polyphasic sleep works, but only with about 70% confidence
“Works” is a very underspecified statement. In particular, I’m curious whether you mean “works for most people”, “works for some people”, or “works for a few people”.
I think it’s pretty evident that it does not work for everyone, but people for whom it works exist. So the question is really “for which part of which population does it work?”
I think it’s pretty evident that it does not work for everyone, but people for whom it works exist.
I don’t think that it’s evident that there are people who can sleep 2 hours per day with Uberman without taking a hit in some important cognitive functions.
There are many more reports on the internet of people who claim to eat nothing and live on light then there are people who claim to be successful on Uberman for longer periods of time.
I think we were arguing not about the applicability of p/s, but about the theory. So I meant “p/s works” = “splitting sleep into multiple phases in certain ways does increase efficiency and makes you require less sleep”
If that is not the case, I wouldn’t say p/s “worked” for me. Trying it would still have been one of the best decision I’ve made in years, but it would only be so because i also fixed real problems in the process. In that world, I would have achieved the same by improving hygiene and nutrition and fixing my schedule to regular, monophasic sleep, and arguably with less effort (although I actually kind of like the nightly hours, I might keep this even if I come to believe that it doesn’t improve efficiency.).
People are different. Few interventions (including chemical) work for literally everyone—e.g. some people just don’t react to common drugs or react in a way which the doctors politely call “paradoxical”. So even in theory p/s might work for some but not all people.
It seems to me that the closest analogy is diet. Take low-carb—does it “work”? The answer is: it depends. It does wonders for some people, does nothing for others, and screws up the third bunch. Unless you understand the mechanism by which it either works or doesn’t, all you can do is provide priors (e.g. works for 30%, does nothing for 60%, screws up 10% [fake numbers]) and say “try it and see if it works for you”.
I would want the same type of answer for the polyphasic sleep. What do you think are the baseline probabilities for, say, four potential outcomes?
OK, so basically you think that one in ten will benefit, nine in ten will suffer (and, presumably, revert back to “normal” sleep), and some small number will think it’s all the same. Hmm…
So I meant “p/s works” = “splitting sleep into multiple phases in certain ways does increase efficiency and makes you require less sleep”
This appears to be true if you must be sleep deprived. That is, if you want to operate at X% function, where X is less than 100, likely less than 70 or so, you would need to sleep a shorter duration on a polyphasic schedule than you would on a monophasic schedule. (“X% function” is somewhat vague, but I trust you understand what I mean.)
However, if you want high X% function (say, higher than 90%) then the required sleep durations appear to be the same in either case. This could easily make polyphasic sleep likely less efficient considering logistics (time spent getting into bed, etc.) and time to fall asleep.
I’d recommend taking a look at Stampi’s book for more information on polyphasic sleep being efficient for sleep deprivation but not so for normal levels of sleep. I also want to note that the estimated percentages I gave above are for illustration only. Look at Stampi’s book for more accurate information. I do not have a copy any longer.
“Works” is a very underspecified statement. In particular, I’m curious whether you mean “works for most people”, “works for some people”, or “works for a few people”.
I think it’s pretty evident that it does not work for everyone, but people for whom it works exist. So the question is really “for which part of which population does it work?”
I don’t think that it’s evident that there are people who can sleep 2 hours per day with Uberman without taking a hit in some important cognitive functions.
There are many more reports on the internet of people who claim to eat nothing and live on light then there are people who claim to be successful on Uberman for longer periods of time.
I think we were arguing not about the applicability of p/s, but about the theory. So I meant “p/s works” = “splitting sleep into multiple phases in certain ways does increase efficiency and makes you require less sleep”
If that is not the case, I wouldn’t say p/s “worked” for me. Trying it would still have been one of the best decision I’ve made in years, but it would only be so because i also fixed real problems in the process. In that world, I would have achieved the same by improving hygiene and nutrition and fixing my schedule to regular, monophasic sleep, and arguably with less effort (although I actually kind of like the nightly hours, I might keep this even if I come to believe that it doesn’t improve efficiency.).
People are different. Few interventions (including chemical) work for literally everyone—e.g. some people just don’t react to common drugs or react in a way which the doctors politely call “paradoxical”. So even in theory p/s might work for some but not all people.
It seems to me that the closest analogy is diet. Take low-carb—does it “work”? The answer is: it depends. It does wonders for some people, does nothing for others, and screws up the third bunch. Unless you understand the mechanism by which it either works or doesn’t, all you can do is provide priors (e.g. works for 30%, does nothing for 60%, screws up 10% [fake numbers]) and say “try it and see if it works for you”.
I would want the same type of answer for the polyphasic sleep. What do you think are the baseline probabilities for, say, four potential outcomes?
Makes things noticeably better
Makes things a bit better
Makes no difference
Makes things worse
You’d have to be more specific. Probabilities if what kind of person attempts or succeeds to adapt which schedule?
No specifics. General population (or if you want, take the half above the median), and any polyphasic schedule.
8.5% – 1% – 0.5% – 90%
But those feel frustratingly uninformed
OK, so basically you think that one in ten will benefit, nine in ten will suffer (and, presumably, revert back to “normal” sleep), and some small number will think it’s all the same. Hmm…
This appears to be true if you must be sleep deprived. That is, if you want to operate at X% function, where X is less than 100, likely less than 70 or so, you would need to sleep a shorter duration on a polyphasic schedule than you would on a monophasic schedule. (“X% function” is somewhat vague, but I trust you understand what I mean.)
However, if you want high X% function (say, higher than 90%) then the required sleep durations appear to be the same in either case. This could easily make polyphasic sleep likely less efficient considering logistics (time spent getting into bed, etc.) and time to fall asleep.
I’d recommend taking a look at Stampi’s book for more information on polyphasic sleep being efficient for sleep deprivation but not so for normal levels of sleep. I also want to note that the estimated percentages I gave above are for illustration only. Look at Stampi’s book for more accurate information. I do not have a copy any longer.