Move forward using experiments

• Published: 2020-05-21 22:29
• Updated: 2023-03-27 19:07

I like experiments for various things. For testing and verifying a piece of information I dug up somewhere. To transfer a new process into my box of tools. Trying new ideas without the pressure of finishing a project.

To me, an experiment is defined by

• fixed, uninterrupted timebox, so it starts and ends (usually 90 minutes)
• lack of judgement: what happens in an experiment is neither good or bad. it just is.
• specific hypothesis, more on that below
• committment to above

I like timeboxes of 90 minutes, cause it's close to how long our brain can focus on any given task before it starts drifting away. It's also a period of time I can communicate clearly with my peers if necessary. So they know I'm busy. Nothing sucks more than sitting in a room, waiting for somebody to interfere. Yes, more troublesome with kids.

Lack of judgement is also super important to me. In the most scientific sense: an experiment always produces an outcome. I'll have learned something. Maybe that I need to run more experiments. Maybe that whatever thing I tried didn't work for me. In that case I'd like to establish why it didn't work. Or maybe I learned that I need more information. Or change something about my approach. In any case: this is about gathering information - which is always valuable.

Hypotheses come in various forms. And depend on whether I want to test something, transfer something into longterm memory, explore an idea, gather a better understanding of something. To give you an example:

Watched something about a technique. Eg 'how restricting the stereo field for ping pong delays can help them create a more cohesive impression of space'. In order to not dump the information where thoughts go to die, I want to fully understand it. I need to experience it myself. Hypothesis: applying the technique to a project, where the space felt incohesive (or all over the place), helps me mitigating the issue.

So, I want an hypothesis to be as specific as possible. When it comes to transfering something by repetition to longterm memory, the general hypothesis is: 'this works, because of cognitive science'. If it's for example a patch programming idea I want to explore, it could be 'e-Piano sounds should be easy with Wavetable. Let's see how many I can patch up in 90 minutes.'

Last not least: committment. An experiment is a deal I have with myself. To prevent myself from getting stuck judging what I do. Which would diminish both: productivity and effectivity. An experiments always ends with me closing a project or the DAW. That's part of the deal. And meanwhile, I won't mindlessly browse the internet for kitten. It would contradict the deal. Eventually, experiments as a tool in general.

But yes, I also call experiments done earlier than 90 minutes if that's simply the case. Or extend an experiment by another 45 minutes. I'm just doing it in a specific manner.