**** it, let's run some numbers!
Using
Greg Stoll's win expectancy calculator, over the past thirty years (to provide a large enough sample of outcomes to remove random noise), a team in the position we found ourselves when Valera was sent on for his fateful bunt (bottom 9, 1st and 2nd, no out, tie game) will win 83.0% of the time, while a team with what would be the optimal outcome of that bunt (2nd and 3rd, one out, tie game) will win about 81.1% of the time.
So already, while it's marginal, we're talking about a bad move. But wait, there's more! Players attempting to bunt, no matter how often or seldom they bunt,
fail to get them in fair territory about half the time on any given attempt. This means, even setting aside the fact that Soto is probably harder to bunt than most given that he throws 100 mph missiles in random directions, there is a 12.5% chance that a hitter never even gets the ball in play, assuming you are crazy enough to have them bunt with two strikes. At the beginning of the plate appearance, that will result in a 12.5% chance of your odds dropping from 83.0% down to 75.8%. So let's look at our revised odds:
Odds if not Charlie Montoyo: 83.0%.
Odds if going for broke on bunting (while being guaranteed you'll succeed if fair): 80.4%.
So, it's drifting, but it's still relatively close. But, again, that's only if there is a 100% chance that the sacrifice will be successful if fair, and we know this isn't actually the case, because Valera bunted the ball in fair territory and didn't get a successful sacrifice.
So how often are fair bunts successful?
About 70-80% of the time. Which means that, from the 87.5% of times where you do get the ball in play, there is another subsection of that where you do precisely what Valera did, and generate an out without advancing the runners, either resulting in the lead runner being out or popping up to the catcher or whatever. So, from the start of the plate appearance, we now have an overall failure rate of about 34.5%, given that you have a 12.5% chance of **** it up so badly that you strike out, and a further ~25% of the remaining 87.5% you'll fail at your job even after getting the ball in play. So now, we have:
Odds if not Charlie Montoyo: 83.0%.
Odds if going for broke on bunting (with normal accommodations for failure rate at sacrificing, but assuming no double plays): 79.3%.
But maybe he was afraid that Kirk, as a fat guy who hits the piss out of baseballs, would ground into a double play, a situation that would result in a runner on 3rd with two outs (leaving a winning percentage of 66.7%). Absolutely a possibility; Kirk hits into quite a few GIDPs. With a runner on first and less than two outs, he has gotten doubled up in 19.4% of his PAs, an extremely high rate that is also a bit of a product of sample size. That's scary! He has also produced 9 hits and 2 walks in that small sample, which means that the odds of him winning the game outright (or at least leaving you with the bases loaded and none out, which is a 93.5% winning environment) are higher than the risk of him GIDPing, which is to say that it's a risk worth taking.
So now stripping out any other game context, we're looking at about a 3.7% decrease in our odds to win the game simply by deciding to bunt. 3.7% might not look like much, but finding things that will increase your odds of winning a game by 3.7% are why you have analytics departments in the first damned place. And the game context doesn't make it better. Let's add that context!
- Gregory Soto throws really, really hard with a great deal of movement. The harder a pitcher throws, and the more movement a pitch has, the more difficult it becomes to successfully lay down a sacrifice.
- Related to that hard throwing and movement, Gregory Soto is really, really wild. He had just walked two dudes. The marginal benefit of trading an out for an advancement of the runners is diminished when there is a 14.5% chance (Soto's walk rate on the year) that the pitcher just advances them his own damned self.
- Gregory Soto, when he isn't walking people, strikes out rather a lot of them. That diminishes the marginal benefit of the 2nd and 3rd, 1 out scenario given that there is now about a 27.5% chance (Soto's K rate on the year) that the next batter won't even succeed in putting the ball in play, thus negating the advantage of having a guy who can tag up and score.
- The defense knew full damned well that the sacrifice was coming and were so close to Valera that they could tell what he had for lunch from his breath. The odds of completing a sacrifice successfully include a lot of scenarios where the defense might think a sac bunt is possible, even probable, but it's well short of guaranteed. This one was guaranteed. They could position themselves optimally for the bunt without hedging in case he were to swing away, because the second Montoyo sent Valera out there, everyone on the damned continent knew a bunt was coming. It's virtually assured that the odds of getting the sacrifice were lower than the norm.
- Valera isn't even a prolific bunter nor notably successful. He has three sac bunts in his career and two bunt hits, in what appears to be seven total bunts in fair territory. 28.5% of the time, when he has bunted, nothing remotely productive game of it.
- I mean, come **** on, you just don't PH Breyvic Valera for anyone other than a relief pitcher. He can't hit worth a damn, and contrary to popular belief being a terrible hitter does not automatically make you a good bunter. Sometimes, you're just not good at baseball! Breyvic Valera is not good at baseball.
So this is a bad decision in a context-neutral environment, gets worse when factoring in environment, and
we still haven't even factored in the part where he was asked to do it with two strikes.
There is a reason why, when analytics became a thing, the first casualty of the nerds was the sacrifice bunt: it's a relatively easy thing to run the numbers on sac bunts because there are fewer variables and outcomes than with a traditional plate appearance, and the numbers stated rather clearly that there are not many situations where a sacrifice bunt with a position player is a good idea. Outs are valuable, and their value relative to base position increases in a high-power, high-strikeout environment, because for instance the odds of scoring a runner from 1st (via XBH) increase, while the odds of scoring a runner from 2nd or 3rd decrease given that more plate appearances while end without the runners even having a chance to advance.
It strains credulity -- badly enough that Merryweather is expected to miss all of 2022 owing to his credulity strain -- that the people whose livelihood depends on telling people what statistically-sound plays are instead forced Charlie to make such a statistically questionable choice.
And that still doesn't even touch on the part where, with everyone in the stadium knowing that Valera would bunt, and a 50% chance that he'd simply strike out, and a further 25%ish chance that he'd fail to succeed at the sacrifice, that Charlie **** Montoyo still asked Valera to **** bunt.