May 31, 2017
Circumstances, Implementation, and the Slow Change of Baseball
I’d like you to stop for a moment and change everything about who you are. Don’t worry, I’m an expert. Here’s a photograph of my Ph.D. diploma. (I pulled it out of storage just for the occasion.)
Well, why haven’t you changed everything?
Again ... Doctor Carleton is speaking to you.
Last week in this space, I talked about how in baseball, when we see some sort of strategic shift, it takes about a decade for it to work its way through the game, and it’s a slow uptake. What was once a novel idea eventually becomes something that everyone does, negating whatever advantage the original team had, but it’s not an instant process. There’s value in being first.
The interesting question is why it’s a slow process. If a team sees that a strategy is working for another team, it’s likely to work for them. So why not do it? This week, we can dispense with the gory mathematical details. Let’s just assume that we’re talking about something that we know works, and the math backs it up. Why wouldn’t a team—once they read the appropriate Baseball Prospectus article, of course—simply change everything that they do?
Well, That Was a Different Set of Circumstances
I have another secret hobby that I don’t often get to share here. I am fascinated by education policy. Part of it is that I'm a child psychologist by training and I’ve done some work in schools, so I come about my interest honestly. The topic is the sort of puzzle that really gets me going, though. You need to design policies that work for kids across a range of abilities, who go to school in different buildings and different cities and different states.
You get into questions like what things should be taught, and then how they should be taught. How should we fund those schools in a way that achieves the goal of actually preparing kids for life, and how do we measure whether we did the job? The blessing (and curse) of education policy is that individual districts get to set their own policies and there are thousands of individual districts in the United States. The inevitable variety that results means that if a researcher wants to look into a specific policy, someone out there is probably using it.
The downside to all that independence and variety is that people fall in love with their own policies and there’s nothing that researchers can do to stop them. If a policy is shown to be effective elsewhere, and a district is following its opposite, the reasonable thing to do would be to reverse course. But then, reversing course on a policy means admitting that the old policy wasn’t so great. And so, people resort to a time-tested refrain. “Yeah, that wouldn’t work here.” The research was done in Massachusetts, and while this isn’t heaven, it is Iowa. The research was done in school districts with very different characteristics. They had some other program that we don’t have that is responsible for those gains. The research was done five years ago and things have changed since then.
In education policy, one of these sorts of critiques might have the benefit of being the truth. It might also be an obfuscating school board member who is more concerned with a political agenda about the way they think things work, but it’s hard to tell the difference sometimes. Maybe it’s an obfuscating school board member who happens to be right about the hole in the research.
In baseball, it’s a little harder to get away with because in all 28 cities where the game is played, the game is fairly standardized, so on-field stratagems are probably pretty transportable. If The Shift works in Houston, it should probably work in Chicago. Maybe there are some edge cases where the setting makes a difference, but they aren’t likely to be many.
Some of the more “cultural” changes, like how to approach development (should players be promoted to the next level in the minors when they are “ready” or should they be challenged slightly before they are “ready?”) might not have a one-size-fits-all approach that works. The same policy might work better in one organization than another. And honestly, when does anything not have some cultural component to it? The Shift might (or might not) work, but it’s also weird. You’re not just asking the third baseman to play in short right field. You’re asking everyone on the team to play a (slightly) different game than they're used to, and I’m sure that in some cases, there were teams that just weren’t ready for that jelly.
Culture Always Matters
Baseball is a game that is now played in a variety of languages, and even when the languages are the same, the cultures that the people speaking those languages come from can be very different. Even within the United States, there are regional variations on issues of culture, to say nothing of how culture is affected by issues like socio-economic status. We don’t often think about how these issues might affect players. For example, how the culture he was raised in views authority and how that might affect how he relates to the coaches who are trying to teach him a curveball.
One increasingly popular intervention for organizations has been to provide better food (often through a catering program) for their minor leaguers. On the surface, that sounds like an idiot-proof idea. Who could be against giving athletes, who need to be at their best physically, better fuel for their bodies? The problem isn’t the idea. The problem is going to be in the execution of that plan. It can be summed up in the question "so, what’s for dinner tonight?”
Food is a cultural product. I found that one out the first time I visited my then-girlfriend/now-wife and her Russian parents for the first time. What they served was edible (and some of it turned out to be downright delicious), but when I first saw it, I didn’t recognize it as food. Or at least, the food that my Midwestern taste buds had grown up eating. It might have been the most nutritious thing in the world, but I’m not a foodie and wasn’t raised as an adventurous eater. If I won’t eat it, does it matter if it’s nutritious? How does one design a food program that will actually get people from all over the world (sometimes literally) to eat? If no one is eating the food, is it worth spending the money on the program?
We’ll Tell Them That One Later
The details on how one implements a big strategic change, even when the changes make sense mathematically, end up being key. You have to pay attention both to the math behind them and the people in front of you. People, as much as we would love to believe differently, are not rational. The idea of a Vulcan manager or pitcher or fan, relying entirely on logic, is pleasant myth.
There are things that I wouldn’t tell my boss, even if I knew them to be true. In public health, there is a principle that when you are trying to lead someone to change their behavior, you don’t start, as you might think, with the piece that will lead to the greatest health benefit. Instead, you start with the piece that will give people a relatively quick and certain victory, even if in the grand scheme of things the victory means little to the health outcome. You don’t tell people to quit smoking cold turkey. You encourage them to go a few minutes without reaching for a cigarette, and praise them for it. The feeling of being able to achieve something is more important, because it begins to build trust and the trust is more important. You have to think about how an intervention is experienced emotionally.
I think one reason that The Shift was so slow to fully catch on is that it eventually produced a ball that was hit into a hole where a shifted fielder would have normally been playing. It also produced a few balls that were hit right at a shifted fielder, which turned would-be hits into outs, but the human brain naturally focuses on the ones that got away. That’s irrational, but that’s how people actually work.
In the past, I’ve advocated that third base coaches should send more runners because the numbers show that they are far too conservative. If a runner rounding third base with two outs has a 45 percent chance of scoring, it might seem like a suicide run to send him, but the other option for getting him home is hoping that the next batter will not make an out. The best batters do that only 40 percent of the time. It makes sense to send the runner, but consider that I’m asking the third base coach to do something where his action is more likely to lead to an out (and an out at home!) than a successful scoring play. That’s going to feel awful when it happens, and it would happen fairly soon after a third base coach tried it a few times.
I’ve also advocated for the idea of using a team’s closer in a tie game in extra innings. My best guess is that this would net a team an extra 0.5-0.75 of a win compared to how things are done now. However, closers are not infallible, and since a tie game in extra innings is, by definition, on the very edge of becoming either a loss or a game where a team has fallen behind with only three outs to redeem themselves, the closer would eventually cough one up. It doesn’t matter that the team’s fourth-best reliever, who for some reason now gets that call most of the time, had an even greater chance of giving up the lead. The emotional experience is that the manager played this new strategy ... and lost with it.
Logically, those early setbacks should be ignored. They’ll even out over time, but you have to consider how willing someone is to live with those setbacks haunting their dreams. This is why you start with the easy stuff where the person whom you’re asking to change gets to feel the sweetness of victory a few times. Best to keep that victory-to-setback ratio as high as you can. So, that means that not only should a good sabermetrician be on the lookout for ideas that might bring a team more runs, but also for the right time to share that information.
The Value of Implementation
Being the team that comes up with a brilliant idea and can implement it before everyone else has its benefits. However, there’s value in being second on the bandwagon. Yes, you’re copying, but you get the benefit of getting in on the action before 28 of your 29 competitors. Last week, I suggested that it takes about 10 years for an innovation to fully saturate the league. I think it’s worth asking why that is. It’s doubtful that teams can’t read what’s going on elsewhere once other teams start adapting some new and different strategy.
I’m guessing that a big piece of the delay in uptake is in figuring out how to actually implement it. The details matter and the details are sometimes really hard to negotiate. How can a team encourage its manager to do something that is counter to his instincts and which is wrong 45 percent of the time? Perhaps the most valuable person in helping a team innovate isn’t the person who finds the regression equation that proves that there’s a two-win hole somewhere in the universe. It’s the person who figures out the way to say, “... and here’s how you get them to do it.”