Short Week: The Stupid Baseball Hypothetical I've Chewed On For a Decade
this is probably not as interesting as I think it is
This is the fourth post in (the first annual?) Short Week at freddiedeboer.substack.com. Since people constantly complain that my stuff is too long, this week all posts will be 500 words or less. We will return to our usual longwinded ways next week.
In the early 2010s I was in a sim baseball league for a few years. A sim league is not a fantasy league; fantasy leagues are based on real-life results, while sim leagues use nerdy computer programs to simulate every pitch of a baseball season. You assemble a team through a draft or auction, trade players and call them up from the minors, and can choose some basic strategies. After that the computer sims the results one game at a time. This was a massive undertaking - we had major league, AAA, and single A rosters to manage - and after a few seasons I gave my team over to somebody else who liked getting into the weeds more than I did. (Despite getting Mike Trout, Nolan Arenado, and Francisco Lindor as prospects through my savvy GMing.) I realized at some point that staring bleary eyed at a spreadsheet to find Vincente Padilla’s WHIP wasn’t actually fun, for me, where it seemed very fun for others. So I stopped playing. These days I barely follow baseball so it was for the best.
But I did think about baseball a lot in those days, and I dreamed up a hypothetical that’s probably not very clever but that I think about all the time. Here’s the scenario:
Imagine a pitcher who gives up one and only one run in every inning in which he appears.
That’s really it. If we must add a little more, I’ll note that this run can be either earned or unearned if you’re bringing him on in relief. Let’s call him replacement-level at batting for a pitcher. He has a rubber arm and can pitch as many innings as you want, again always surrendering a single run each inning. So: how valuable is he?
In a traditional sense, he’s pretty damn bad. He’d have an ERA somewhere near 9, depending on whether batters are on base when he gets up, which is more than 3 runs worse than the worst ERA in MLB in 2021. As a starter? This past season the team with the highest runs per game, the Houston Cheaters, only scored 5.41 runs per game, so trotting him out there seems like a really bad idea. And yet he’s situationally one of the most useful relievers I can imagine; what manager would not be thrilled to know that he can guarantee only giving up one run in a crucial inning? Also, in any game in which the team leads by more runs than there are innings remaining victory is guaranteed. And he’s an innings-eater.
What’s interesting is that I have posed this question to many serious baseball fans and gotten very different answers. Some argue that he’s not worthy of a roster spot; some say that he’s a borderline All Star. This is most likely not a very clever scenario, but still - what do you think?
I agree with the general sentiment here that he's very valuable. But let me give you the best negative case I've heard: to value this player is to overvalue the certainty of middling performance relative to the variability of good to bad performance. As the guy put it (and I won't really do it justice), take a reliever with a 2.5 ERA who pitches 60 innings in a season. There are times when the variability of his performance will cost you a game that you can't predict, yes, and the hypothetical pitcher will only lose you a game when you choose to. But amortized across a season, the regular 2.5 guy is adding much more run probability than our guy. (I don't know if this is true.) And so the advantage of this guy is adding certainty, which we can't quantify and which doesn't add wins.
My response would be to say that baseball has a postseason, where certainty is essential because there's potentially no tomorrow. Analytics work much less well due to the small sample sizes and randomness. Billy Beane famously never won a World Series with the As due to this dynamic. So while the 2.5 guy is probably more valuable over a season, or hypothetical guy is a postseason hero.
As for how this could happen, maybe he is a brilliant pitcher with anxiety. He has been on this streak (one run per inning) for years, and every time it comes close to ending, he psychs himself out and sabotages it.
When the run comes early in the inning, he relaxes, because he already failed. Then he pitches perfectly and the rest of the players strike out. But when he starts strong, with two outs, he panics and allows a home run.
He's neurotic and miserable, and part of him wants his career to end -- but the manager loves his consistency so he keeps slogging along as a relief pitcher, year after year. Fans pick up on the pattern and give him humiliating nicknames. Reporters cover the streak, and sports blogs have trackers showing how long it has been since he allowed more than 1 run (or no runs).
Eventually, a sports columnist gets the idea that next time a team is ridiculously ahead or behind, they should sacrifice an inning on purpose just to end the streak. One day it happens: Our team is up by 10, and they send in our guy. The opponents conspire to strike out on purpose, just this once. The first two batters succeed. But the pitcher freaks out. He's not ready. He throws a sloppy pitch that hits the next guy’s bat – then does it again and again until there’s a run. The streak lives on.