Comment by Cthulhu_ | Hacker News Reader

Comment by Cthulhu_ | Hacker News Reader

Parent Full thread

* Three hours of sleep - if that - is far, far too little for any normal human being to function normally. Also yes, working until 6 AM is an all-nighter, it's dawn then.

> Second, how many times did you pull a few all-nighters in a row during college? I can't even count. I certainly didn't die.

Never. I don't get why people do it either, it's not like it makes you a better student. If anything, having to pull all-nighters for college can, in my limited and breezing-through-school opinion, mean only a few things:

* You only started to work on an assignment / study for an exam at the very last moment. * The US school system is terribly broken and puts a disproportionately and unhealthily high workload on its students, turning it into a Chinese sweatshop instead of a nurturing educational facility.

In software methodology, there's an expression called 'Sustainable Pace' [0]. When building software, it's pointless to go and do 'crunch time', since, one by one, your people will start underperforming or even outright crashing, leaving the remainder under duress to compensate. Doesn't work.

I don't know what these 'deals' are, but I can't see how they'd require anyone to work 21-hour workdays for any duration of time.

Personally, I work a 9-5 job as a software developer / consultant, and I close the door behind me at 5. I resist my colleague's excitement over DOING ALL THE THINGS!11one, simply because I know I can't deal with it. I've seen enough colleagues (and my superiors, lol) work too much and unable to let go of their job and becoming unfit for work. So yeah, I like to dramatise this a bit, and not allow myself to fall into the same traps. It's not like I'd gain anything from it, anyway.

Work to live, not the other way around. etc.

[0] http://www.extremeprogramming.org/rules/overtime.html

College students who pull all-nighters do so two reasons:

1. They tend to be terrible at time management and prioritization. They procrastinate until the last minute and then have no choice but to cram.

2. They do it to create a narrative that can protect both their ego and their image. "I failed the exam despite pulling an all-nighter - it was just crazy hard!" sounds a lot better than "I failed the exam because I'm dumb" or "I failed the exam because I didn't start studying until the last minute."

Yes, I see having to pull an all-nighter as a big failure on my part. I did once or twice, but I noticed it's just counter productive. I was adding negative value at 2-6am, it would have been better to sleep and continue the next morning, or just let it go.

I'd like to disagree - at least for my university, which to be fair is notorious for being brutal in terms of workload. It's not unusual for me and many of my friends to have so much work for classes and extracurriculars over the course of a week or month that there literally isn't enough time to do all of it and also eat and sleep in the 168 hrs/week that you have.

Allowing that there may be exceptions such as your own case (assuming you are properly assessing your situation), would you still agree that the generalization holds for the vast majority of cases?

Not at my university, but I wouldn't be surprised if it was somewhat true generally. I think a valid comparison might be between a hard sciences student at a top school who's also involved in a lot of extracurriculars and up at all hours just to stay on top of everything, and someone working on an early stage company where there's always something burning. The opposite end of that would be the investment banking intern spending 100+ hours a week at the office spinning their wheels or college student who does nothing all week and then stays up to make themselves feel better about failing the exam. That's sleep deprivation for the sake of appearances.

There's that, yes, but I think there's a point to be made for workload. I remember back in college all the courses would inevitably pile the workload on at the same time.

Around the same 2 weeks during a semester the profs will load up on all the projects, labs, and other must-dos, across all your courses. Taken separately these workload increases are modest, but as a whole you literally start running out of time.

My coping mechanism back then was to strategically not do some of these things, but then again my only intention going to school was to get a piece of paper that would unlock the Magical HR Gates at companies. My GPA isn't exactly great.

>>Around the same 2 weeks during a semester the profs will load up on all the projects, labs, and other must-dos, across all your courses. Taken separately these workload increases are modest, but as a whole you literally start running out of time.

At least for me, this became an issue only if I didn't do the homework assignments and study the class notes. What would happen is that we would get a large project, and I would realize that I didn't even have a firm grasp of the material. So I'd have to work two or three times as much, first to understand the prerequisite subjects, and then start working on the project itself.

After I figured this out and started to study regularly, projects started to become progressively easier, and as a result took less time. Even when professors loaded up on all of them within a short period of time, I could knock them out quickly because I was already very familiar with the subject.

It's like picking up a new programming language. There's the learning curve of the language itself, and then there's the learning curve of all the tools. But if you are already familiar with the tools themselves, you can learn the language much faster and get projects done much more quickly.

I'm of the opinion that people can adapt to less sleep. I spent quite a while sleeping exactly 6 hours per night, and then spent about 11 weeks alternating between 3 and 4.5 hours of sleep on weeknights and about 7.5 hours of sleep on weekends. This almost certainly wouldn't be a healthy thing to do for years on end, but for the 11 week period that I did this I felt pretty good (maybe 90-95% mental capacity) and was able to get a ton of stuff done with my ~20 hour days.

Again related to your post, this sparse sleep schedule was caused by college. I go to Caltech (for those who haven't heard of it, its probably pretty similar to MIT in terms of workload/intensity) and have a fairly typical sleep schedule among my peers. I'm pretty sure that every term (10 weeks) that I've been here I've had at least one 40+ hour period with 0 sleep, and had a term where I would sleep from about 7AM to 10AM every weekday (and then at least 10 hours/day on weekends). These longs hours are mostly caused by problem sets + studying, and things get a bit worse when you're actually too busy working to have time to go to class (which is a common occurrence).

As for why students do it, its to get everything done. Typical courseload is 5 courses/term, and starting in the middle of sophomore year most people are taking courses that would be intro graduate level courses at most institutions. 4 advanced math + 1 social studies/humanities course definitely keep you busy.

Also, among my group of peers, we typically define an "all-nighter" to be where you just complete the planned activities for the following day without any sleep at all. This generally involves staying awake from one morning (10AM) to dinner time on the next day (~5PM), making an all-nighter be ~30 hours. If an all-nighter were defined more leniently, we'd have a depressingly large number of all-nighters.

Note: Most of the above applies to my more difficult terms. I've also had terms where I slept 7+ hrs/night and maybe only worked until the sun rose about 2 or 3 times.

> I'm of the opinion that people can adapt to less sleep.

Your opinion has been proven wrong by every single scientific study on sleep. Humans do not effectively adapt to less sleep. Error rates go way up, productivity goes down disproportionally, focus goes down, etc.

Although I could have expressed myself better, you still quoted me out of context. I was particularly talking about people dealing with less sleep on a relatively small timescale (less than 3 months). I'm choosing the 3 month window because it replicates the length of a college term and greatly reduces the risk of burnout (which I don't think is as big of an issue in college as in industry, because typical college schedules give at least a few weeks off after each term).

Are there any studies that compare how much people can done in a few months on different sleep schedules? I'm imagining some sort of cognitive task (like reading books and passing a comprehension test, or working through all the proofs in a math book) where the goal of the participants in the experiment is to accomplish said task as many times as possible in a fixed time period. This would be very difficult to design because one needs to decide on cognitive difficulty for the task (if the task was very easy, then less sleep = more time would almost certainly help) and have roughly equivalent ability between the 2 groups. This study wouldn't capture dealing with context switches or retention of information; just how much can be done in a fixed time period.

As far as I know, such a study doesn't exist. Anecdotally, I've seen many people make the decision to trade sleep for more time even while working on cognitively difficult tasks. Personally, I feel like I accomplished more during my terms in school than I could have sleeping 7+ hours/night.

On the other hand, I acknowledge long term sleep deprivation is not good for long term health, and I've personally seen my ability to run with intensity suffer during periods of low sleep.

In summary, I don't doubt the long term effects of low sleep on health, I slightly doubt the negative long term effects of sleep on productivity (know people who have been doing uberman for years and swear by it), and I heavily doubt that people can't be more productive over <3 month window by sleeping less than the recommended 8 hours per day.

Those studies have been performed several times [1][2][3][4]; most of them indicate that moderate sleep deprivation (<6 hours) produces the same amount of cognitive impairment as being legally drunk. Many of them also show that people are unaware that their cognitive function is impaired when sleep-deprived: they may feel like they're productive, but their actual performance is riddled with errors that then need to be corrected.

[1] http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1739867/pdf/v057...

[2] http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1071308/

[3] http://psycnet.apa.org/psycinfo/1997-07865-006

[4] http://en.wikipedia.org/wiki/Sleep_deprivation#Effects_on_th...

Here is my problem with people like you pulling out their study reference lists against every anecdote: results drawn about people in aggregate do not necessarily mean much for this specific person in this specific situation.

"I am not the exception" is a good principle to live by, but it is also important to note that sometimes you are the exception, you are in exceptional circumstances, or you can become the exception (through exceptionally hard work for example, or exceptional stupidity. Being the exception is not always a good thing).

My experience is similar to lightcatcher's. When I was younger, before I was married and had kids, it was common for me to sleep 0-3 hours for a few days while I was working on something and then crash for a few days, sleeping 10-14hrs. Basically I slept when I felt tired, stayed up if I still felt alert. At the time I did not feel less productive and I did not feel like I was putting out lower quality work. Looking back objectively those were the most productive times of my life to date, and the quality of the work was on-par or better with anything else I have done. Temporary sleep deprivation under specific conditions was highly effective by all the objective measures I can think of.

Okay, but driverdan was saying "people can adapt to less sleep" not "I have adapted to less sleep." As if it's reasonable to expect that from people in general when, empirically, it's not.

Fair enough.

> I'm of the opinion that people can adapt to less sleep.

Fortunately, we have facts about this, so we don't need opinions. Its not true, although it can appear true over short periods of time. More importantly, the amount of sleep a person needs can vary a lot between individuals - you might only need 6 hours/night, but that doesn't make it physically possible for everyone to survive like that.

Looking at the just the productivity aspect of sleeping less, I have yet to see a study that accounts for the extra waking time that people have when they sleep less.

How much you get done = rate of work * time working.

Many studies have shown that the rate at which people can "work" (which means something like respond to visual stimulus) decreases when people sleep less. I haven't seen a study that compares this decrease in rate to the increase in time available for work.

If I have to do a task 1000 times each day and each time takes me 2 seconds longer on 5 hours of sleep than it does on 8, it'll take me 2000s ~= 34 extra minutes to do the task 1000x when I'm on less sleep, but then I still have 2.5 hours left out of the 3 that I decided not to sleep. (Numbers from this inspired by [1])

I acknowledge the health issues of chronic sleep deprivation.

[1] http://www.huffingtonpost.com/2012/07/30/sleep-deprivation-w...

You provided a single example and then skipped the alternatives.

Reply on news.ycombinator.com