Hacker News new | past | comments | ask | show | jobs | submit login
Sleep quality and stress level matter more than languages or practices (twitter.com)
244 points by ognyankulev 35 days ago | hide | past | web | favorite | 52 comments

I think this is true for the most part, although poor development practices can eventually catch up to you (I once worked in a place with 5000 programmers working on the same thing and whose productivity averaged 1 line of code per day -- it can happen.. :-P ).

However, while many people will rightly complain about stress in the workplace and long hours coding, I'd like to also point out that stress and lack of sleep can come in a variety of packages. I've seen good programmers destroy their careers because they thought there was no problem staying up drinking all night on workdays. I've experienced (unfortunately) being in a bad relationship and arguing with my SO all night until it was time to go to work.

Stress and sleepless nights will cost you in your career. It might cost you a lot. It's not sexy to be the person that avoids drama in their life and manages their sleep schedule well, but it can have pretty significant rewards. If you think that in SV the difference between a top and middling performer can easily be in the $100K per year area, it puts a pretty hefty price tag on "Oh, I'm pretty sure I can catch up on my sleep on the weekend".

From the "hard to talk about, but I wish I knew this 20-30 years ago" department :-(

> being in a bad relationship and arguing with my SO all night until it was time to go to work

If that ever happened to me I'd seriously consider my life choices in the romantic department.

That's pretty much my point ;-) At the time you think, "I have to make this work". You feel like you are failing big time and you have to try harder. You don't necessarily realise the cost of that effort. It's weird to say, "Is this really my priority?" Relationships don't always work out and that's fine. Don't let a bad relationship bulldoze the rest of your life too is what I'm trying to get at. It's great that it's never happened to you -- some people are either lucky or skilled in this area :-) Until I met the person who is my wife now, I was neither ;-)

...you’ve never had an all night argument with your partner?

No, never.

I've been with my current SO for over 10 years (and married) and had a couple long term relationships before that.

We've had arguments like any other couple, but never an all night argument.

The big picture is trivially true -- if you don't sleep, you die. Where's your favorite language, now? Similarly if you don't eat -- yet few people would make a rant about "food quality matters more than languages".

There's still a lot of nuance. What studies there are don't negate the existence of variability either. Well, then what's the point of bringing up the subject? Taken broadly sleep is important as life itself, but then it has no place in an earnest discussion around improving productivity. Taken narrowly, it's very context sensitive, so it's again rather useless in discussing improving your own or even your team's productivity unless you're already in the weeds of knowing individuals' detailed backgrounds.

There's a fun fact johnc's comment in that thread points out. We can all agree if we want that if people aren't getting their 8 hours (or some number close by), that has to happen first, and we can ignore whatever nuance/variability/context objections one might have. But with 8 hours, there are still over 100 other hours in the week. 80 if we're just talking M-F. What are you going to do to fill those more productively? If switching languages lets you do something in half the time as otherwise, why wouldn't you switch, and rather than just taking half the day off, continue working and do twice as much as you would in the other language for the day?

Is the point that you shouldn't unnecessarily waste your waking life for some company? Fine, but you shouldn't do that even if you're getting 10 hours every night and are only asked to work 30 hours a week if those 30 hours are truly a waste. On the other hand, you might be in a situation where 80+ hours this week (hopefully not forced to be compressed to M-F or intrude on your sleep requirements) spent on work is the most non-wasteful thing you can think to be doing with your life at this time. Circumstances are different.

Sleep is only a part of it — avoiding burnout (which is the real goal here) requires taking time not only to sleep, but also to work out, do your laundry, cook and eat food, see friends... we’re not just our jobs. We need lives outside of work or we will burn out — often at the most stressful and least opportune moment. If you’re working 100+ hours a week, you’re almost certainly offloading all that extra work to someone — usually a romantic partner, which can damage the most important relationship in your life.

That leaves at most 60 hours a week to truly dedicate to work. The people who work more are sacrificing quality of life for productivity. While a very small minority of people really can maintain that, most of us can’t. We shouldn’t take advice from the outliers who can. We shouldn’t work for companies where a schedule that leads to burnout is the norm. It’s a signal that the company wants to suck you dry then toss you away.

Furthermore, with knowledge workers... we often never stop thinking about a problem. The thinking is “real work” — I’ve had inspiration about how to solve a problem strike during dinner. Whip out the phone, send myself an e-mail, then deal with it in the morning. But the obsession with maximizing individual productivity at all costs is literally killing us.

I would think, in absence of other evidence, that investing in microoptimizing sleep and stress would have more benefits than any language or tool choice.

So it's not just 8 hours, it's also midday naps, circadian rhythms, it's letting people sleep when they need to.

And you didn't even mention stress and there are of course myriad ways to improve mood levels and reduce stress.

And obviously some of that stress will come from your choice of tools and methods, but again, we have little evidence as to which tools and methods produce the least amount of stress.

It's not what you do, or even how you do it, it's the circumstances in which you do it that have the greatest impact on your ability to do it well.

Work 80 hours, work 5, play 50, dig ditches, cure cancer - doesn't matter, fix sleep and stress before worrying about the little things.

Given that your big picture conveniently ignores stress it is not "trivially true" after all. :-P

I intentionally left out talking about stress, but I think it's also trivially true at the big picture level, too much stress can kill you. X is true, Y is true, X and Y is also true... But stress is a much more vague and hazy notion than sleep when you get down to specifics, we have few rules of action to agree on even in principle like "get your ~8 hours of sleep (even if that involves napping) before you look at other potential boosters" and we have contradictory notions about certain types or levels of stress actually being better for you in the long run whereas chronic sleep deprivation seems to have no long term benefits.

The book ‘Why we Sleep’ has a lot of the detail on how sleep (or the lack of) impacts you.

Surprisingly, the Joe Rogan interview covers most of the book and is pretty funny: https://www.youtube.com/watch?v=pwaWilO_Pig

If you have a long commute, I can recommend it.

Great episode! The part about memory centers sort of reminds me of GANs. Discriminate during day, generate during night.

Definitely strikes a chard in me - I've completely stopped coding when I'm tired as it has often led to disasters. I've just learned to live this limitation but its not been too hard.

I read here recently, "For a time, I worked 12 hour days. I'd spend the first 4 hours of the day fixing mistakes I made in the last 4 hours the previous evening."

I’ve been wondering about this WRT the 969 policy at companies in China.

Why did they not immediately see a drop in productivity when implemented?

Is China’s software industry making itself less competitive by not letting its workers sleep enough?

There's no drop in productivity if you don't care about quality.

The sibling replies neglect to mention that midday naps are very common in China (across all age groups), with lunchtimes typically being 2 hours long.

It's really not that different to Google having nap pods.

Google has nap pods? Wow...that remindsof mats they used to have for us in preschool. Sheesh.

Can't tell if you're being pejorative or not, but I make it a point (as a regional director) to take naps when I'm tired in full view of anyone who wants to see.

Yo don't always have the luxury of waiting till you're in a nice cool bed with the environment and time of your choosing.

People need sleep when they need it. Pretending otherwise is foolish.

Stigmatizing naps is so boneheaded, petty and backwards that I'm surprised to see anyone on this site so casually write something like this. Literally read anything on the subject, please?


Read deeper into the post, Im not stigmatizing naps. Im big on sleep. Love it so much & think it's fundamental to physical and mental health. Its so important that if you need it at a place of employment, maybe that place of employment needs to do a better job of load-balancing.

Looks like Mr. Grump here could benefit from a nap!

Because they don't care about the afterfall. They have an expendable work force and bribes to work around penalties.

Maybe some people have the time to do creative and patient work, while other engineers crank out GUIs and iterative corrections. Presumably Andrew Ng isn’t being asked to demonstrate his loyalty or commitment.

Isn’t it the same in all professions?

I think part of the problem is that it's so difficult to actually measure productivity in software development.

Measuring productivity is impossible because you can’t do repeated experiments

I've never heard of this 969 policy. What is it?

It's rather 996 : work from 9am to 9pm, 6 days a week.

This is why I recommend every developer read Deep Work by Cal Newport and Why We Sleep by Matthew Walker.

Deep Work in particular made a bigger impact on my programming than any thing else in ~25 years, including languages frameworks and tools.

What I’m saying is, good sleep and the principles from Deep Work made more of an impact on my productivity than switching from Java Spring, to Ruby on Rails, or to functional JS and Serverless. Put that in your pipe and smoke it.

Just bought Deep Work because you mentioned it; Newport's So Good They Can't Ignore You was refreshing.

Yes, this comes up often. I think it is difficult to be disciplined about it. To catch yourself when you are not well rested and act accordingly.

If you are going to write that much, please use a better platform than Twitter.

> Twitter is my stream of consciousness thoughts, blog (http://hillelwayne.com ) is my heavily edited and revised thoughts


That explanation actually kind of makes it worse.

My favorite talk on Sleep by William Dement https://www.youtube.com/watch?v=8hAw1z8GdE8 September 23, 2008 Google Tech Talk

Very entertaining.

Additionally, a low stress non-crunch environment is better at retaining talent, which is significant on the short term but even more so on the long term (in a whole host of ways). Keeping people around longer means more retention of experience, less disruption to the corporate culture over time, more predictability in execution on projects, a more stable work environment (and less stress) for everyone else, etc, etc.

Of course. This is the least controversial opinion out there and it borders on fact. Anyone who claims otherwise is an idiot or an exploiter. There's simply no substitute for sleep in any discipline or in life in general. It simply doesn't exist as far as we know. People who short themselves on sleep are self harming, often out of stupidity. I've been there as have many because it's part of our stupid work culture, but once one wises up, there's no going back. Sleep quality and stress levels matter more than anything. Period.


Then again, this is not repeated enough! Moreover, I have the idea that companies don’t know, they should.

I had chronic insomnia for several years. But then I’ve stopped coding right before sleeping at least 1-3 hours in advance and now I can finally sleep. Who would’ve thought.

I am just finding that out now. I read something that it could be due to the blue light from LED screens, TVs, phones, etc. I hope to end my insomnia too.

Blue light before bed has been clearly shown to affect circadian rhythms and sleep. Some references here to get you started:


I've felt this for a while, been actively monitoring my sleep for over a year and try to hit at average 7.5hr of sleep nightly. (This is what I found works best for me).

I've found it helps with my mood, and my stress and also productivity. I've also found that it is possible to have the same effect of not enough sleep, by getting too much sleep.

Do you find any variance in the 7.5 hour 'sweet spot' with the seasons (assuming you have them) or with your activities?

Not much, no.

I sleep well since i use Clojure

This is gold.

Related: determination matters more than intelligence

Maybe tools don't matter much either, but... essays on twitter?

What's with this trend of multi-tweet posts being seen as noteworthy articles on HN?

Probably because the quality of the content mattered more than the platform it was posted it.

Sure, but languages and practices contribute significantly to sleep quality and stress level.

Yeah, man, but I'm doing well. I'm awake at 0730 San Francisco time after sleeping more than 8 hrs before, I eat a balanced breakfast, time my lunch and dinner appropriately, deadlift a respectable-though-not-special 350 lbs, run a 6 min mile, make about half a mil a year, have friends I love dearly and who love me, and everyone I know closely has a good life.

Some dumbass downing Red Bull to type `console.log("got here")` for the fiftieth time in the last 36 hrs without sleep isn't on my level.

All these things aren't things for Mr. Red Bull. They're for me trying to close the gap in productivity to Carmack.

Or rather that's all true for my friends. I can only deadlift 345.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact