Fast-forward five years or so. And 17 seconds. MLB has entered the brave new world of the pitch clock. Ryan Webb is standing on the mound, finishing out yet another 8-3 game in which he will not get a save, a fact which will be discussed at length on Effectively Wild podcast #1595 (a Friday episode, naturally).
Webb looks in for the sign, but he’s having trouble coming to an agreement with his catcher. Webb wants to go fastball here, but his catcher is thinking a changeup would do the trick. Webb relents and nods to the changeup and gets himself mentally ready to throw the pitch. He feels a little twinge in his upper arm. Nothing major. A tiny cramp. It’s August. He wiggles his arm and the twinge goes away. Then he starts to wonder, why is it that I never get to pitch in the save situations? Why anything? Why am I here in the middle of this field? "Save." What a funny word when you say it over and over again.
But now he realizes that he has a problem. With all of the pitcher-catcher negotiation, the setting himself, the quick muscle maintenance, and the existential angst, the pitch clock has snuck up on him. It’s at 3 seconds. Webb needs to hurry up, but the twinge knocked him off his train of thought mentally. He feels like he’s not completely ready to throw the pitch, but if he doesn’t, the umpire will call a ball. Still, a ball is better than giving up a home run, even if this isn’t a save situation.
Article continues below ...
What should Ryan Webb do?
It’s easy to say he should throw the pitch and avoid the penalty, but the penalty is only a ball. What if his mechanics are a little out of tune and he hangs that changeup? Something much worse than a ball could happen. Maybe it’s just better that he takes the ball, gets himself set and throws the pitch that he wants to. Welcome to the wonderful world of pitch clock management! It doesn’t exist yet, but it’s looking more likely that it will.
The Arizona Fall League experimented with the pitch clock this past autumn, and MiLB will introduce the clock in Double- and Triple-A stadiums this year. New Commissioner Rob Manfred is a fan of using clocks to speed up the game, so it seems a matter of time before someone is having that discussion at the MLB level. Teams might want to start planning now for what seems inevitable and start teaching their minor leaguers. How to build a between-pitch routine that lasts less than 20 seconds. Where the escape hatches in the rule are. And most importantly, what to do in a crisis, like the clock winding down.
There are going to be a few knock-on consequences that the clock rule will have to deal with. One is that if there’s a runner on first, an easy way out is to make a sham throw over to first to "check on" the runner — or just step off the rubber and look menacingly over at him. The balk rule might be changed that if the pitcher steps off, he would need to make a throw, but you can’t outlaw those throws without changing the dynamic of the running game. If the rule is going to have any teeth and actually make the game shorter, baseball is going to have to address the workarounds. Or abandon the project. But somehow, I doubt that baseball has come all this way to let it die so easily.
At some point, someone, maybe even Ryan Webb, is going to be faced with this scenario. Should he throw the ball or hold it?
Warning! Gory Mathematical Details Ahead!
Well, figuring this one out isn’t all that hard. Let’s talk about the value of an automatic called ball. Obviously, the only real effect that it has is changing the count, but we know that the count can make a big difference. Ever wonder why this catcher framing thing became a big deal? Balls and strikes matter a lot.
For example, if a plate appearance starts out 0-1, that plate appearance is about 10 percentage points more likely to end in a strikeout than one that starts with a 1-0 count. If a count goes to 1-0, the pitcher might come back and get to a 1-1 count, but he has no chance at an 0-2 count during that plate appearance. In 2014, plate appearances which featured a 0-1 count resulted in a slash line of .221/.261/.331, while those which started 1-0 resulted in a line of .267/.373/.420. Some of that is a function of the fact that better hitters are more likely to find themselves ahead in the count (because opposing pitchers are nibbling?), but the implication is pretty clear. Getting ahead in the count is worth a lot.
In a similar way that we use to get an expected value for a base-out state — runners on second and third with one out — we can get the expected value for a count. A team that has second and third with one out might not score for the rest of the inning. It might score five runs, but we take the average of all the times that it happened. Similarly, a 1-1 count might lead to a strikeout on a 2-2 pitch or a home run on a full count or a groundball to third on that very next pitch, but we look at all events. Using a linear weights approach, we assign a value to each event, we can find the average number of runs produced by a hitter after a 1-1 count.
I did that for 2014 for all 12 count possibilities. In 2014, the pitch clock rule wasn’t in effect, and we assume that most pitches were thrown when the pitchers were completely ready to throw them and weren’t feeling hurried. So, we’ll take that as our unhurried baseline. With that pitch, they might have thrown a ball, gotten a strike somehow, or the ball could have been put into play. (On a two-strike pitch, there could also be a foul ball.) Still, we’ve accounted for what that might eventually lead to.
If the pitcher finds himself running short on time, though, he might end up taking an automatic ball. So, let’s take a look at that decision on a 1-1 count. If our pitcher takes the automatic ball, he accepts a 2-1 count. In 2014, plate appearances that reached a 2-1 count had an expected run value of about 0.1346 total runs. If the pitcher had made that 1-1 pitch, we would have expected a value of 0.0901 total runs, for a difference of .0445 runs. By accepting a ball, a pitcher concedes four-hundredths of a run to the batter — assuming that he was a league-average pitcher facing a league-average hitter — over what we expect he would have done had he just thrown the pitch he was supposed to.
Here’s the "penalty" of a ball at each count:
Penalty (in expected runs)
The penalties are actually rather small early in the count. They aren’t zero, but especially on 0-2, it’s not much. The pitcher usually wastes that pitch since he’s ahead in the count, maybe trying to set something else up, and it often goes for a ball anyway. On 3-0, an automatic ball leads to a walk, but it’s likely that the pitcher is in big trouble anyway. Chances are the batter would have walked anyway and frankly, the pitcher is going to have to throw some nice contact-friendly pitches to battle back from this hole. A walk does hurt him to the tune of roughly an eighth of a run, but it’s not an absolute nightmare.
Now, while the penalties are small, one might ask why a pitcher would ever willingly accept a penalty. The trick here lies in strategic thinking. A pitcher looking at a pitch clock at 3 is in a bad spot. In a perfect world, he never would have gotten to this point, but here he is and he must choose between these two imperfect options. The automatic ball is going to hurt him somewhat, but taking it does buy him something. It buys him the ability to not have to make a hurried pitch. Sometimes, you have to pick the option that does the least damage to your cause.
But let’s talk about the value of not being hurried. We have almost no empirical data on what effects that might have on a pitcher. There are Pitch FX data available for the 2014 AFL, but the pitch clock data are not embedded in there and it was at only one stadium. Pitching is such a high-wire act to begin with that the effect most certainly can’t be positive.
For one, we know that pitchers tend to "live on the black," referring to the outer edges of the plate. Go further out, it’s a ball. Get too much of the plate, the batter can get good contact. Pitchers spend a lot of time working a very small area. If we assume that a hurried pitcher might suffer in his location or his movement (the pitch doesn’t break all the way out to the "good" zone), then we have a couple of possibilities. One is that the pitch will go well out of the strike zone for a ball, and the penalty for a pitch clock violation is a ball anyway. But there’s also the possibility that the pitcher will not execute the pitch properly and it will stay out over the plate, begging Giancarlo Stanton to hit it 500 feet.
In 2014, there were 688,102 non-intentional balls thrown. There were 13,172 doubles, triples, and home runs hit. That means that despite having all the time he wanted, the average pitcher saw just under 2 percent (1.9) of his pitches go for extra bases. Yikes. That’s when he can settle himself totally. The average double or triple or home run was worth 1.06 runs in 2014. So, with that information, we can at least solve for how much additional extra-base risk a pitcher should be willing to assume when he makes that pitch.
On an 0-2 count, the penalty of the automatic ball is .0200 runs. Our pitcher can take that or he can run some additional risk of a 1.06 run event. A little quick algebra:
.0200 = 1.06 * p
p = 1.88 percent
If the pitcher believes that it is 2 percentage points more likely than normal — roughly double the rate that it is now in the unhurried world — that he will hang a rushed pitch on an 0-2 count, he should simply stand down and live to fight on at 1-2. In reality, I’m probably under-selling this risk, because there is additional risk that a rushed pitch will hit the batter or become a single where it would have been an out, but let’s just stick with this.
On a 3-2 pitch, that break-even point jumps to 22.3 percent, which means that our pitcher would have to think himself one-quarter likely to hang one to the batter (and that the batter would crush it) before he would simply allow a walk without a fight. At least if he’s playing the game correctly. Men in their early 20s are not known for making fantastic strategic decisions. The urge to prove that he can do anything might overtake him.
The problem is that we don’t really know what effects hearing the crowd yelling "3, 2, 1 …" will do to a pitcher. Will the effect vary from player to player? (Probably.) Would the effect be a trifle? A major problem? Baseball has never had a pitch clock before. We have very little in the way of a starting point. Neurologically, we know that pitchers are likely drawn from the extreme upper end of the distribution in terms of their motor execution and motor coordination. A pitch requires a complex sequence of movements with just about every body part getting involved. But the game’s ecosystem has never selected pitchers for processing speed or the ability to make fine motor movements in the presence of a time-based stress reaction.
So What Should Ryan Webb Do? What Will He Do?
I don’t know whether anyone has a good answer to the question of how likely it is that a hurried pitch would become a hanging pitch. Even if a pitcher wasn’t totally ready to throw a changeup, he does have the benefit of having done this a few thousand times before. Muscle memory is a nice fallback. Still, we know that time pressure increases anxiety and anxiety increases interference in procedural memory. So, that’s going to cause some problems. Maybe the lion’s share of the "errors" that the pitch clock causes will be in the form of balls outside, rather than balls that catch too much of the plate, but it seems likely that the risk of hanging one goes up. How much will have some effects on game strategy.
We need to stop for a moment and take stock of where the game is going neurologically. For a long time, people have romanced the idea that baseball didn’t have a clock. Clocks do weird things to people, even when they’re arbitrary clocks. Ever been tempted by a sale at a department store that’s "limited time only," despite the fact that you know they’ll be having another sale next week to celebrate Bastille Day or Arbor Day or Groundhog Day? Teams have never had to think about how to talk to their young players about clocks, nor confront these issues, because it was never relevant.
Beyond that, a clock is somewhat antithetical to a game that requires intense periods of low-stimulation concentration (read: lots of "boring" down time.) Maybe one of the reasons that there isn’t a pitch every 20 seconds is because the brain needs resting periods between bursts of concentration. Try to push the button too many times too fast and the circuitry will wear out. It might lead to pitchers being exhausted on the mound and making mistakes even when they aren’t under time pressure. Maybe that increases scoring and makes the game more exciting, but it most certainly is a fundamental change.
Coming back to practical matters, teams are going to need to start preparing their pitchers to handle this sort of a situation. The data here suggest that they should instruct pitchers to take the penalty when well ahead in the count, and then do some research on what generally happens. It’s possible that there are several counts where taking a penalty makes sense. Funny that a rule intended to get the pitcher to stop messing around and pitch might actually cause him to take a little extra time between some pitches.
Teams are also going to have to start having a discussion about strategic thinking with their pitchers. There might be cases where the "correct" course of action is to hold the pitch, perhaps even ceding a walk. Humans commonly make a mistake of feeling the need to "do something" in the presence of a winding down clock, and the obvious thing for a pitcher to do is to throw a pitch. Young men don’t like to be dictated to by a silly clock. It’s not manly to give up ground without a fight. All of these issues come into play because of the introduction of a clock.
The pitch clock might actually do what it says on the label and make games shorter, but it will also introduce issues that baseball has never confronted before. It’s easy to think that the only thing that will change will be the time of game. I doubt that.