Why feedback trumps theory
Originally published @ samxsmith.com
Feedback loops are a naturally occurring phenomena. Homeostasis is achieved in the body with a number of feedback loops that alter our physiology, and our behaviour. When we get cold, we shiver to create heat energy, and sweat when it’s too cold. When we are scared we get a hit of adrenaline that raises our heart rate ready to defend ourselves or run. These processes respond to an environment or state, and provide a reaction to correct for it.
We can make use of our own feedback loops to correct and improve, to learn and test.
Feedback loops are worth your time
I dedicate a disproportionate portion of time to feedback loop design, a specific kind of system design where the aim is a cycle that nudges us in ways big or small towards our goals. I think a lot about feedback loops and how they supercharge deliberate practice. When I am learning or trying anything new, feedback loops are the first thing I think about.
“If I’m going to try this, how will I know if I’m doing it well? How will I know how to get better?”
Getting this right is a force multiplier to everything. Improving a feedback loop has a compound effect on the speed of improving at anything else.
This would be my TedTalk!
Feedback loops trump theory
On this episode of the Tim Ferrriss Show, an episode with more than a few absolute gems, Josh Waitzkin talks about Chess players of the past, and how they would need to study for many years to gauge the best moves for the millions of permutations of a Chess match. Over decades they’d build a wealth of experience that would enable them to select winning moves. Today things are a little different, because we have computers that can give highly accurate assessments of the best move in a given situation in seconds. A practicing chess player today can make a bad move and be instantly notified. Now they can be nudged in the right direction, and internalise the correct strategies, without years of broad theoretical study just to get to the point of identifying the correct move.
For a recent client, I was helping a team to extract some of their legacy system. As we went along there would be many times when the behaviour of the existing system was unclear. There was an engineer in that team who had been there since the inception of the legacy system, and the team would routinely turn to him with questions when we hit these obstacles.
