What about position players pitching with a huge lead?
The Indians were still a long way from wrapping up a 17-4 pounding of the Twins on Saturday when the speculation began among fans watching and writers covering the pound-ee side.
Who would be the position player who got the once-in-a-lifetime chance to walk out to the bullpen and see how the other half earns its more precarious living, albeit pitching with no pressure given the infinitesimal leverage of a blowout of such magnitude? Mike Berardino, Twins beat writer for the St. Paul Pioneer-Press, called his shot early and nailed it.
There was no speculation on the other side, though. Who would be the position player to march out there for Cleveland and protect the 13-run cushion that you don’t really need a pitcher to protect? There would be no reason to speculate.
Two hundred and eighty-six times in baseball’s recorded history, a position player has taken the mound to pitch – not counting the Brooks Kieschnick-esque two-way types and anyone else who pitched more than five times in his career. Not one of them came in when his team was leading.
The most obvious reason beyond the chances of a comeback, which might not be as realistic as common wisdom might suggest, is that the team that’s leading isn’t likely the team that just got a short performance out of its starting pitcher. If you’re giving up double-digit runs, you’re likely to have put some stress on your bullpen.
However, in this era of big bullpens, position players are used in situations that fall far short of dire need. The Twins, according to Berardino, still had four relievers left when they handed the ball to Robinson, who flashed a knuckleball in pitching a scoreless inning around a walk.
Early this season, Russell Carleton wrote a piece for JABO titled “When Should A Team Give Up” that talked about and went into the gory math on many aspects of the decision to bring in a position player to mop up a game that’s gone so far wrong. Among the benefits of the white flag: saving a reliever from pitching back-to-back days (the second day being tomorrow’s game), given the scholarship that shows that in that second game, velocity will on average be diminished. So why throw him today in a game you can’t win when you might need him to be at full strength tomorrow?
Additionally, even if you knew he wouldn’t be needed tomorrow, why use the bullets, stress out the ligaments, when we know that repeated use is an issue for pitchers?
Russell does a terrific job on the math of win probability and showing that teams are often chasing tiny chances of winning with their real relievers, but let’s get back to today’s topic at hand. Is the question really that much different when substituting out “a game you can’t win” for “a game you can’t lose?” And is there such a thing as a game you can’t lose if you have a position player on the mound?
To see just how out-of-reach a game would have to be for a position player on the mound to be reasonably safe, the win probability metrics won’t help, so we sort of have to create our own using some assumptions. The assumptions will generally skew conservative, because if you’re gonna do something weird and be the first to do something weird, you better be pretty darn sure.
Against the 286 moonlighting position players, hitters have a .387 on-base percentage, something that’s going to have some fluctuation among eras, but is perhaps a misleading stat for another reason. Many of those are in blowouts where the team ahead might not be trying all that hard. In the interest of being conservative, I narrowed down the sample to just those where the position player was pitching in extra innings because a team really ran out of arms, rather than the blowouts. Not surprisingly, hitters hit better against position players in the tight games, so if they do sense a comeback, this is a more sensible and always more conservative number to use.
Hitters have a .454 on-base percentage in the much smaller sample of extra innings and a .456 OBP in extra-inning games where the position player came in with the game tied.
Take the 13-run game like the one Cleveland sought to finish off on Saturday against the Twins. Assuming the .456 worst-case scenario and assuming that the last hit is a home run, what are the chances that given one inning of work, the average position player pitcher would give up 13 runs (in this case, 13 baserunners) before recording three outs?
Of course, this is assuming that the manager would be sitting on his hands watching and hoping for the hitter to hit it at somebody. In reality, when it reached a certain critical point, he’d certainly start to get a real reliever ready and at some other point, use that guy he would have used anyway.So if a team is up by 13, there’s a 1 in 775 chance that the other team would, not even win, but just come back to tie.
So let’s say the team decides that if the lead ever shrinks to five – or more precisely if they were in a situation where if everyone on the bases scored, the lead would shrink to five – that they’d bring in the real reliever. The win probability for a team pitching the bottom of the ninth with a five-run lead fluctuates year-to-year but generally dances around 99 percent. So we can redo that table in two parts, where the trailing team needs to both get the deficit down to five and then come back from down five against the reliever (we’ll even be overly conservative again and assume no outs were recorded to get the 99 percent).
Move the threshold to six or seven runs and it gets that much more unlikely. And so on. Any math on this is going to show that the chances of blowing that lead are tiny and that many of the reasons for pitching a position player in a blowout loss still apply.
There’s the usual “don’t get fired by doing something unconventional” argument here. If you blow a 12-run lead in an inning you started with a position player, it’s going to look a whole lot worse than if you are down by 12, have your position player give up two more in the top of the ninth and then score 13 in the bottom of the ninth to fall one run short.
But there’s also an unwritten rule at play here. Baseball has few unwritten rules closer to being codified than the one that says if you’re blowing out a team, you shouldn’t try too hard. Don’t steal up by 12 runs. Don’t play the platoon advantage, hit and run, bunt, etc. But there’s a corollary to that in sports that says don’t make it too obvious that you’re not trying. Don’t jog the last 100 meters. Don’t start kneeling on the ball with 10 minutes left.
And probably don’t throw a position player with a double-digit lead.
In a world where the goal is the World Series and nothing matters en route to that goal, though, it’s not a bad idea. Certainly a better one than the zero times it’s ever been tried would indicate.
