I spent about £19 on half a year's worth of 1mg Melatonin pills. I swallowed one last night at about 2300, then went to bed at midnight.

Thoughts:

I didn't notice any extra sleepiness during that hour interval between taking it and going to sleep. This may mean that melatonin as a solution to hyperbolic discounting may not work for me. Alternatively I just went to sleep too early and had I stayed awake, the melatonin would have kicked in and made me want to go to sleep anyway.

I woke up over an hour before my alarm, feeling /much/ more refreshed than usual, and with almost no desire to sleep in. If this was the work of melatonin then it will be very useful to me in stopping my habit of turning off my alarm and then going back to sleep again.;

Obviously, this was only one night, so things like placebo and random variation in effects are there. I'll continue with this experiment and take notes. I can't be bothered to do a blinding experiment like gwern did with his nootropics experiments, but given that the effects of melatonin are well-researched and that most people react the same way to it, I'm fairly confident the effects I'm getting are real. Even if it is a placebo, I'd still be satisfied with it given how much money I paid.