There’s a little-known secret which surrounds relief pitchers, and it goes like this:
Most of them aren’t very good.
Good pitchers usually remain in the rotation. Teams guide them and protect them, treating them like the crystalline figures they are. Those that don’t work out? Off to the bullpen they go, where they will be thrown into blenders and asked to stop the blades until they either somehow become a diamond or are tossed into the great waiver wire/compost bin in the sky. The reason for this is simple: A healthy reliever will throw roughly one third as many innings in a season as a healthy starter. Teams, believe it or not, give more innings to better pitchers.
It’s harder to remember this these days. Teams have figured out how to use their bullpens as toolboxes more than scrap heaps. Over the last almost–decade (since the start of 2015), relievers have allowed fewer runs per inning than starters. It’s close, but it’s true. Teams have learned how to best leverage their relief arms, turning every member of a bullpen into a specialist. But while bullpens as a whole are often better at preventing runs than rotations are, individual relievers are not as good as individual starting pitchers. Going by fWAR, the top 53 pitchers of the last almost–decade are all starters.
There are exceptions to the rule that relievers are bad. Some relievers—Josh Hader and Joe Kelly among them—were initially moved to the bullpen because the rotation was already full. But most relievers failed as starting pitchers. Some, like Kenley Jansen, failed as position players.
Why don’t people notice this? Because most relievers don’t stick around. Since the beginning of 2015, 2,241 pitchers have made an appearance out of the bullpen. The median number of innings thrown by those within that cohort? 29. This isn’t precise, but the average relief pitcher throws around 30 major league innings before being DFA’d for the last time.
Most relievers aren’t very good.
Another secret:
The measure of a good reliever is not his single-season stats. It’s his career body of work.
Single seasons are small for relievers. If you broke Joe Kelly’s down into a series of starts, it would look like this:
- 4.2-inning outing to open the year. Six runs allowed, five earned. Bad start, but settled down after the second inning.
- 5.1-inning outing his next time out. Two runs, both earned. Not a quality start, but probably a win.
- 3.1-inning start before exiting with injury. No runs allowed.
- IL Stint
- Six-inning start in return from IL. Two runs allowed, both earned. Quality start. Yielded his first home run of the year in the final frame.
- 4.2-inning outing his next time out. Currently ongoing, halfway through his third trip through the order. Four runs allowed so far, all earned. Not going amazing, but his team is in the game.
What does that sample tell you about this hypothetical starting pitcher that is Joe Kelly?
Nothing!
Absolutely nothing.
Maybe he’s having a good year, maybe it’s bad. His 4.88 ERA isn’t great, but he’s got a 4.08 FIP and 4.10 xERA, so there’s probably been a little bad luck involved. You’d keep this guy in the rotation, hope he got out of this ongoing fifth inning cleanly, and look forward to seeing what he could do in his sixth outing of the year. There would not be wailing in the streets on days he started.
I bring all this up because our moles within the Dodgers fanbase have detected an anti-Joe Kelly movement in the wake of the recent home runs. Some even expressed outrage that Joe Kelly wasn’t DFA’d in place of Brent Honeywell, a man who’s been below replacement-level his entire career. (No hate towards Honeywell. He had some big innings in there, and I wouldn’t mind my favorite team giving him a shot.) This is stupid. This is very, very stupid. Dodgers fans complaining about Joe Kelly are fundamentally stupid people. It’s not their fault, and many are not beyond saving. But they are dumb. They are dumb, dumb idiots.
The best way to tolerate the reliever roller coaster is to tell yourself it’s an experience. Tell yourself relievers go hot and cold. This isn’t really true—the small samples make things look like hot streaks and cold streaks that would again, for starters, all take place in the same night—but it’s the best way to keep yourself from becoming a dumb, dumb idiot who complains about one of the best bullpen arms your generation has seen. “Lights out” relievers are rarely lights out for long, and truly bad relievers are usually cut. Those who remain, like Joe Kelly, remain because they’re good. This might surprise some of you, but the people who run the winningest team in baseball since 2006 know what they’re doing.
Joe Kelly pitched last night. It went well. He made old friend Justin Turner line out to right field, he made Josh Rojas ground out on a little chopper to second base, and then he froze Dominic Canzone with a fastball down and in to finish the eighth inning. The Dodgers, invigorated by the sight of Joe Kelly warming up in their bullpen, scored three runs right before he entered, grabbing the lead. Joe Kelly preserved that lead before Evan Phillips finished things off.
A bunch of Dodgers fans moaned the whole way through.
**
Joe Kelly only threw 13 pitches last night, and his 27 pitches over the last week are nearly the lowest number on the team. It wouldn’t be a surprise to see him back out there tonight, although it’s possible all the pairs of back-to-back outings led to some of the home run troubles these last three weeks. Our guy looks sharp. He’s a good reliever. A very good one, in fact. Since 2012, his first season in the majors, he’s 71st in reliever fWAR among the 646 pitchers good enough to reach the qualified innings threshold. Over his last 100 innings, he has a 3.17 FIP. That’s almost identical to Josh Hader’s.
When he enters, Dodgers fans? Don’t wail. Don’t complain. Don’t act like your manager just called upon Josh Osich to keep the game alive. Keep your head about you. Maintain perspective. Don’t be a dumb, dumb idiot.
