my polyphasic sleep experiment

April 22, 20203 min read

I’ll write about this more in-depth soon, but wanted to get a quick post out while my polyphasic sleep experiment is still top of mind.

I switched to a polyphasic sleep schedule (specifically the Everyman-3) for 11 days. Last night, I returned to a monophasic schedule.

At the start of the experiment, I was extremely exhausted. I would fall asleep in the middle of reading, playing video games, or even standing up.

After a few days, the exhaustion faded.

I could wake up at 2:30 am every day, and stay awake throughout the day aside from my three 20-minute naps. In fact, my body got so used to being awake all day that occasionally I would be unable to sleep during nap-time.

Occasionally I got hit by a wave of tiredness, but I’d combat it with a physical activity, like juggling or push-ups.

I started to feel that maybe I *could* have it all: I could sleep for 4.5 hours a day and suffer minimal consequences.

I was using an internet version of the Psychomotor Vigilance Test (PVT) to test how “sleep deprived” my body was. The real PVT is used by astronauts on the International Space Station. I thought a modified PVT would be a good proxy for My Brain on Sleep Deprivation.

Throughout the experiment, I scored ‘normal’ on the test.

But since ending the experiment yesterday, I’ve realized that I was suffering significant effects from sleep deprivation.

Thinking felt like swimming through a thick liquid.

I would forget what I was doing, in the middle of doing it. This could happen multiple times if I was doing a long or complex task.

My abstract thinking capability plummeted. I’m a programmer, and bugs that should have taken me an hour took me days. Scott had negative feedback on my PRs, which is unusual. (When I told Scott that I had ended my sleep experiment he replied: “Thank God. I thought I was going to have to roast you at our one-on-one because of your PRs”).

I’ve only had one night of normal sleep since ending the experiment, but my brainpower was 10x today. For the last week, the little work I’ve done has been based on pre-existing knowledge, and even then it was a slog. Today, I learned something new (re: Django filters) and applied the new thing. I don’t think I could have done that when sleep deprived.

I’m not at full brainpower just yet. That’ll take another day or two, I think.

One of my lessons from this experiment is that I need to choose better measurements to monitor my state. During the experiment I thought I was doing “okay,” but my qualitative assessment was hampered by my compromised ability to think.

And my quantitative assessment that was meant to provide some objective feedback on my state failed too. I’ve since found this study about PVT effectiveness. There is little standardization among PVTs, but the PVT I used was inferior in a few ways: the main measure of performance was average response time, and the test lasted 2 minutes when longer tests might give more accurate results.

I will miss this experiment a little, because having infinite time is awesome. Especially on the weekends when I didn’t have to apply myself.

I spent my extra time on my quarantine hobbies (learning juggling and handstands), watching TV (Tiger King, Unorthodox, Run This City), hanging out with Andrew and my housemates, and playing video games (Dark Souls). I also wrote a lot.

But overall, I’m happy to have my brain back.