Prototypes should emphasize speed over polish
It needn't look good, or even have labels. Make just enough to evaluate the idea. Then decide whether to go straight or turn. Identify the intent of the prototype. What hypothesis are you testing?
Transition from exploring to refining near deadline.
Reduce the amount of explorations as deadline approaches and start to focus on what's working.
Delete code as you go. Be ruthless.
Delete branches as you go (Mike uses git branches to manage his work).
Make your process reproducible.
A build system provides machine-readable documentation. Accelerate the reuse of parts from past projects. Here is the blog link he is refering to.
Try bad ideas deliberately.
You can't evaluate a visualization absent the data.Don't get too attached to your current favorite. Don't get stuck at local maximum; go down to go up. Design is chaotic.