Thursday, April 10, 2014

Trial and Error

This is a great TED talk by Tim Hartford. It has some solid advice, both for how to view the world and how to practically navigate complex problems. I won't explain too much of the talk here (watch it!), but one core take-away I'd like to drill into: some problems are so complex that our reason and logic will fail us. We can't solve the problem in a theoretical space. This is a scary thing to admit. Don't worry! In these cases, we can lean on trial-and-error.

When we say trial-and-error, what we're really advocating is a healthy acceptance of failure. We know we won't get it right the first time. That's okay. As long as we keep course-correcting and trying new things, we're moving toward a solution. As Tim points out in the talk, this is how nature works - it's evolution.

Evolution, however, can take a very long time [citation needed]. To practically apply the idea of trial-and-error to my work, the key question I keep coming back to is, "how can we make iterations cheaper?" This is similar to another question I've heard repeated: "how can we fail faster?" Here are some things that can help:

  • make better tools. As a tech artist, I may be biased, but I think most people under-invest here. Automate what's worth automating, be thoughtful about your process. Eliminate needless steps.
  • understand your MVP (Minimum Viable Product). What's the least amount of work you can do? If you're designing a character, do you care about the materials of the clothing? Probably not before defining the body shape. Focus on the broad problems first. Don't waste time on the details you'll end up throwing away.
  • don't be timid - go too far, then dial back. If something is too small, make it bigger than you think you need it to be. If a color is too cool, make it fire red. It will ensure that you've tried every option and, if it still doesn't work, maybe you didn't understand why the first solution failed. Tweak something else.
TL;DR - Don't be afraid of failure. You won't know all the answers. Try new things, fail faster, innovate more.