasted with the decision Rays manager Kevin Cash had made earlier in the game, when he pulled starter Blake Snell for Nick Anderson -- ignoring Snell's dominant performance so far, and going with analytics that said Snell is less effective the third time through the order.
"Analytics is a huge part of our success," Rays center fielder Kevin Kiermaier said after the loss. "And sometimes it can bite you in the butt."
That makes second-guessing the manager in the playoffs a game in and of itself. Ride the starter? Trust the middle of your bullpen? Extend your closer beyond just the ninth inning? Then you factor in that you have to win at least 11 postseason games and that massaging a pitching staff through October is a lot more difficult than just starting Gibson in Games 1, 4 and 7 of the World Series.
So managers have to get more creative, especially if they don't have a great bullpen. Of the eight teams left in the playoffs, here are the four best equipped to deploy that strategy:
Houston Astros
The Astros have six good starters -- Jake Odorizzi, who was left off the ALDS roster, had the worst ERA of the group at 4.14 even though he had been pitching well with a 2.72 ERA over his past eight outings -- although manager Dusty Baker already said that Zack Greinke moving to the bullpen for the playoffs is a "distinct possibility." Greinke has been out since Sept. 19 due to neck soreness. Still, Baker has plenty of depth in Framber Valdez and McCullers, with Jose Urquidy and Luis Garcia in the mix to shuttle back and forth between starting and relieving if needed.
The Houston bullpen is ... OK. Closer Ryan Pressly has elite numbers and wipeout stuff, so no worries there. Kendall Graveman's control has wavered a bit since coming over from Seattle, but he's the likely setup guy, with Ryne Stanek, Phil Maton and lefties Blake Taylor and Brooks Raley after that. The Astros have 33 saves and 26 blown saves -- an indicator that the middle relief has been shaky at times.
It feels like Baker often waits for a pitcher to get into trouble before replacing him. We can devise a little proxy for this theory by looking at the percentage of relief appearances made with the bases empty when the reliever enters. Turns out the Astros are middle of the pack in the majors in this area (17th). Still, with the additional flexibility Baker will have at his disposal, he has multiple strategies to consider.
Los Angeles Dodgers
Anything goes with five quality starters. Maybe Roberts plays it straight this year and simply moves Gonsolin to the bullpen, especially since Kenley Jansen and Blake Treinen look more reliable than a season ago. Still, although the bullpen has good overall numbers, it hasn't been clutch: The Dodgers are 6-13 in extra-inning games (although the zombie runner rule disappears for the postseason). No doubt Roberts will be tempted to use Scherzer or Urias in relief in a key moment (probably less so with Walker Buehler, who started Game 1).
Atlanta Braves
The Atlanta bullpen has a higher ERA and worse peripherals than the starters, and closer Will Smith has been a little shaky. Charlie Morton has done the starter/reliever thing before and Max Fried has been outstanding in the second half. Atlanta's bullpen was its strength a season ago, so it's hard to get a read on how Brian Snitker will run his staff this October, and there isn't really a reliable fifth starter to give that extra security if you use another starter in a relief.
Milwaukee Brewers
The Brewers are maybe the most intriguing team to see here because they have the tremendous 1-2-3 in Corbin Burnes, Brandon Woodruff and Freddy Peralta, plus Eric Lauer and Adrian Houser have been solid starters as well (Lauer has a 1.78 ERA over his past 13 starts and allowed more than two runs just once). But they also just lost premier setup man Devin Williams, likely for the entire postseason, with a broken hand. There is still plenty of bullpen depth behind closer Josh Hader with Brad Boxberger, Brent Suter and rookie Aaron Ashby, but Craig Counsell will have to deviate from his usual late-game plan.
With a comfortable lead in the NL Central for most of the second half, Counsell hasn't had to ride his staff hard at all. In fact, the Brewers have used a six-man rotation most of the season -- Burnes, Woodruff and Peralta have combined to make just eight starts on four days of rest (two each from Burnes and Woodruff and four from Peralta, and two of those came after short outings of two innings).
Using a four-man rotation in the postseason means mostly starting on four days of rest. So on one hand, Brewers starters enter the postseason more rested than some of their competitors; on the other, they aren't used to pitching on four days of rest. Maybe that's not a big deal; maybe it is. Likewise, Hader hasn't pitched more than one inning in a game all season, but without Williams you have to think Counsell will consider Hader for longer saves.
Then consider that Peralta and Lauer often get pulled after five innings. Peralta hasn't pitched seven innings in a game since June 16. If Counsell wants to concentrate as many innings as possible in his best pitchers, that means we could see Burnes, Woodruff or Peralta out of the bullpen at some point for a crucial inning.
Indeed, that's another factor to this postseason. We've heard managers talk all season about trying to protect their pitchers after the shortened 2020 season. How will that affect decisions with pitchers? Who is still going strong and who is gassed? Stay tuned.