There's been a fair amount of lament on Twins Daily about Rocco's propensity for using the bullpen game. Let's take a look at how it's gone.
First, the context. At the end of last season, the rotation set up as Berrios and unproven starters. "Impact Pitching" was needed.
There was lots of conversation on TD about what Impact Pitching meant, with some thinking that it meant get the highest-impact pitcher. Realistically, Gerrit Cole was not going to happen. The next tier was seen as Wheeler, Keuchel, Bumgarner, Ryu and maybe a couple others. None of them happened either.
Instead, rather than going for an Impact PitchER, they went for Impact PitchING. They offered a Qualifying Offer to Odorizzi, who accepted. They traded for Maeda. They re-signed Pineda, even though they knew he'd miss 36-39 games. They signed Hill and Bailey, and they offered Chacin a minor league deal. Add in Dobnak, Smeltzer, and Thorpe, and they entered spring training with eight options for the opening day rotation and anticipating two "mid-season acquisitions" in Hill and Pineda. Of the 10, seven had some track record of recent major league success and three were rookies, at least one of whom had shown significant promise in the prior year.
Then we got 2020ed by COVID-19. That made Hill a candidate for Opening Day and changed the nature of the schedule. General prognostication was that while the Twins may or may not have had an Ace Ace, they had a stable of guys ranging from lower-case aces to Nos. 2 and 3-type guys and shaped up to have one of the deeper rotations around.
Then we got 2020ed by injuries and the depth was deserted. Chacin didn't pan out and exercised his opt-out. Thorpe and Smeltzer didn't grab hold of rotation spots. Bailey pitched Game 4 and got hurt. Hill pitched Game 5 and then missed the next 19. And Odo missed the first two weeks before pitching three games, getting 2020ed by a comebacker that knocked him out of the rotation for a time.
That left Berrios, Maeda and a very successful Dobnak for three spots throughout the season to date. Practically speaking, the timing of the availability of Bailey, Hill, Odo, and now Pineda has essentially filled one more spot.
So what to do with the fifth spot?
But before diving into that, let's take a look at an important sub-context. The offseason move for Impact PitchING also included beefing up the bullpen. Romo was re-signed, and Clippard, Wisler and more were added to add depth to a bullpen that was already pretty deep. In addition, the roster was expanded to 28. When you look at the roster construction, it seems clear that the Twins used the two extra spots for the bullpen and have spent most of the season with a 15-man pitching staff. That means, once you take out the four-ish starter spots listed above (Berrios, Maeda, Dobnak, Bailey/Hill/Odo/Pineda), the team has spent most of the year with an 11-man bullpen.
So now, let's look at what's happened in the fifth spot, the eight games that weren't started by one of the seven guys in parentheses in the previous sentence.
Rocco's primary options were:
- Go with Thorpe and ride out the storm.
- Go with Smeltzer and ride out the storm.
- Call up Duran or someone from St. Paul and ride out the storm.
- Hope that Falvey would get reinforcements from somewhere else, likely needing to give up significant assets to do it.
- Go with Bullpen Games.
Well, No. 4 didn't happen. We don't know how hard they tried, and we don't know what it would have taken. But it didn't happen. For whatever reason, they also didn't go with Nos. 1, 2, or 3.
Instead, they've gone with No. 5, the Bullpen Game. And what have been the results?
I went back through the box scores and checked the performance of Twins pitchers in those eight games. Six were regulation games and two were doubleheader games. It wouldn't be fair to look only at the starter, because that would only be an inning or two per game. Instead, I looked at the performance of the bullpen in the first six innings of the regulation games and the first four innings of the doubleheader games. I chose four innings for the doubleheader games because Rocco has talked about managing these games as if the opening pitch was the first pitch of the third inning. Thus, four innings of the game equates to the end of the sixth in a regulation game, with the high-leverage guys set to work the final three innings.
That makes 44 total innings [(6 x 6.0) + (2 x 4.0)]. In those innings, a total of 11 guys have combined to give up 37 hits and 17 walkouts, for a 1.23 WHIP. They've struck out 39 batters, for a K/9 ratio of 8.0 and a K/BB ratio of 2.3. They've given up 17 earned runs for an ERA of 3.48.
Realistically, how can we be disappointed in a 3.48 ERA and 1.23 WHIP from our No. 5 starter? Additionally, in five of the six regulation games, it would have been considered a Quality Start. (In each of the two doubleheader games, they gave up two runs through 4.0 innings, which is the same pace as a QS.) The team has gone 5-3 in those eight games. Again, what team would be disappointed with going .625 in the games started by their No. 5. That's essentially 100-win pace and is actually a bit better than their overall record in the small sample size.
But wait, it gets better. In actuality, before fully going the bullpen route, they first tried No. 1 on the list above and gave Thorpe an opportunity on Aug. 3. He gave up six hits and four walks in allowing three runs over 4.0 innings. He didn't do well, and they pulled the plug on option No. 1.
Four days later on Aug. 7, they tried No. 2 and gave Smeltzer an opportunity. He actually pitched okay, giving up four hits and walk in allowing two runs over 4.1 innings. However, even with that respectable start, he only lowered his ERA to 8.00. He'd also thrown 62 pitches, so he wouldn't have been ready the next time they needed the No. 5 spot on Aug. 11.
So instead of going full-on to option No. 5, the Bullpen Game, they actually tried other options in Thorpe and Smeltzer. Those two starts are included in the numbers above, and I'd argue that they weren't really Bullpen Games. Rather, they were attempts at finding a fifth starter.
Take out those two games, and in the six true Bullpen Games, here are the numbers:
- 32 innings
- 23 hits + 11 walks for a 1.06 WHIP
- 11 ER for a 3.09 ERA
- 31 strikeouts, for an 8.7K/9 and a 2.82 K/BB
That's potentially an All-Star, folks. It's not quite as good as Maeda, but better than Dobnak or Berrios and probably better than the Hill/Bailey/Odo/Pineda position. I'm confident that few, if any, other teams have gotten that kind of performance from their No. 5 starter. (And I haven't referenced option No. 3 above, but do we really think Duran would have pitched that well, including going 6.0 innings per game?)
And speaking of 6.0 innings, of the remaining starters, only Maeda has averaged that much, so the Bullpen Game stats are actually being held to a higher standard, since they are leaving the rest of the bullpen with nine outs to go, whereas games from other starters have typically required more than nine outs from the bullpen.
Before calling it an unqualified success, we have to acknowledge the "yeah, but it leaves the bullpen depleted on other days" argument. I don't really have a good way of telling whether this has been the case, but anecdotally it doesn't seem like it. Consider:
- The bullpen is two guys longer than it is in a normal season, so they have more options, period.
- That mix of innings includes none from Romo, Rogers, and Duffey and only 3.0 from May, so they weren't needing to use the high-leverage guys to pull it off. Backing up the high-leverage guys, Clippard has pitched 6.0 innings over four of the Bullpen Games, but has still been available to pitch in 11 other games. The bullpen games have clearly not unduly limited Rocco's options for the ends of games.
- In only one of the eight games did they use as many as four guys to get through 6 (or 4) innings, so they still had a lot of bodies available (though not necessarily all rested), including a mixture of high-leverage guys (see previous bullet) and low-leverage guys.
- In all six of the true Bullpen Games, Maeda has started either the game before or after, meaning that there's a good likelihood the bullpen went into the game reasonably rested or could look forward to a good possibility of a long outing by the starter in the next game. In several of the occasions, the Bullpen game has been sandwiched between Maeda and Dobnak, strengthening that effect.
- In actuality, it's perhaps increased the depth of the bullpen, in that it's given guys like Wisler (13.1 of the 44 innings) meaningful innings.
Is planning to go with a Bullpen Game throughout the year, over a 162-game season, with a 26-man roster a good approach? I don't think that's been proven yet, and I don't think a team should plan for that. For this season, we may have seen the last of it, in that if Odo is ready by Tuesday, we would have a healthy Maeda, Berrios, Dobnak, Pineda, Hill, and Odo, with no more scheduled doubleheaders and two off days in the remaining 17 days of the season after next Wednesday and Thursday's off days that will fully reset the staff. If Odo is not available by Tuesday, we'll probably have a Bullpen Game in Tuesday's Game 2.
But with all this at play, I think that in the current context, the Bullpen Game approach has been the right decision and it has totally worked out so far.