Feedback loops teach better than theory

let’s define these terms

  • teach: Changing of behaviour in a way deemed an improvement, from the subjective perspective. This might mean to teach in the sense of helping improve a skill. It can also mean to modify behaviour for the good, or alleged good, of an external group.

  • theory: Information gathered by others on their experiences performing the same task or function, in this case for the aim of instructing or guiding others.

Using theory

Study will give you a goal, and perhaps a starting point. It won’t tell you what you’re doing wrong and how to correct it.

Study gives you one approach, proposed by somebody else. There are many ways to do something. You should usually try and find the easiest and most natural to you.

No matter how much information you gather, at some point you need to try it out. You can’t learn to play tennis by reading a book. Though you might be able to recite the desired technique, implementing it, and making it work, requires practice.

And once you’ve started, you need to fix problems to improve. Theory can be a slow way of diagnosing an issue. It relies on your ability to know the root of the issue, and that someone else has suffered the same problem, for the same reason as you.

Usually, to find the solution to your problem, you need to be able to express it clearly. This can be difficult, especially in a domain you’re new to. It also relies on your ability to action their recommendation. Compare Googling “why the tennis ball goes wide” to hitting hundreds of balls, adjusting your shot, with the help of a friend or coach, rapidly iterating as you identify problems.

Whether it’s tennis or learning to drive or any other pursuit, there is often a difference between what you think you are doing, and what you are actually doing. Theory cannot help us here either.

Using feedback

Once you start to practice, you need a way to diagnose problems effectively, and then fix them. Good feedback will tell you what’s wrong. Embracing feedback loops will empower you to quickly iterate on continued feedback to reach a desired outcome. Through rapid trial and error, you can efficiently develop.

Unlike theory, trial and error will encourage you to develop a method that works naturally for you. Rather than consideration of the best approach, repetitions with a strong feedback loop help you quickly internalise effective ability. Deliberate practice is high reps with a good feedback loop.

Josh Waitzkin talks about how in Chess feedback loops have removed much of the need for theory. Machines can give them move-by-move feedback, so players don’t need to study ahead about why a move might be wrong, when they can internalise that through play with specific, accurate & tight feedback.

Only experimental physicists publish findings

Rather than pondering what the best approach is, dive in and test. Get quick feedback, and improve. Get going, to get good. This is the scientific method. Leave theory to theoretical physicists, and define tests in the field, like their experimental colleagues.

Unlike theorisers, by the time you’ve figured out the best way, you’re already most of the way there.

How to design good feedback loops

If feedback loops are more valuable than theory, then you ought to dedicate more time to designing and improving them. Good feedback loops have the following attributes:

  • good feedback loops are tight
  • good feedback is highly specific
  • good feedback is accurate

I’ll go deep into each of these in upcoming posts.

Occasionally I send out an idea & ask for your thoughts.