These Bullpen Games Are Tearing Me Apart

Rob Schumacher/Arizona Republic/USA TODAY NETWORK

In Game 4 of the NLCS, Torey Lovullo ran out of starting pitchers, so he sent Joe Mantiply out there, and decided to mix and match with the rest of his relievers until the game was either won or lost. And it worked. Phillies manager Rob Thomson also went to his bullpen early, and before too long we had a close matchup of dueling bullpens. The hope in such a strategy is merely to survive to see the endgame, and pit one’s high-leverage relievers against the opponent’s offense, just as would be the case with a conventional starter.

Both teams got that far. Andrew Saalfrank had a little bit of a meltdown, but Orion Kerkering and Craig Kimbrel had a big meltdown, so the Diamondbacks won. Faced with an identical conundrum in Game 4 of the World Series, Lovullo played “Freebird” again and called on Mantiply once more.

How did it go? Not too badly for Mantiply, who was a bit unlucky to take the loss after striking out three batters and allowing two hits in an inning and a third. But after that, the Arizona bullpen imploded like an overripe tomato under the wheel of a dump truck.

In a flash, the Rangers scored 10 runs, and the game was basically over. We did see some drama near the end, but only because the Rangers’ bullpen isn’t exactly trustworthy either, and manager Bruce Bochy had too much Diet Mountain Dew before the ninth inning.

I’ll just come out and say it: That stinks! We only get seven World Series games a year, tops, and when one of them has you thinking “maybe I should go to bed” in the top of the fourth, that’s a real drag.

The whole affair was like taking a gamble on a skunked can of beer; it brought on feelings of discomfort and queasiness, which Tyler Kepner of The Athletic captured well in his postgame column: “On the World Series stage, a bullpen game just feels wrong.”

The following evening, we got a classic pitchers’ duel. Zac Gallen and Nathan Eovaldi traded zeroes through six innings; Gallen took a perfect game in to the fifth and a no-hitter into the seventh. It was like something out of the 1960s, when men were men and relief pitchers were merely a necessary evil.

The long and short of it, for people who remember Randy Johnson and Curt Schilling the last time the Diamondbacks made the World Series, or read about Bob Gibson vs. Mickey Lolich in 1968, is that the World Series ought to be a platform for exceptional performances by exceptional pitchers.

We’re seeing fewer of those. Starting pitcher workloads have gone down in general as the job has become more physically intense and strikeout-focused. Jay Jaffe wrote extensively last week about the decline in starting pitcher workload volume, both in the regular season and in the playoffs. This isn’t one of those boiling-the-frog-type slow descents into the dictatorship of Johnny Wholestaff, it’s happening more or less overnight:

These workload trends have really only gotten noticeable in the past five years, 10 at most. Which sounds like a long time, but if you’re a 50-year-old general sports columnist who only periodically dips into baseball, or a one-team fan whose ballclub hasn’t made the playoffs in a while, what you’re seeing now is quite jarring.

I want to highlight two points about the increase in postseason bullpen games. First, the problem with Game 4 is not that too many pitchers were used, it’s that the game was out of hand two innings in. In his column, in addition to mentioning the Curly Ogden Maneuver in Game 7 of the 1924 World Series, Tyler quotes Lovullo as saying that the game has changed since the days of the Big Red Machine. He then follows up by pointing out that in Game 6 of the 1975 World Series, Sparky Anderson pulled his starter after two innings and used seven relievers. If the game is close, nobody is going to give a damn how many pitchers are involved.

I’ll go one step further into the annals of history. Go back up a couple paragraphs and look at that chart again. See that blue dot in the middle, way below the others? That’s the postseason with the record lowest percentage of starters who made it through five innings: 1947.

That year, the postseason comprised only the World Series, and in seven games not a single Dodgers starter lasted long enough to qualify for the win. Only four of seven Yankees starts lasted that long, one of those being Bill Bevens’ famous no-hit bid.

Back then, relief pitching as an art was only about a generation old, and pitcher usage was a bit haphazard. High-leverage relief opportunities went to starters on rest days, as well as high-workload multi-inning relievers. The 1947 Yankees had one of the best closers in baseball, Joe Page, who went 14-8 with 17 saves in 44 appearances (leading the league) totaling 141 1/3 innings. In the World Series, he closed out Game 7 with a five-inning relief stint. (“Closed out” might not be the right term, since Page was credited with the win.)

And this despite the Yankees having a good rotation, with four above-average starters, three of whom went by Spud, Spec, and Bobo. God bless America.

The other point has to do with seeing what Andrew Heaney and Ryne Nelson did amidst the chaos on Tuesday night. Here, while both bullpens imploded, two pitchers each threw at least five innings of one-run ball, Heaney in a starting role. Had Nelson started the game as well, this might not have turned into a classic, but at least it would’ve been sequenced in a way that was more watchable.

Here’s what Nelson had to say on the issue: “I can’t expect them to put me into the game like that when I haven’t had the results to earn it. It’s frustration toward myself because I know what I can do, and I haven’t shown it to earn that trust.”

Between this answer, and Gallen and Merrill Kelly’s forthrightness about their own ups and downs this postseason, the Diamondbacks seem to have assembled one of the most candid, plainspoken pitching staffs in the league. It’s pretty great.

But the point is this: Nelson was only in this game because it was a mopup opportunity. In the regular season, he made 29 appearances, 27 of them starts, and over 144 innings he had a K-BB% of 8.1 and an ERA- of 121. In his previous playoff appearance, he allowed three earned runs in 2/3 of an inning. You’d have to be out of your mind to start a guy with that track record in a World Series game, and Nelson knows it.

The same is true of Heaney, who was good during the regular season but had never before faced more than 14 batters in a playoff appearance. This year, he was solid one-and-a-half times through the order in his first postseason start against Baltimore, got wrecked in less than an inning in his second, and has been a mopup guy ever since. Bochy, a manager who was a couple hours from inserting his hideously overworked closer in a six-run game, would never have let Heaney throw five innings if the game was close.

One reason I keep coming back to Tyler’s piece is because of what he doesn’t say. The bullpen game is time-consuming, it flies in the face of tradition, and it sets up a house of cards that can often result in the disaster we saw in Game 4. It’s ugly. If I’m being completely candid, it is not baseball as I would like it to be played.

But he never said it was the wrong move on Lovullo’s part.

To paraphrase MLB’s Mike Petriello, this is how teams have to make do when there are not enough good pitchers to go around.

It wasn’t too long ago that teams would just throw their no. 4 starter out there like normal for a critical playoff game. The star-studded mid-2000s St. Louis Cardinals threw Jeff Suppan in Game 7 of the NLCS twice in a three-year span. The second time his opponent was a young Oliver Pérez, who’d posted an ERA of 6.55 that year.

Both Suppan and Pérez pitched well that night, but it’s wild that we ever lived like this, going into the most important games of the year and handing over more than half of it to pitchers who had trouble getting anyone out.

The state of the art seems to be that using more pitchers in shorter stints is the way to go, even if you get the occasional stinker from your fourth-best reliever. And it’s amazing how extreme that trend has gotten in the past 10 years. Even in 2016, when Terry Francona went bullpen-mad in order to white-knuckle an injury-riddled Cleveland team to the playoffs, things weren’t this extreme. Except for cases of injury or extreme ineffectiveness, every Cleveland starter that postseason either went through the lineup at least twice or pitched into the fifth inning. Francona used Corey Kluber on three days’ rest and let Josh Tomlin work into the late innings. He even called on a babyfaced rookie named Ryan Merritt, with all of one big league start under his belt at the time, to pitch into the fifth inning in the ALCS clincher. He never came close to doing what Lovullo did twice this postseason.

There are, at minimum, 150 rotation spots available in the majors at a time: five per team, 30 teams. That’s not accounting for six-man rotations, injuries, or other contingencies. With 12 playoff teams running four-man rotations in October, that leaves 48 starting rotation spots up for grabs in the playoffs.

I am not sure there are 48 starters in the entire league who could hold up for a full season and turn over a lineup more than twice against playoff lineups. Starting pitching is so demanding now, the margins so small, the risk of injury so high, that almost nobody in the league has that combination of stuff, nerve and stamina.

Only 44 starters qualified for the ERA title this year, and of those, 33 had an ERA- under 100. Two of those pitchers, Dean Kremer and Bryce Elder, actually made exactly one postseason start this year and ended up with an ERA over 20.00. Sometimes a rookie with great stuff puts up horrible numbers early in his career before figuring things out in time for the playoffs. That worked out great for Brandon Pfaadt, not so much for Grayson Rodriguez. Then there’s Jacob deGrom, Walker Buehler, Sandy Alcantara, and a number of other top-end starters who were excluded from this postseason by injury.

It’s difficult to even identify the trustworthy playoff starters from season to season or even game to game. Framber Valdez, who was lights-out last postseason, allowed 13 runs in 12 innings over three starts this year. Cristian Javier, who threw the pointy end of a no-hitter in last year’s World Series, got knocked out after a third of an inning in Game 7 of the ALCS. Managers just don’t know whom they can count on anymore.

Regular-season starter and postseason starter are almost different jobs at this point. The Rangers signed Heaney this past offseason, paid him $12.5 million, and didn’t use him much more liberally than the Diamondbacks did Nelson. The Phillies signed Taijuan Walker to a four-year, $72 million contract last winter and got a solid regular season out of him. He didn’t throw a single pitch in the playoffs.

With eight- or even nine-man bullpens, that leaves what we saw on Tuesday as the best option when teams run out of quality arms. It’s not traditional or pretty, but it gives them the best chance of winning.

And I wish it weren’t the case. I don’t know how to change this, short of hopping in a time machine and preventing Candy Cummings’ parents from ever meeting. Maybe there’s a roster rule change that would prevent the cavalcade of relievers without causing further harm to pitchers’ health, but I don’t have an easy answer.

Until then, it is how it is. Bemoaning the bullpen game is in the same bucket of grievances as complaining about teams’ reliance on home runs in the face of overwhelming evidence that the best path to victory goes over the fence. It’s something that’s going to happen if the teams that play under current conditions are playing to win.





Michael is a writer at FanGraphs. Previously, he was a staff writer at The Ringer and D1Baseball, and his work has appeared at Grantland, Baseball Prospectus, The Atlantic, ESPN.com, and various ill-remembered Phillies blogs. Follow him on Twitter, if you must, @MichaelBaumann.

37 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
mariodegenzgz
4 months ago

Great piece. As you said, getting mad at the teams over this is ridiculous. Their job is not to worry about aesthetics, it’s to win as many games as possible, as efficiently as possible. They’re going to push the limits of the rulebook with no regard for anything else. The avalanche of position players pitching in recent years should be a clear indicator of that.

If people want to get mad, get mad at MLB for being so slow to react over a decade ago and allowing teams to distort their product. Its possible the cat’s out the bag now, though I still hold hope that it isn’t. The fact that both Manfred and the MLBPA seem to agree on the idea of reducing the size of the pitching staff is a good early sign, but I’m not holding my breath here.

68FCmember
4 months ago
Reply to  mariodegenzgz

How sure are we that bullpen games are actually more efficient? I get it for the D-backs when their 4th starter is Ryne Nelson who has been significantly worse than any of the pitchers who would be expected to make up the bullpen game.

But for the Rangers or Phillies, Heaney, Dunning, Gray, and Walker were all around average starters. Is an average starter really all that worse than a bunch of good to average relievers, the best of whom have been worked hard and all of them have faced the opponent multiple times in the series?

Those kind of pitchers should generally be expected to be able to give you 5-6 innings allowing 3ish runs. They were all better than that in the regular season. Is that really worse than you expect to get out of the underbelly of the bullpen game? Plus it saves the relievers that would been pitching both the fatigue and the exposure if they are needed later.

hglmanmember
4 months ago
Reply to  68FC

Were they better than that vs the best teams in the League? There is a huge difference in terms of bat quality, game stress, etc., in a playoff game. Relievers are generally much better; if you have to win a single game, you should do a bullpen game. Maybe you can make a strategic argument for trying to get 5 out of the 4th guy if you’re up, but that seems likely to be over-optimization.