Tinkering: Experimenting using trial and error, and feedback to create, improve, or fix something

Often, in reality, there exists an asymmetry in the payoff function between success and failure. In these situations, it makes sense to optimize to try as many viable options as possible in the shortest amount of time. Knowledge can help in removing infeasible options but actually trying ideas out is often the only way of finding out what works in reality. Crucially, tinkering allows us the option to throw away the error and continue.

Tinkering involves iteration which involves building something based on information gained through feedback. Some of the advantages of scrum and agile practices lies in how they allow us to tame complexity through aggressive tinkering. Sprints and MVPs can be seen as constructs to aid iteration and gain feedback. Iteration quality can be judged through the number of options being tried simultaneously as well as the cycle time i.e. the time needed to gain feedback as well as the time required to use this feedback to create the next iteration.

One invention that that arose from tinkering that went on to change the world is the steam engine. James Watt, an instrument maker, had a chance to fix a steam engine. He then went on to work on it tirelessly for many years, and managed to achieve a 4X efficiency1.

Natural selection too seems to operate by a process of tinkering. The gene pool of a species constantly iterates generation by generation simultaneously experimenting with mutations. This allows them to keep the upside if it arises while the downsides are cheap and can be discarded.

This is not to say that tinkering is always the optimal strategy nor that theorizing has no place. Often, theoretical methods can be far more efficient. Furthermore, theory can often help establish the validity of options helping us prune the search space for faster tinkering.

In my life, I try to tinker constantly. One way I do this is to keep the cost of mistakes low and reduce the cycle time of new iterations whether it be new foods, people, books, or project ideas. Perhaps I’ll write about these techniques in a future topic.


  1. Faster, Better, Cheaper in the History of Manufacturing (Goodreads link) ↩︎