Switch to mobile version

Experiment No. 22 – Studying my own sleep

The purpose of this experiment is to track the quality and length of my sleep under different conditions. I’m using an app to track bodily movements (and therefore deep sleep phases), as well as total sleep time. Over the course of five weeks I’ll be testing out different sleeping arrangements. I’m going to try earplugs, total darkness, maybe some earlier and later bedtimes, before-bed meditation, and caffeine abstinence.

This experiment has finished. The final report is the last entry on this page.

Here’s the rough schedule:

Week 1: (Feb 29 to March 6)

  • A few control days, where I only track and don’t change anything.
  • A few days where I make my bedroom as dark as possible.

Week 2:

  • A few days with earplugs
  • A few days with earplugs and maximum darkness

Week 3:

  • A few days with a short meditation session right before bed
  • A few days with no electronic screens at least an hour before sleep

Week 4:

  • A few days with no caffeine after 2pm
  • A few days with no caffeine at all

Week 5:

  • A few days with a longer overall sleeping time (8 hours instead of 7)
  • A few days with some yet-to-be-determined difference

These are subject to change, but it definitely begins February 29 and ends five weeks later.

I’ll be adding diary entries below, and posting data occasionally. At the end I’ll post a full report.

Experiment Log

You can view my data spreadsheet here. It should update automatically, although I might not update it every single day.

After Night 1: 2/30/16

Had my first “control sleep”. It wasn’t a bad sleep although I went to bed a bit later than expected, and woke up later than expected because I had set the alarm incorrectly.

I’m giving each sleep a subjective rating out of five, according to how good it seems to have been when I woke up. I have no way of knowing which sleeps are actually the most restful or beneficial, so this rating is really just how close it is to my ideal sleep.

Here’s the sleep graph for the night.


Basically all it measures are bodily movements during sleep, and during deep sleep your muscles are deactivated. I won’t post the graph every day, but I wanted to post it today so you know what kind of data I’m looking at. There is more data in the spreadsheet linked above, including caffeine consumption during the day and a few other things.


After Night 2: 3/2/16

Today is technically a control day, but it also happened to be a night I had concert tickets for. I went out with friends, and alcohol became a factor in my sleep. I ended up going to bed quite late, and sleeping in. My sleep was very fitful, and once I woke up I couldn’t sleep any more, which has always happened to me when I drink alcohol.

Tomorrow is the last control day, and then I’m going for as close to total darkness in my bedroom as I can get.


After Night 4: 3/4/16

So I had my last control sleep last night. I am feeling kind of weak/tired these days, and I know part of it is because I’m not getting enough sleep. Another factor might be Seasonal Affective Disorder, which my winter-addled city feels particularly hard this time of year. And a third factor might be that I’m eating a caloric deficit because I’m cutting weight. Neither of the latter two factors should mess up the experiment too much, because they will be affecting me throughout the five-week course.

Mostly though I’m still recovering from the alcohol-perturbed bad sleep on Tuesday night. Usually I need one good sleep-in to recover from a bad sleep, and I haven’t done that for myself yet.

So far there isn’t much to report, but tonight I will introduce my first variable: complete dark. On my bedroom window I’ve got vertical blinds plus heavy drapes, and I’ll make sure no light is coming in the door one way or another. I’d rather not use a sleep mask because it might annoy me enough that I don’t sleep as well as I would in pure darkness.

After four sleeps, my spreadsheet is showing a few early patterns. Three of the four sleeps have nearly identical stats, and the post-concert one is the outlier.


After Night 7: 3/7/2016

I finally got to sleep at a reasonable hour last night, shortly after 11:00. My bedtime had been creeping later for quite a while now. But I’m waking up quite a lot, and I feel kind of restless when I do. I don’t have trouble getting back to sleep but it feels like something is preventing me from sleeping as deeply as my body wants to. I think it’s the caffeine.

I can’t quite get my room to total darkness. The double-layered blinds still let a bit of light in. Not nearly enough to see anything, but it isn’t that perfect depthless blackenss that you would get in say, an isolation tank. I considered trying a sleep mask but I don’t think I would sleep well with something stuck to my fact. In any case I think I’m getting almost all the benefits of darkness but it’s not adding up to much.

Tomorrow night I will try earplugs.


After Night 9: 3/9/2016

Last night was the first night I used earplugs. It’s too early to say whether they had an effect — I’ll be using them most of this week. I did take them out at some point during the night, although I don’t remember doing it. I know they didn’t fall out because they were standing neatly on my nightstand when I woke up this morning.

The quality of my sleep is slowly improving, as I start going to be earlier. Last night was an exception, because I had a friend over a little later than expected. (We were playing The Witness together and wanted to finish it — highly recommended.) I’m not seeing many correlations in my data yet except the one between going-to-bed time and subjective sleep quality. I have simply been going to bed too late on weeknights and that’s becoming clear.

The elephant in the room is still caffeine. My cutoff time has crept up from 3pm to 6pm over the last few months, and I didn’t think it was a problem because I don’t drink that much overall and I seldom have trouble getting to sleep. But my sleep is clearly not as restful as it once was. I wake up in annoyance several times most nights and I have more than a hunch that’s what it is.

Tonight, earplugs again.


After night 11: 3/11/2016

I had a great sleep! For the first time in forever, I only woke up once. I woke up feeling very well rested. It was pretty close to an idea sleep.

I suspect this happened because I did something I hadn’t done in a long time: I spent my evening reading a book. The evening was very quiet and didn’t involve many screens. By bedtime I felt calm and easy. Recently I’ve been watching a lot of videos and shows and reading very little printed material. This is consistent with my emerging hypothesis, which is that a winding-down period before bed has a defining effect on quality of sleep.

So I am unable to keep earplugs in my ears all night. I take them out without noticing. Earplugs have been very useful for getting to sleep while traveling, when you end up sleeping in a room above a bar, or in a hostel dorm room. But at home they don’t seem to do much because getting to sleep in the first place hasn’t been a problem.


After night 13: 3/13/2016

Two crappy sleeps, and both times I had unusually late doses of caffeine. It’s like clockwork.

Tonight is supposed to be my last ear plugs night, but I think I won’t bother. I always remove them. But they are useful for getting to sleep when there’s unusual amounts of noise.

Tomorrow will be my first of four nights with a short meditation session before bed.


After night 17: 3/17/2016

Another complicating factor emerged in two of the last four nights. Occasionally, if it’s been too long since I’ve eaten, or the last thing I ate was low-fat, low-fibre, I will have low blood sugar at the time I’m trying to get to sleep. My mind gets really active and I feel a bit ill. I start seeing images swirling when my eyes are closed. If I fall asleep in that state, my dreams are very vivid and even kind of violent, or at least jarring. Usually I’ll wake up shortly after that and have to eat something (a piece of toasted fibre bread usually does it) or that will keep happening. This happened on night fifteen and sixteen, which delayed my bedtime and made my sleep kind of crappy.

Daylight Savings Time has also messed me up a bit I think. I’m maintaining the same “clock” bedtime, but that time is actually an hour earlier than it was last week. So my mind is more active come bedtime and I’m still adjusting. Last night I finally had a good sleep again, even though I forgot to meditate before bed, which is the variable I’m experimenting with now.

There are variables all over the place so it’s hard to know exactly what is doing what, but some patterns are emerging in the graphs and data. My sleeps are more pleasant and restful when I a) spend more time in deep sleep and b) wake up less often. This sounds obvious but it’s nice to see the data supporting it, because it gives me a measurable goal. I am gradually getting a firmer picture of what is needed to have sleeps like that. So far it seems that some kind of winding-down at the end of the day is the most helpful actor. Reading, stretching, slowing down physically, limiting artificial light.

I will do a few more days of before-bed meditation and then move on to no screens an hour before bed.


After night 21: 3/21/2016

I have a few bad–or short, rather–sleeps at the end of last week. I had another night with that weird blood sugar issue, and then another few nights where I stayed up too late. I almost never sleep in to compensate for that, because I’m afraid the next day will disappear on me if I begin it too late. So I’m coming back from a bit of a sleep deficit.

Last night I was extremely tired, and went to bed “early” although it was still after 11pm by the time I finally turned on the sleep tracker and closed my eyes. And my sleep was great. It seems that the earlier I get to sleep the better I sleep.

Bedtime meditation seemed to be healthy for me, although I kept my sessions very short (5 minutes) so that I would actually do them. The data doesn’t show any effect on my sleep numbers, and subjectively I don’t think I noticed much of a difference. I think a full 20-minute session before bed would be extremely beneficial on all sorts of levels, but I’m not ready to commit to that.

Anyway, it’s time to finally test the think that is the most-recommended suggestions for improving sleep: no screens an hour before bed. I have a good book going now so I’m looking forward to that. I almost always have an electronic screen in front of me until shortly before bed, so this could make a big difference. I’m going to do it for five days in total.


After night 29: 3/29/2016

Just wanted to make a quick update because I haven’t in a while. Preparing for Camp Calm season 2 has kept me busy. For the last week I’ve been testing two related things: no screens an hour before bed, and screens with the blue light electronically removed (apparently it’s the blue light that convinces our bodies it’s daytime).

But as I continue to track my sleep I’m really not seeing any definitive patterns. Some days I have a great sleep, some days I don’t. Of all the factors I’ve tracked (and observed informally) the ones that seem to affect sleep quality in a predictable way are: how recently I had caffeine, the total length of my sleep, and the temperature in the room (cooler is better).

I wish I had tracked my sleep for longer, because I am generally more restless with my sleep than I was a year ago. The only major change I made that I can think of was I started lifting weights regularly six months ago. Exercise is supposed to help sleep, but maybe it is upsetting my rhythm in some way. Or maybe my diet changed to high-protein when I started lifting, and that’s it. It’s hard to know. Part of me feels like I’m learning a lot in this sleep experiment, and part of me feels like I’m learning nothing.

I have made an important decision though: I am getting a new mattress. This one is eight years old this April.


After night 36: 4/5/2016 [FINAL REPORT]

So it’s officially over, and I suppose I’ve learned some things. I didn’t notice a big change between the different variables I tried, however. Sometimes I slept well with no screens, with earplugs, or in total darkness, and sometimes I didn’t. I don’t think I nailed down the necessary factors for a good sleep.

I think I did nail down some surefire ways to get a bad sleep though. Late caffeine (after 6pm) being the most obvious one. Temperature was also a fairly consistent determinant (cooler is better for me). Also, a “winding down” period was helpful, though it doesn’t seem to be necessary for a good sleep, nor does it guarantee one.

Another thing that seemed to matter was the total length of sleep. Seven and a half hours seemed significantly better than seven, judging by my subjective opinion of the sleep at the moment I wake up. My best sleeps were a bit longer than usual.

I don’t think I will do a full blog post about my findings, because they’re really not conclusive in any way that might be useful to others. I had planned to use the sleep tracking data to graph correlations, but there really aren’t any consistent ones. I don’t think I will load the rest of the sleep data into the spreadsheet either. At first I thought there was a clear relationship between sleep quality and percentage of time in deep sleep, but there were too many days that contradicted this.

But I am very glad I did this experiment. I’m still very interested in sleep quality and plan to keep trying things, including (maybe) bi-phasic sleep.

Perhaps most importantly, this experiment has led me to the decision to get a new mattress. It will definitely make a major difference. I get slightly congested whenever I get into bed and I wonder if it’s the decade-old civilization of dust mites that must have established itself in there somewhere.

And I’m learning a lot about the racket that is the mattress industry. If you’re looking yourself, I highly recommend going to oldbedguy.com and reading everything he has to say (begin with the Mattresses 103 article). He is an octogenarian with 50 years in the mattress industry, and he’s seen all its tricks and lies. He has taken it upon himself to give the most informed and impartial personalized mattress recommendations to confused mattress customers. He does not accept money for his services (and instead asks people donate to a local no-kill animal shelter). What a godsend.


A Raptitude Community

Finally! Raptitude is now on Patreon. It's an easy way to help keep Raptitude ad-free. In exchange you get access to extra posts and other goodies. Join a growing community of patrons. [See what it's all about]


Flor March 29, 2016 at 2:06 pm

Hello David,
I’ve been trying to carry out this experiment on my own as well (with some variations) but I have the impression that the factors don’t influence the quality of my sleep itself, they only make falling asleep easier.
I am surprised at your results – especially with sugar-; they offer a new and unknown perspective and also suggest that not everyone’s bodies react in the same way.
Speaking of new perspectives, I recommend that, if you find some time when you have finished your experiment, you read “Night School” by Richard Wiseman, if you haven’t already. It is a very instructive book full of social experiments which try to examine myths about sleep. It would be interesting to compare your results with what you read.

{ Reply }

Bernadette April 4, 2016 at 7:45 pm

Hi David, I sent you a private email a few weeks ago about a new supplement I’m taking (and which I’m a distributor for). The stand-out common feedback is the greatly improved quality of sleep people experience. Personally I am having the best sleeps ever, consistently! I fall asleep easily, sleep soundly, and get back to sleep easily after a trip to the loo. It’s formulated to reset gene expression that has become less-than-optimal over the years.

{ Reply }

Ville Sorsa April 15, 2016 at 10:00 pm

Hello David!

If you want to get really scientific on this thing, I suggest reading, or at least skimming this article:

It’s written by a polish researcher of memory and learning.
His strategy for optimal sleep is so called free running sleep:
Go to sleep if and only if you feel sleepy enough, and sleep as long as you need to, without using an alarm clock.

{ Reply }

Andy May 17, 2016 at 9:32 am

Hey David,

Not sure if you made that new mattress purchase yet or not. In case you didn’t, I thought I’d mention that I bought a Casper mattress about six months ago and have just completely fallen in love with the thing. I’m not affiliated with Casper in any way/shape/form, but I feel inclined to recommend them whenever I can. It’s like a Tempur Pedic, but for 1/4 the price.

{ Reply }

Abhijeet Kumar July 20, 2017 at 7:18 am

Hi David,

Sleep has been a personal favorite subject of mine over the last few months, actually the last year. I haven’t tracked or run experiments on my sleep with the various factors involved, or wait I have, but it was less an experiment and more a part of me desperate to make sleep sound and deep and long enough everyday, that I get lost like you mentioned. This has been far from achieved.

Before I go further, let me clarify I do not drink caffeine at all, unless it is offered to me masked artisanally. I meditate everyday twice. After waking up and a couple of hours after dinner. I practice meditation along with yoga.
Exercise regularly.

My worst sleep experiences recently have been a combination of 1) I can’t sleep, I am too active. This usually happens when old emotions/unwanted experiences rear up and I have had some success escaping, 2) I go to sleep fine, have some vivid dreams, feel warm, may be a bit hot, with feelings in my belly (not necessarily bloat), and wake up after 4-5 hours 3) I have gas/bloat before sleep. I do things to pass it out/wait it out before I feel ready to sleep. But more often than not, I am gonna wake up with the gas and not well rested, could be 2-3 hours.

For 1) and 3), there are obvious factors to eliminate and work on. 2) however is interesting. In 2), I actually am mentally and physically well rested and capable of going about my day normally. But for quite sometime, I had psychological concerns that there is something wrong with me. Until one fine morning when I wake up after approximately 5 hours of sleep, and a switch flips and tells me, you are well rested, enjoy the morning. I will need to verify that 2) is actually a case of me needing a little less sleep, as I meditate (have been for over 2 years, and some days at the retreat levels).

{ Reply }

Leave a Comment

Desktop version

Raptitude is an independent blog by . Some links on this page may be affiliate links, which means I might earn a commission if you buy certain things I link to. In such cases the cost to the visitor remains the same.