It’s Game 4 of the World Series, and up on the marquee in blazing lights is tonight’s starter for the Los Angeles Dodgers: TBA. Even when they announce it, it won’t feel right. Ryan Brasier? Really? That’s the plan?
Yeah, as much as it may not be fun for the fans, a reliever is the plan, and it’s been the plan all along. In the regular season, a bullpen game usually marks a quirk in the schedule where a certain team may just not be able to field enough healthy starters and needs to try to remain competitive. But it’s become a postseason strategy, perhaps because of the off days baked into the schedule, or also because throwing multiple relievers at a lineup can be a successful way to win a big game.
Is it successful, though? Or is it just something you do if you don’t have enough starters? That’s the state of the Dodgers currently, though in times past other teams might have just started Landon Knack and left him in as long as he was successful. If you sum up all the times when a starting pitcher recorded fewer than 30 pitches in the postseason, you get a 16-24 record in 40 games since 2015 began. But that includes games like Luis Severino’s blow-up in the 2017 American League wild-card game — not really the spirit of the matter, that’s the unplanned version.
When the Oakland A’s used Liam Hendriks as the first pitcher in their 2018 wild-card game, it marked the first time in the Statcast Era that a team had decided upon a bullpen game in the playoffs. That was the plan. If you go back through all of these 40 games by hand and remove the cases in which a traditional starter just had a miserable time in the first inning and was removed, you get 32 games that were bullpen games in the way we understand them. The record in those games for the bullpen team is 14-18, which makes it seem like a slightly better, but still worst-case scenario, type of strategy.
This mimics what happens in the regular season, as far as the numbers can tell us. STATS Perform looked at all games since 2015 in which the starter threw fewer than 30 pitches and found that teams were 428-449 with a 4.52 ERA in those games — but if you limited it only to games in which the starter threw 30 or fewer pitches *and gave up three or fewer runs* to try and remove the blowups and find the “planned” bullpen games, the team’s record improves to 422-403 with a 4.19 ERA. Planned bullpen games are a viable strategy from April to October.
But not all teams are alike. The Dodgers might have been in the postseason a lot since 2019, but they’ve also thrown a whopping nine bullpen games in the past 10 postseasons. The Houston Astros have been in more postseason games and have only thrown two bullpen games. You can add up the New York Yankees’, Atlanta Braves’ and Cleveland Guardians’ postseason bullpen games — they’ve had a combined 145 games in the playoffs since 2015, compared to the Dodgers 102 — and you get five.
Oh, and the Dodgers’ record in those nine bullpen games in the playoffs? Six wins, three losses. They’ve won six of the seven series in which they’ve deployed the bullpen game. They’re winning those games at a 67 percent clip while everyone else has won 35 percent of their postseason games started by a reliever who threw fewer than 30 pitches.
Why?
One reason is probably just quality depth. We only have pitch modeling numbers since 2020, but over that time frame, the Dodgers have probably had the best bullpen in baseball judging by the movement, velocity, spin and locations of their pitches.
| Team | Stuff+ | Location+ | Pitching+ |
|---|---|---|---|
 Dodgers |
107
|
101
|
103
|
 Rays |
108
|
100
|
103
|
 Astros |
112
|
98
|
102
|
 Padres |
107
|
100
|
102
|
 Guardians |
107
|
99
|
101
|
The Yankees have had similar stuff over the same time frame (a 108 Stuff+) but showed worse command (98 Location+) that kept them out of the top five here by Pitching+, which puts together the elements of stuff and command to create a third process-based metric.
This year’s Dodgers crew is led by Michael Kopech in terms of stuff, but went 15 pitchers deep with average stuff or better, and included important relievers with both command and stuff like Alex Vesia, Daniel Hudson and Ryan Brasier. This is a team that is active on the waiver wire (Evan Phillips) and makes both trades large (Kenta Masada for Brusdar Graterol counts, probably) and small (Dylan Floro for Vesia and Kyle Hurt) to bolster their bullpen for moments like these. They don’t often spend top-end cash in free agency on this one facet of the team, perhaps because they spend freely on the rest of the team, or perhaps because stuff doesn’t age all that well and they prefer younger flamethrowers to their veteran counterparts.
Part of their plan seems to be related to the old “clock face” idea with the Tampa Bay Rays: each reliever is a little different than the last, if not as obviously defined by release point as Tampa Bay has featured in the past. Kopech throws a four-seamer with great ride and a power slider with little movement. Blake Treinen throws a super sinker and a sweeper, and everything seems to go sideways. Brasier might have a similar fastball to Daniel Hudson, but their sliders are very different and Brasier mixes in a sinker. The idea is that even if they are all righties, they aren’t throwing the same shapes at you.
Familiarity with shapes might be an important way that batters can get comfortable against relievers. We’ve seen research suggesting that familiarity is the source of the third-time-through-the-order penalty, and now that we’re looking at a third-time-against-a-reliever penalty, it’s fair to think that the source is the same for relievers.
The third time through the order penalty, where pitchers get worse results as batters see more of their pitches, shows up as a residual in my models.
It is decorrelated from my xRV which shows that it is due to familiarity and not pitchers throwing lower quality pitches. pic.twitter.com/HN0JxWOnyG
— Cameron Grove (@Pitching_Bot) November 21, 2021
One last reason may have nothing to do with the relievers they use. The Dodgers allowed the second-lowest batting average on balls in play (BABIP) this year. Include last year, and they’re top three. Including 2022, they’re first. Go all the way back to 2015, and they’ve allowed the lowest BABIP in baseball. By some defensive stats, the team has been below average over that same time frame, and yet they gobble up balls in play. It’s possibly because they are better at positioning their defenders than other teams. Here is how they ranked during the regular season by defensive value derived only from the positioning of the players, courtesy of Sports Info Solutions:
| Team | Positioning runs |
|---|---|
|
Dodgers |
51
|
 Tigers |
33
|
 Cardinals |
33
|
 Yankees |
29
|
|
Guardians |
28
|
The Dodgers have a huge research and development team with great minds powering their strategy, so maybe their success with bullpen games is due to something other than stuff modeling, shape variance or defensive positioning. The reality of the situation, though, is that they’ve had the most success with this sort of game. Just because they’re going to the bullpen in the first does not necessarily mean they are ceding the pitching advantage in Game 4.
(Photo of Dodgers manager Dave Roberts taking the ball from Anthony Banda in the fourth inning of Game 6 of the NLCS: Sean M. Haffey / Getty Images)
