Finding the Best Sleep Tracker | Hacker News Reader

Finding the Best Sleep Tracker

karpathy.bearblog.dev

61 pointsnreece17 comments

Native Apple Watch wins in most scientific papers and also in tests done by “the quantified scientist” by a large margin. Apple Watch does not use HRV but accelerometer for movement and breathing rate.

Apple Watch also wins for HR during exercise.

So if you want one device for everything, there is little competition yet.

If anyone wants a tool to show you this analysis automatically, I make an app that links with Whoop and Oura and Apple Health and it shows you all of those relationships: https://apps.apple.com/us/app/reflect-track-anything/id64638...

I'd love to see this comparison between Garmin and Oura. I ditched Whoop because it was too easy to get 100 sleep scores, but my Garmin watch is much harder to please. I do think it offers better signal, but I've never scored over a 90, so maybe it is too critical?

I like garmin, and have to keep working to get a good sleep score.

that said, when I do get very good or very poor sleep, the score reflects it.

I've found Garmin's sleep tracking to be unreliable. For instance, if I lie in bed reading before I fall asleep, it often incorrectly logs that time as sleep.

I had over 90 scores, and only when I get earlier in bed and if there are medications involved. you do feel like you had a great sleep so for me it's very accurate.

Why did the author not evaluate the sleep tracking that comes with the Apple Watch out of the box?

From TFA: Maybe there is a better app on Apple Watch for sleep tracking, but I haven't found it.

And what other apps were tried? As you ask, why not try the default? From what I hear, it is “better” (for whatever that means) than Garmin, for example, on sleep tracking. I dunno, it is kind weird to start comparisons with some random paid app rather than the built-in that Apple offers for free.

Apple's health app is extremely careful not to say what is 'good' or 'bad' about the metrics it monitors, I find. The sleep tracking really just gives you a breakdown of sleep stages and length of sleep, rather than a sleep score that the o ther apps give you. I believe that is what Karpathy is comparing, not the telemetry itself, rather the measured sleep quality mapped to his perceived sleep quality, which Apple doesn't really give you.

Nice article but I with there was something more authoritative to compare it to. I know controlled in-lab sleep studies are expensive but this would be the key data to have in the mix.

The Quantified Scientist YouTube channel by Rob ter Horst does a good job imo at comparing trackers against an industry standard baseline. For sleep he uses the Z-Max EEG device.

Really surprised that Autosleep performed so poorly. I've been using it for over a year and I'm pretty satisfied, and chose it because it was the only apple watch sleep app I could find that didn't require a monthly subscription. Admittedly I don't look at the 'score' so perhaps that's it, I just look at the overall 'quality sleep' hours and have found that tracks pretty well with how off-my-game I am that day.

I enjoy Peter Norvig's writing style—it's rich in detail and backed by solid statistics. I've also been following Andrej Karpathy for the past two years, and I have to say, his writing has become my top priority to read first.

My Garmin Instinct Crossover watch does a great job of sleep tracking. It correlates well with how I feel the next day and how active I was the day before. On the other hand, Sleep as Android inflates the scores too much to be useful.

I have a (LONG discontinued) zeo, but stopped using it after I ran out of the contact patch things. Too bad, it was a true EEG sleep tracker.

I just dont get it.Sleep tracker? Dont you just know?, about something as personal and intimate as your own state of restedness?,energy levels? And quite honestly I worry about the cost of optimising ones life and managing everything for what?, a better performance review?, and does this kind of minute management and reliance on external sensors and data, leave a person able to make criticl descisions when facing anything outside of a habitual and specialised routine. Perhaps these could work as training and awareness devices?, which could be revealing, but I think habitual use is only going to work, for the most dedicated ...military precision types.... but for them, it will be competitive, and it's bad news for anyone outside of that group, to compete with them. And as a reminder, there is nothing quite so luxurious as well earned sleepyness, and the general feelings of well bieng and satiation that creates this, vs an off switch.

Sorry, but I think this guy (The Quantified Scientist) has the article beat in terms of rigor and analysis (As some other peoole mention) https://www.youtube.com/watch?v=kCVf3QPm-8w

- Similar graphs

- Medical grade baselines

- His whole channel is dedicated to this kind of stuff

- The original post is good, its just this channel is insanely good

Reply on news.ycombinator.com