How long should a Polyphasic nap be?If you are looking to use a "standard" polyphasic schedule, such as Uberman or Dymaxion, then the obvious answer is: however long the schedule calls for, 20 and 30 minutes respectively. But, perhaps there is more to the lengths of these naps then first meets the eye, or mind.
So, what happens during sleep when one adapts to polyphasic sleep? The first thing that is generally accepted as happening is that people enter REM sleep more quickly. This should be obvious. A normal sleep cycle for a monophasic sleeper is around 90 minutes and REM occurs at the end of that 90 minute period. See our article on how sleep professionals usually respond to the idea of polyphasic sleep or even better our article on sleep phases. So, since REM sleep is very important, our human bodies adjust to the need to get it by entering into REM sleep more quickly. The most data collected on polyphasic sleep was done by Dr. Claudio Stampi. There is a youtube video about his experiment with one voulunteer below. According to the video, his subject went into REM sleep almost immediately after going to sleep. But, he indicated that the subject would then go into deep sleep. I should note that this was for 30 minute naps.
The next best documented sleeping was by Steve Pavlina with 20 minute naps. The indication by Steve (and other bloggers on Uberman) is that they seem to be waking up from REM sleep. Obviously, if they are entering REM quickly and are in REM when they wake up 20 minutes later, there is only 2 options: 1 REM sleep phase or a REM sleep phase followed by deep sleep followed by another REM cycle. The likelihood of the second option is low, so the most likely thing is that there is only one cycle of REM. Of course, what this means is that the person on uberman is actually NOT getting any deep sleep.
If this is the case, then this would fit with the generally agreed upon idea on-line that Dymaxion sleep is harder to adapt to than Uberman Sleep. Awaking from either phase 1, phase 2, or REM sleep allows the person waking to feel refreshed. Waking from deep sleep (phase 3 or phase 4) will not provide this. Rather, the person waking will feel groggy and will desire to go back to sleep. People, such as Steve Pavlina, have indicated that once adapted to Uberman, they do wake feeling refreshed. This is contrasted to the subject in Dr. Stampi's experiment who, even after adjusting, would still have difficulty waking up when he was supposed to.
This seems to make the most sense. There is some research on sleep deprivation which indicates that the body will quickly enter REM sleep when deprived. So, what would seem to make sense is that on an Uberman cycle, one "simply" is training their body to always get REM right away when they go to sleep, similar to how the body natually responds to sleep deprivation. Then, train themselves to wakeup when exiting REM. This sounds complicated. It is, which is why it takes time and effort to adjust. But, this is easier than the adaptation to Dymaxion sleep which means either includes a stretching or compressing of sleep cycles. If the same adaptation occurs as in Uberman, then an additional stretching of the REM sleep time period needs to occur so that the REM cylce doesn't end and cause waking until the 30 minutes is up. There is another theory that the normal cycles are actually being compressed to 30 minutes rather than adding a REM cycle. But, Dr. Stampi's work does not seem to support this. Additional sleep studies including EEGs of both Dymaxion and Uberman sleepers is needed to confirm these conclusions.
Another thing to take into account is the amount of time it takes to fall asleep in the first place. For most people on nap-only schedules, such as Uberman or Dymaxion, the falling asleep is often not a problem. Most people report being able to fall asleep within a minute or two. But, for Everyman, it is much harder to fall asleep. Either way, each individual is different and will have different difficulty falling asleep for his naps. This time to fall asleep needs to be considered when discussing nap lengths. Obviously, we set alarms and then Lie Down. We do not set them right before we go to sleep. So, if one wants a 20 minute nap but it normally takes roughly 8 minutes to get to sleep, then setting the alarm for 28 minutes might make sense. Or, setting it for 30 might even be better in case it takes a bit longer. The problem with this is that one runs the risk of sleeping past the desired REM cycle and into a deep sleep cycle. This is undesirable.
So, to make a long story short. There is limited evidence, and some logic, that would point toward 20 minute sleep cycles being more natural and more easily adapted to. Of course, something in between, such as 22 or 25 minutes might also work and actually be optimal. We recommend that you play around with the time and find what works for you, most especially depending on how long it takes for you to fall asleep, even after adapted. Unles ou fall asleep immediately every time, you will have to make a trade-off between getting as much sleep as possible and not falling into deep sleep which will be harder to wake from.