Total Daily Sleep Time Inaccurate

[mechanesthesia]

If you have ever slept more than once in a day in your life, forget about using Fitbit to accurately track your sleep.  No, you are not confused and there is nothing wrong with your eyes.  Apparently the developers at Fitbit are unable to program any sort of function to add your total sleep for the entire day.  In the meantime they are stealing money from us users who purchased these products and even wasted $50 for the "premium" service so we could track our sleep and get accurate sleep reports.

Nope.  Fitbit only tracks the LONGEST sleep record.  

Took a 2 hour nap? Doesn't count!

Woke up in the middle of the night, and went back to sleep?  Doesn't count!

Studies have already suggested that a monophasic 8-hour sleep schedule is a myth.  This sleep pattern has only recently been adopted with the industrial age, but yet a 1/3rd of Americans are unable to adopt this type of pattern.  Historically, a biphasic sleep schedule was more common.  They would call it "1st sleep" and "2nd sleep," where people would normally be awake for a few hours in between... Now doesn't it make sense why sometimes you wake up mid-sleep only to fall asleep a little while later?

Check out these links for more info:

http://www.bbc.co.uk/news/magazine-16964783

http://www.huffingtonpost.com/2012/03/05/8-hour-sleep-myth_n_1322376.html

Also, let's not leave out the polyphasic sleep enthusiasts...

It's frustrating just how little Fitbit understands their target market.  For those of us who like to track everything, we are on top of stuff like the aforementioned links...  Why isn't the company who is SELLING this on top of these things?

I am very upset that the people at Fitbit would go this long without noticing such a simple thing and have the audacity to sell this product like this while misleading their customers.  

EXAMPLE:

Today I took two "power naps" as seen on my log.  One for 17 mins and one for 20 mins:

In the dashboard, it only shows the longest sleep record, which is 20 mins.  This is not accurate:

The sleep report you pay $50 to see also has this erroneous information, it only shows one sleep record per day.  

It gets better.  Guess what, this is how the $50 report portrays the information to you, as a graph showing how you rank with everyone else.  If you actually add it up and average it out, it DOES NOT take into account this "missing sleep time."  It only takes into account the longest sleep record.  So if one day you slept 5 hours, then you slept another 4 hours, those 4 hours don't count and are not added to your average or to the daily totals, and if you do this more than once in a week, you can already see how your total sleep will be under-reported.

This is EXTREMELY misleading.  In my case, I actually showed these graphs to my doctor and was prescribed medicine based on it.  Can you imagine the consequences?

Moderator Edit: Subject edited for issue clarification and word choice

[life-spiral]

Your doctor prescribed medication based on a print-out from a website without looking into the specifics of the data, much less verifying its accuracy? That sounds like a much bigger problem than Fitbit not adding up your sleep times!

[mechanesthesia]

That is besides the point.  The point is a service misleading its customers.  It wasn't the SOLE factor.  My doctor obviously took into account other factors like my personal accounts.  I was simply told, like many doctors do in these situations to keep a personal diary of my sleep, and I used Fitbit instead.  Obviously a personal "sleep diary" is going to be more inaccurate than something that is attached to your wrist tracking all your movements for the entire day, if it WORKS correctly.  He also had a Jawbone up and was familiar with the technology.  And please like, in the US you can't just walk into a doctor's office complain about feeling sad and have Prozac shoved down your throat.  I'm not acting like I'm some kind of "victim."  I am just using it as an example how these records can obviously mislead people to think there is a problem with them, when there may not be or it may not be as severe.  If my "sleep average" is shown as 4-5 hours per day, and it is inactuality 6-7 hours, that can make a HUGE difference in your perception of your health and your life as a whole, and make you think that you have a problem where there is one.  

The way the data is displayed, in a graph form comparing you to others, CLEARLY implies that yes indeed there is a problem.  Why cause customer to pay $50 in order to be deceived into that type of unneeded stress?  People use these problems to FIX their issues by getting accurate information, not to WORSEN them.
I really don't understand what people don't understand here.

[Colinm39]

@mechanesthesiaLet's make positive use of your findings and make this a better product. here is the link for Fitbit Support the Link and they will pass it onto the engineers.

Why not put your suggestion into the Product features because I find Fitbit proactive, takes time, but it happens, and based on the lack of number of requests to add multiple sleeps Fitbit need something to push the design priorities up the list.

[life-spiral]

I agree that it would be great if they offered more options for tracking sleep, I just find your sense of moral outrage wildly out of proportion to the actual offense. I mean, okay, I happen to wear my Fitbit One clipped onto my bra rather than my pocket/waistband, and that is one of the options advertised. BUT sometimes when I walk around carrying something in both arms it doesn't register my steps! And then when I look at the premium section that I PAID for, it tells me that I walked fewer steps than I actually did and so when I compare my semi-acurate information to the semi-acurate information from a non-random selection of other people (giving me information that is interesting but clearly in no way medically pertinent), I have no option but to believe that there's something WRONG with me! And, you know, AT LEAST 1/3 of American women sometimes walk while carrying something in both arms. Right? LOL

[Blade7658]

I agree with Colin, interesting findings, but the negativity makes the point less powerful.  It would seem this is a software fix that could be obtained by grassroots call for a change.

[mechanesthesia]

You are all right. I'm being very negative and I know that. I am just so frustrated with this whole situation. I contacted support several times and feel as if I have been just given the run around. And I had to submit my case a couple of times because it didn't seem they were actually paying attention to what I said. Like I was just given some boilerplate generic response based on skimming the message I sent.
I had the Nike Fuelband and went out and
bought the Fitbit specifically to track sleep. This is their competitive advantage and one of their main selling points.
Unfortunately in life and especially as a customer service, sometimes you have to resort to negativity and shock to be listened to. It's not meant to be pretty. I tried being polite. I am also angry that this is something that has been happening to other users and no one is being helped. I felt that it was time that someone had to collect the facts and speak up.

[mechanesthesia]

Thanks guys for listening! I feel better. 🙂

[gerrymcd]

I think your expectations of a $99 pedometer are widly overblown.  A device like this is mainly for keeping the owner mindful (and honest) about their levels of activity, time spent sleeping, etc. It isn't a full blow sleep or physiology laboratory - so we shouldn't expect similar levels of information, accuracy, or precision.  

[mechanesthesia]

My expectation for basic logic and addition is not " widly overblown."
My expetation that 4 + 3 = 7 hours daily total and NOT 4 + 3 = 4 hours daily total is not "wildly overblown."

This is in regards to the $130 Fitbit Force, which has similar function to the Jawbone "Up" that also tracks sleep.  Some of you on here don't seem to be familiar how it works, but sleep records are manually tracked though the stop watch feature.  You press it when you go to sleep, and press it again when you wake up.  Then you can edit it on the website.  I would expect if I am manually entering the time of an activity, that it would add that time for the day in my sleep reports and in the numer it uses to calculate my average sleep time that week.  This is deceptive and ridiculous that it would just disregard any time there is more than one record per day.

[ILuvFun2]

Have you tried manually entering your sleep time from First time sleeping to last time up. 

I think fitbit will calculate teh actual time slept and show you awake or restless the other times.

NOTE: If you are sitting still in a recliner and not moving that may count as sleep time.

[Meanderer]

You can add sleep manually for each day.

[_dt]

Check out the help topics in Flex 101 and look for the articles on sleep tracking. One in particular, "Why is my sleep graph different on my dashboard than my mobile app? Addresses part of what you're looking for. With the understanding (from the article) that there will be differences, I just check my 11 Mar 2014 graphs where I slept, got up and went back to bed. I stopped and restarted my fitbit flex during the time I was up. 

First to set the scene:  I started sleep tracking at 23:50, stopped at 05:52.  Restarted at 06:09, stopped at 09:17.

The dashboard tile (website, not mobile phone) shows just before 24:00 to just before 06:00 [@6 hrs] and says 3 hr. 19 min of sleep.

When I click on the tile, the pop-up shows both time periods.  When I hover over the first, it says 3 hr 9 min and the second says 57 min.

When I click on the arrow in the bottom right, it takes me to the sleep log which show the two graphs you show in your initial post.  They say 3 hr 9 min and 57 min the same as the popup. 

On the right edge of the sleep log page are the 30 day sleep stats.

Time Asleep for 11 Mar says 3.15 hrs (3 hours 9 min)  which is a match for the longest period.

Total Time Asleep says 4.1 hours (4 hours 6 min) which is correct for the total.

I don't have premium, so I can't speak for that.

Do you really only get 30 minutes of sleep a night?

[Leopardfish1]

I think mechanesthesia is exactly right.  This "$99 pedometer" should be just fine at adding 2 or more sleep times in a day so that a total sleep time can be accurately calculated and shown on the dashboard.  As a matter of fact, IT SHOULD show an accurate tabulation of sleep throughout the day.  For it to do otherwise certainly does qualify as a major short coming of the technology.  I hope that Fitbit catches that short coming and corrects it because it is in error that only the longest sleep time is counted.  How useless.  It might as well not even track sleep and just allow for manual entry of sleep stat's.  Fitbit programmers take notice please.

[Sleeper_Awakens]

I agree that sleep tracking could use a little more sophistication and hope that fitbit can choose to improve upon it. But I'm not sure it's as simple as adding up time per "day". Has anyone noticed all of the "Sleep Stats" (the right column of the log) for a single "day" aren't satisfactorily accurate? _dt has one example.

My example: my current "Sleep" record shows "Times awakened = 13". However, if I hover over "Apr 12" on "Times awoken over the past 30 days" I only see 7 times. That suggests to me sleep tracking is based on a midnight to midnight schedule. Therefore, in addition to being able to sum a "'day's' worth of sleep", tracking needs to be aware of what you consider the start to your "day".

Let's not let mechanesthesia's "OVERBLOWN" (not my word) frustration detract from the fact that there is something wrong and it is worth making an effort to improve it.

What would be your "requirements" to make the Sleep Tracker/Log more accurate?

[AimGal]

This is becoming a great conversation.

To answer the last question posed, I'd LOVE to see a "total sleep time" way of tracking. My typical sleep pattern is sleep 4 hours, wake up for 1.5-2 hours and read, sleep another 3 hours. Ideally, I'd love the fitbit to show my 1.5-2 hour insomiac period as "awake" time in the middle of a longer sleep time, as that's what it is. It would help me to be able to show my doctor hard evidence of true insomia hours as opposed to my own report. Why? Because if you say "I'm awake two hours in the middle of every night" many doctors figure that you're like most people and overestimating. It's not that I'm waking multiple times and thinking I haven't slept in between. I'm truly just not sleeping. If fitbit could find a way to track this, I'd be able to use it to help show how my insomia may or may not correlate with other health issues I'm having by charting. I'd LOVE to be able to do that. It becomes objective rather than subjective information, and that's invaluable.

[idahopacker]

I agree with mechanesthesia, fitbit needs to address the way it calculates sleep.  I have worked in the IT field for over 30 years and this should be an easy fix for their programmers. 

[NEq1Guy]

I am pretty appalled that the website has this sort of deficiency.  The iPhone App can correctly tally up the data from the exact same data back end without any problem.  There is literally no reason why the website can't do that as well.  To say I am disappointed in their inability to address such a trivial deficiency is an understatement.

[Colinm39]

@NEq1GuyMy Android won't let me sync but are you saying that the iPhone displays the total sleep time on the tile. Is it broken into the times during the day or is it one total for the day. ?

[NEq1Guy]

Ironically the app is better on the summary but not on the break down.  I can't look at the detailed sleep of anything but the first sleep period.  However on the summary screen:

You can see it's tallied up the details of the two periods:

If you take this into a weekly summary plot it gets the values correct as well.  It may be acting a little buggy still but for the most part it seems to work in the iOS app.  What behavior are you seeing?