An Experiment: Polyphasic Sleeping

I attempted the Uberman sleep cycle about a year ago. I managed to get by with 20 minutes sleep every 4 hours. I lasted 2 weeks before bad things started happening (hallucinating, extreme weakness, and major cognitive deterioration). It was a fun experience but something I am convinced is fabricated and doesn't work despite some peoples anecdotal evidence.

Eventually I ran into this blog and when I read it I immediately quit the cycle http://www.supermemo.com/articles/polyphasic.htm

If anyone has any questions about it/are thinking of trying a sleep cycle I would be glad to answer!

It's hard for me to say anything on this topic without resorting to hyperbole. I have a serious sleep disorder and in desperation, I tried a polyphasic experiment about 7 years ago.

I was determined to see it through no matter what the cost. The conclusion I came to was that it is a complete scam, perpetrated on life-hackers who are often, productivity hounds. Polyphasic sleep does not work; it does not improve productivity; it is not healthy. The only people who have slept like this on a consistent basis are soldiers or other people in extreme situations, where wakefulness is equivalent to survival and so the damage to your mind is a trade-off. If you are not fighting to stay alive, then you gain nothing by not following your body's natural rhythms, and you can seriously mess yourself up by experimenting with alternatives.

Then again, you're young so you'll probably recover. Probably.

I had a biphasic schedule about 8 years ago for 9 months. I slept at night from 1:30am to 6:00am and in the afternoon from 3:00pm to 4:30pm. I was never tired. I felt like I had two days in one. It was great. I worked from 7am - 2pm, ate lunch, and went back to sleep. When I woke up at 4:30pm, I had 9 hours of free time until I would sleep next.

I don't have a flexible work schedule like that anymore, and now that I live in San Francisco, the environment is not exactly quiet enough at 3pm for quality sleep.

I recommend trying it out if you have the flexibility. Avoid caffeine. Treat the afternoon sleep time, the same as you would the night time. Block out noise and light. Be consistent. Having two days in one is pretty awesome.

I experimented with this though college. I found the idea that you could have extra months in your year amazing. But I could never stick with it for more than a couple days at a time. It was primarily this that kept kicking me off:

"I also have a theory about just being conscience that long, you start to lose a sense of time there is no finality just this never ending loop of waking up and sleeping"

I really decided I needed significant time sleeping to reset my brain. Otherwise it did just feel like a single extended consciousness.

And if you believe John Medina in Brain Rules (http://www.brainrules.net/) sleep is what solidifies memory. Which missing that isn't so helpful to a college student.

I'm the dad of a five-week old and, give or take a little, this seems pretty much the pattern I had for weeks 0-3. The thing I found most surprising was how I did manage to function after waking, and in a relatively short space of time after waking. The second most surprising thing was how hellish, for me, it felt to be waking after the naps.

Edit: Since about week 4, I've tried to pack more 'regular' sleep in.

I saw an article that debunks a lot written by a doctor, that might be biased but it's still worth a read: http://www.supermemo.com/articles/polyphasic2010.htm

Don't get me wrong, I've tried it, and failed like many others before me. I'd tried the uberman method for about 2 weeks before falling, and I didn't feel very productive during that time at all (although there were times that I felt more awake than I ever have done, and time seemed to merge, days didn't exist... it's a weird feeling)

I've now adapted the old version of polyphasic sleep, take a nap in the afternoon and sleep for about 6 hours on a night and it seems to be doing me well :)

But! I am tempted to give the uberman method another try, and I will write about it if anyone would be interested?

This comes up on HN about once a year. Here's my blog where I wrote about trying it: http://www.pchristensen.com/blog/category/polyphasic/

tl;dr version: - the first 5-7 days are just awful but after that you're not tired - deviating from the schedule (say, staying awake 7 hours one time instead of 4) is devastating - you do feel different and perceive time differently. Kind of like San Diego weather instead of Boston.

Interesting experience, worth a try, I'd do it again in a heartbeat if my home and work schedule accommodated it.

I wouldn't even try something like this, so good on you for giving it a shot and writing about it.

However, having researched polyphasic sleep before, my reaction at coming to the end of your essay was surprise that you'd ended the experiment prematurely. It seemed like you hadn't given yourself time to fully adjust. I expected a Week 3 during which you were able to drop immediately into REM and wake up refreshed.

I assume you just ran out of vacation, which is understandable. But it seems unfair to judge polyphasic sleep based on an incomplete run.

If you are interested in polyphasic sleep, the best long run experiment I know is from Steve Pavlina. 90 days of polyphasic sleep:

http://www.stevepavlina.com/blog/2005/10/polyphasic-sleep/

Interesting experiment, thanks for sharing.

Original polyphasic sleep was divided into two periods of night sleep. The historical evidence shows this was considered normal in most pre-industrial societies. http://www.historycooperative.org/journals/ahr/106.2/ah00034...

Is there any actual evidence that anyone has ever been able to stick to one of these polyphasic sleeping schedules for a long period of time? All I ever hear about it is vague unsubstantiated claims about how a bunch of really smart people did it their entire lives and stories about how people tried it for a week or two and then stopped.

I have a sort of strange sleep schedule.

I work for international teams with no one I work with being on my timezone, I work from home.

So every day my sleep schedule is different. My wake schedule is different. I should record my schedule since it's difficult to remember since it changes every day.

To give some impression, yesterday I had meetings at 2am, 6am, 6pm, and 10pm. Through that I slept, and had social time. I ate and had normal experiences.

They just weren't on any sort of real schedule. Tired? I sleep. Awake? I go do things or work.

Anyone else do this sort thing? So far I don't ever feel "wrong" about the schedule physically, because it's always changing. Now constant change feels right.

Sleep all night tonight? Sure. Wake up at 2am and work till 10 am this night? Sure. Have a meeting a 3am, then sleep till 6am, then work some more? Sure.

I just follow whatever seems like the most reasonably course of action. Including work 10 hours today, and 6 hours the next. Whatever makes sense.

International-Logic-Phasic Sleep. Heh.

I've gotten locked into strange sleep cycles simply out of necessity (I am an undergraduate student at Carnegie Mellon). I've found that, while my body will accept a psuedo-equilibrium of somehow getting ~ 20 hours of sleep a week, it will almost always become unsustainable after 2 weeks... in which I'll need another day or two to recover. But after that (usually I take an entire week off, where I get at least 30 hours of sleep) I go back into my regular routine (note that I have a 105 minute sleep cycle, which is a little longer than most people's:

M3:30 T0:20 W3:30 R0:20 F3:30 S7:00 U3:30

This is a perennial topic on HN. Prior discussion highlights: http://news.ycombinator.com/item?id=1464142 and http://news.ycombinator.com/item?id=652747

Just some notes; on what I thought about polyphasic sleeping; I will answer any questions people may have.

I tried what might be called polyphasic sleeping twice. When both of my kids were born. It sucked. Babies do it, and their parents are forced to join them for the first few months of life. It's a damned good thing we teach kids the normal way of sleeping, when it's night.

The BBC published an interesting article recently about how it was possibly customary until fairly recently to sleep in two sessions at night, with an hour or two of wakefulness in between them. People would get up, read, eat, talk with their partners etc. in the wakeful hours.

TBH, I'm not sure the evidence is convincing, but it'd an interesting concept.

http://www.bbc.co.uk/news/magazine-16964783

[edit: earlier HN discussion of the BBC article: http://news.ycombinator.com/item?id=3620742]

Well, I'm not sure that's a very good idea. Apart from the fact that sleep helps solidify facts learned during the day (not sure whether this is really true or not but from my experience I'd say there is something to it) I remember reading a paper claiming sleep could have a crucial function in cell refreshing and rest even on a molecular basis. In fact, the former physiological and psychiological effect could be very well implicated by the latter one actually.

"I found myself constantly switching between reading, coding or the next shiny thing finding it hard to focus on any given one."

This is a symptom of ADD, however it is caused by lack of sleep. I experience this, so I'm very familiar with it. I've tried CPAP, oral sleep appliances, and everything else. Nothing has worked perfectly.

Well what about the siesta? Or, before lighting, people more commonly had segmented sleep. http://en.wikipedia.org/wiki/Segmented_sleep

This seems to be a modern, social issue, not so much a physical one.

The sailors that cross oceans single-handed use some variant of this. The reports I've read is that the first two days are quite hard to handle but once in the quite issolated routine it works for certain folks.

Hm, no concrete data on mental performance on benchmarks, or recordings of sleep data (eg. via a Zeo).

Oh well. Someday someone will do the experiment better.

Maybe it's just my lousy Dell laptop... but the difference between those two colors on the graph is nonexistent.