Icosian Reflections

…a tendency to systematize and a keen sense

that we live in a broken world.

[Polyphasic Sleep] Greetings, Polyphasic Sleep, Chives


I'm blogging now? Apparently. Expect a hodgepodge mix of assorted things, roughly corresponding to the set of things I'm interested in, in no discernible order. Maybe at some point, I'll get my act together and start composing sequences of posts, but for the initial brain-dump phase, expect no more than, well, a brain dump.


Today, I begin transitioning to a polyphasic sleep schedule. If it works, I'll develop the ability to REM sleep in naps of twenty minutes, and survive on fewer hours of nighttime sleep by napping during the day. (The particular cycle I'm aiming for is "Everyman-3", three hours of night-sleep and three twenty-minute day-naps.)

Well, perhaps the word 'survive' is misleading. I'd like to do a lot better than 'survive'; I'd like to continue functioning at my present mental capacity, for extra conscious hours. While efficiency-times-duration calculations indicate that I should be willing to accept up to a 20% drop in efficiency along with the 25% increase in waking-hours, I suspect that the nighttime hours that I'll be awake will be less useful than the daytime hours I'd be taking a productivity-hit in, which means that an 20% hit is a loss. I'd like to see no more than a 5% dip in mental-function test scores, and if I see more than a 10% drop, I'm going to need to seriously evaluate the utility of being awake when everyone else is asleep.

My outcome expectations are as follows:

  • 50% I give up before completing the transition
  • 20% I observe significant mental-function decline post-transition
  • 30% my test results exhibit no significant decline

Within the 30% chance of success, though, I expect a 1/3 chance that the mental handicap simply wasn't caught by the tests, so I'll expect, to first order, about a 20% probability of successful, symptom-free transition, as well as a 30%-ish chance that I'll quit, but retain the ability to REM-nap, which may prove useful in the future. (A gentler 5:00/0:20/0:20/0:20 cycle that totalled six hours, for example, might prove more useful, and just as restful as a monophasic 8-hour.)

Why? Because 20% odds of success * 4 hours/night * 365 nights/year * 3 years = 876 hours. Assuming that I spend the next two weeks out-of-commission on account of the toll of the transition schedule, I lose 16 hours/day * 14 days = 224 hours. To first order, running the experiment is beneficial on average to the tune of a several hundred hours. Or several thousand dollars (assuming my prospective pay-grade at my previous employer), or approximately five lives saved (by donating to the AMF). Which is definitely overgenerous, as it assumes that every hour gained is optimally-productive. But an estimate in the range of 2-4 lives is probably the right order of magnitude.

That is, assuming that the best thing I can do with my time is do industry work for money. Considering the unique (and time-sensitive!) opportunities available to a Harvard undergrad, extra hours in these years are almost certainly worth more than $21/hr. But that's a topic for another day...


My chive plant is dead. Either I suck at taking care of living things, or it never was meant to live in partial sun. Will we ever know?