Washington Nationals: When the pen isn’t mightier… (Cont.)

The Washington Nationals bullpen coughs up a vital Friday game to waste seven strong innings by Stephen Strasburg

On Friday night, Washington Nationals starter Stephen Strasburg built a 5-3 lead against the New York Mets through seven innings. Then Nats manager Dave Martinez turned the game over to his bullpen, which re-purposed that 5-3 lead into a 7-6 loss.

Strasburg had thrown 97 pitches at the point Martinez yanked him, a significant but hardly oppressive workload for that number of innings. Perhaps more pertinently, Strasburg had retired the last six batters he had faced and 10 of the previous 11.

By most standard measures, Strasburg had the situation under control. It was Martinez’s decision to entrust his lead to his bullpen that turned the outcome. In the bottom of the ninth, after Washington had padded its lead to 6-3, closer Sean Doolittle allowed the game-ending four runs on six hits.

More from Call to the Pen

• Philadelphia Phillies, ready for a stretch run, bomb St. Louis Cardinals
• Philadelphia Phillies: The 4 players on the franchise’s Mount Rushmore
• Boston Red Sox fans should be upset over Mookie Betts’ comment
• Analyzing the Boston Red Sox trade for Dave Henderson and Spike Owen
• 2023 MLB postseason likely to have a strange look without Yankees, Red Sox, Cardinals

By the standards of modern managerial strategy, Martinez’ decision to pull Strasburg for the eighth reflected nothing more than a devotion to Standard Operating Procedures. Yet perhaps it is time to question those SOPs. A look at the broader bullpen performance suggests why.

On Friday night, 30 major league starting pitchers worked a total of 164 innings, and they allowed 77 earned runs. That’s a 4.21 ERA. They were followed to the mound by 97 bullpenners who pitched 94 innings, allowing 59 runs. That’s a 5.68 ERA. Starters for Los Angeles and Seattle joined Strasburg in watching from the bench as relievers tossed away leads they had built.

Managers, in other words, replaced starters with pitchers who as a group produced an ERA that was more than a full point higher than the guys they were called upon to assist.

Something’s wrong with the SOP.

Nor was Friday especially unusual this season. For the year as a whole, the average starter’s 4.52 ERA is only marginally higher than the average reliever’s 4.46 ERA. Their OPS’s are virtually identical: .763 for starters, .750 for relievers.

Through games of Thursday, the Nationals were among 14 teams – nearly half of MLB – whose relievers had compiled a higher ERA than the team’s starters.

In Washington’s case, the difference was particularly dramatic: 3.57 for the starting staff against 6.00 for the bullpen. But they were hardly alone in that respect. Here are the starter and reliever ERAs for several other teams in divisional or wild card contention:

           Team                                  Starters  Relievers       Difference

• New York Mets                    4.02        5.09                    1.07
• Tampa Bay Rays                 3.38        3.78                     0.40
• Minnesota Twins                3.93        4.48                     0.55
• Philadelphia Phillies           4.50        4.79                    0.29
• Cincinnati Reds                    3.89        4.41                    0.52
• Arizona Diamondbacks     4.18        4.50                    0.32

Obviously, the fact that a rotation is more efficient than a bullpen does not in and of itself establish that use of the bullpen should be abandoned. It is, however, fair to ask in such circumstances whether the reflexive reliance on that bullpen in critical late-game situations is a wise strategy.

The Washington Nationals are Exhibit A in this respect. Friday night’s surrender to the Mets was the 16th time this season that Martinez had pulled a Nationals starter with his team either tied or in front, only to see the bullpen lose the game. Strikingly, the Nationals offense produced 20 runs of support for that pen in those 16 games, yet the Nats still lost because the pen allowed a breathtaking 71 runs. That’s almost exactly twice as many runs as the Nats starters had allowed (36) in those 16 games. And the relievers created that carnage in half as many innings of work.

It wasn’t as if Martinez was acting to protect a sub-standard rotation. In six of the 16 games, the starter removed from the game in favor of that bullpen was Max Scherzer, twice while holding a lead and four times in ties. Strasburg was removed three times.

Washington Nationals

The instinctive counter-argument to criticism of Martinez’s decisions would be that his pitchers would have been tiring, pitch counts would have been approaching dangerous levels, or that they should not risk giving opposing batters a third look at them. But those arguments are highly situational and often do not withstand scrutiny.

Take the six games in which Martinez pulled his ace, Scherzer, in favor of his pen. On April 14, Scherzer was lifted after 8 innings of a 3-3 tie game. He had thrown 98 pitches and retired 14 of the last 15 opposing batters. Notably, he was not removed for a pinch hitter. Wander Suero took his place on the mound in the ninth and surrendered a walk followed by a run-producing double.

On May 6, Scherzer was lifted after five innings with the score tied at 2-2. It had been a relatively labored outing, but Scherzer was working on a stretch of four hitless innings when he was lifted. The Nationals lost 5-3.

Two weeks later, Martinez pulled Scherzer after six innings with a 1-0 lead, the starter having retired six of the previous seven batters. Washington lost 6-1. In his next start, Scherzer left after six innings having retired six of the last eight hitters and leading 2-1. The Nats lost 3-2.

The problem with the “third time through” argument is that it has to be applied selectively in order to be valid; as a broad-brush caution, it stinks. For the season, the batting average of hitters facing Scherzer for a third time is .190; it’s .240 on his first trip through an order and .218 on his second trip. He allows a .564 OPS facing batters a third time, about 70 percentage points better than during his first look at the lineup.

The same, by the way, is true of Strasburg. Facing him for a third time, batters hit .202, nearly 60 points worse than when they get a second look at him. In other words, familiarity conveys an edge to Strasburg and Scherzer, not to the batters.

This is not to say that managers should never lift starters in favor of relievers. It is to say that they ought to have a legitimate, situation-based reason rather than a reflexive “that’s what we always do” rationale. Granted, that requires additional thought, and it may also require a post-game explanation for those reactionaries in the media who are used to reflexive managerial responses.

The Washington Nationals are as of this moment a wondrous example of the danger of reflexive management, for Friday’s outcome continues a season-long stretch of apparently reflexive decisions leading to reliance on a sub-performing asset, and in turn leading to a defeat. Leading the NL wild card race by a mere game and a half, they do not have the margin to rely on reflexivity.