Days when I go biking-or-something in the morning tend to be much better. The obvious question is whether there's a causal arrow, or a common cause.

So I'm going to do an experiment: for 30 days starting Nov 1*, I will:

• in the morning, flip a coin:

  • heads: eat breakfast, start work

  • tails: exercise, eat breakfast, start work

• in the evening, rate how energetic and competent I felt, from -25 to 25:

  • -20 is "day is pretty much a total loss"

  • -10 is "definitely somewhat dumber than usual"

  • 0 is "eh, normal, I guess?"

  • 10 is "definitely somewhat cleverer than usual"

  • 20 is "on fire in a good way"

• (I expect the distribution to look something like ~Normal(0, 10).)

At the end of the 30 days, this market resolves PROB(50 + avg rating from days I exercised - avg rating from days I didn't).

I won't bet in this market after the experiment begins.

* (skipping days when there's an obvious reason to, like "I'm sick" or "I have company"; this extends the end date of the experiment)