My Superpower: OODA Loops

I’ve pushed the value of OODA loops in various posts over the years, but I don’t recall ever writing a post specifically about them. Which is weird, because they’re a superpower. So much so that I think they are at least 60% responsible for what enables me to lead a life I enjoy (the rest is some mix of luck, privilege, and high intelligence). And unlike those other things, you can learn to use OODA loops.

OODA stands for Observe, Orient, Decide, and Act. The concept was created by John Boyd. His version has a lot of details I’ll probably leave out as it’s not how I actually think of them (more on that later). The loop consists of these parts:

Observe. Notice what’s going on. Stop and pay attention to what’s happening around you. This is the skill of getting your expectations about what you are going to observe out of the way and actually noticing what’s happening in front of you. This is the skill of seeing the world with fresh eyes so you can respond to it.

Orient. Now is the time when you make sense of what you’ve observed. Apply your priors to come up with some prediction about what’s happening in the world. Use the models and context you have to make sense of things. This is where you might start to get things wrong, jump to the wrong conclusions, and so on. That’s fine, it’s all part of the process, just do the best you can.

Decide. You’re going to have to decide to do something based on the information you have, even if it’s to take the default action of continuing to observe and orient until you can come to some other decision. The decision making process usually looks like choosing from among a few options that became apparent during the orient phase, although when it’s functioning smoothly orient will suggest a single course of action to decide on: the best one you know.

Act. Now that you’ve decided what to do, do it. Don’t hesitate, you’ve already decided, just perform the action. Once you’ve done it, observe what happens and start the process all over again.

Okay, so that’s a fairly stilted way of describing OODA loops. It’s like that because some people are really bad at iterating on feedback to make things better, both on short and long time horizons, so it needs to be somewhat explicit to help them break the skills down and work on them until they become good enough at them to do it naturally.

In fact, I don’t really think about OODA loops much except as something like the Schelling concept to describe what I internally just think of as the natural feedback process:

get some input about the world
put that input into context
let the updated context suggest an action
take the action
repeat

OODA loops sound fancy, but really the process is just what you naturally do anyway, only if you know it’s what you do then you can turn the process on itself and make it better.

What do I mean? Well, you can learn to get better at each of the stages.

You can get better at noticing. This is actually pretty easy to train: you just have to stop doing anything else and pay attention. And don’t try to pay attention, just actually pay attention. For example, sip a cup of tea and just drink the tea. Watch a sunset and just notice what you’re seeing without trying to tell a story about it. This is one of the key skills being taught by meditation: sitting still and paying attention to reality as it comes into existence.

You can learn to get better at orienting. This is what rationality is great for. Most people bumble around being terrible at orienting, often because they skip so quickly past noticing to orienting that they miss information (see the previous paragraph about how to notice) or because they’re bad at updating based on the information they get. Rationality is largely about the art of getting better at updating. Use those skills to orient better.

You can learn to get better at deciding. If you’re doing orienting really well, there’s nothing to decide because you’ve already figured out what’s best to do. But if you’re not quite there, and honestly no one seems to be totally free of having to explicitly decide at least some times, you can continue to use the skills of rationality to make better decisions. Perform quick tradeoff analysis, do expected value calculations, and so on. Also, learn to trust your decisions. Once you’ve decided, then that’s what’s been decided. Second guessing gets in the way of acting with confidence and will hold you back. Whole-ass it!

You can learn to get better at acting. The saying “practice makes perfect” is tired, but it’s also true. There’s a lot of learning that goes on in our body-minds that isn’t visible to us. You generally can’t get better at something by sitting around thinking about it (or at least can only get better to some limit), you’ve got to get experience. Why? Because experience creates new observations that let you iterate by feeding the OODA loop.

You can take time to practice with each part of the loop. Maybe meditate 30 minutes to get better at observing, set a 5 minute timer to think through things you observe, make decisions using expected value estimates, and do lots of things and see what happens. It won’t necessarily be obvious at first, but over time continuous practice will infuse itself into your OODA loops, so not only do you take better actions, but you continually get better at taking better actions. As I think of it, bring kaizen into your life.

So as I say, OODA loops are basically my superpower. I think this is because they are not-so-secretly the superpower of all agentic beings, but most living beings in the world aren’t lucky enough to be able to improve their OODA loops. Humans can, but we have to learn to do it. If you get good enough at them, they become “systematized winning”.