Found a significant bug:

seagete ces-pas

My experience has been consistent with this:
the total calories number appears in (at least) five places:

• Watch itself (accurate)
  
• Web dashboard steps widget (accurate)
  
• Web Steps Page (Inaccurate)
  
• Web Reports (Inaccurate)
  
• Android App (Inaccurate)
  

On some days they are all the same and everything is fine, bit on some days 3, 4 and 5 are inflated by some weird double counting. I thought it was related to activities, but I found at lease one day in which the problem happened and I didn't log any activity.

Myfitnesspal reads the inaccurate number, so on the days when the problem happens it will tell you to eat more, if you don't realize that double counting is happening.

This is good to hear that they are now aware of it. With the support guys I'm in contact with after 3-4 email back and forth they still don't see the discrepancies so just now I've had to take screenshots and hightlight the dates where there are discrepancies. I didn't know that I now work voluntarily for Garmin!

I reflect exactly what you have seen. I would also add that it's also distance that can be wrong/different on some days too and that the calorie/distance discrepancy is not always doubled by an activity - sometimes is just a slight increase with no known real correlation with anything else (i.e. could be an increase of about 80kCal or so).

Some more data points that someone from Garmin will hopefully take notice of...

It's now 3pm and i've been a sloth all day. I got to bed after 2am this morning and didn't get up until 11am. From then i've mostly sat around relaxing, only accumulating 672 steps. Yet somehow according to both the FR235 and GC, i already have 240 active calories! If i was running that would take ~4km to achieve!



Looking more closely, i see that early this morning (while awake) my HR apparently measured 129BPM for ~30 minutes. Exercising with such a HR would explain the 240 active calories. But of course i wasn't; it was just an anomalous HR reading that took half an hour to get updated.



My take away from this: daily calories are NOT calculated from steps, but from the 24/7 HR (though it could be a combination of both). Hence, with the low HR sampling any high readings (either accurate or not) will have a large impact on your daily calorie estimate. It will always push you higher.

Finally, from my experience both the watch and GC have been consistent (and consistently wrong). This makes sense, since the HR data comes first from the watch.

Some more data points that someone from Garmin will hopefully take notice of...

It's now 3pm and i've been a sloth all day. I got to bed after 2am this morning and didn't get up until 11am. From then i've mostly sat around relaxing, only accumulating 672 steps. Yet somehow according to both the FR235 and GC, i already have 240 active calories! If i was running that would take ~4km to achieve!



Looking more closely, i see that early this morning (while awake) my HR apparently measured 129BPM for ~30 minutes. Exercising with such a HR would explain the 240 active calories. But of course i wasn't; it was just an anomalous HR reading that took half an hour to get updated.



My take away from this: daily calories are NOT calculated from steps, but from the 24/7 HR (though it could be a combination of both). Hence, with the low HR sampling any high readings (either accurate or not) will have a large impact on your daily calorie estimate. It will always push you higher.

Finally, from my experience both the watch and GC have been consistent (and consistently wrong). This makes sense, since the HR data comes first from the watch.

Yes, that matches with my conclusion. The calories on the Daily steps line and not directly linked to the calorie value and seem to be HR related. Of course if you do a lot of steps outside of a recorded activity then your HR should go up and you may get calories allocated. However low sampling rate can throw these readings out and I have had exactly the same situation as darrencroton where the sampling has fallen on two peaks and assumed a similar HR for that whole period and allocated calories. Sometimes as others have noticed I get erroneous HR spikes recorded and this can throw the readings off also. I'm sure even a small increase in sampling rate could make things far more accurate than they are now....

I'm starting to think my issue is more with the resting calories. It's 1:21 pm, and my resting calories are at 1591. I'm 5'6" and 132 and 31 years old. According to most BMR calculators, that is more than I burn in a day from just being alive. It usually isn't that high this early in the day due to my workouts usually being in the afternoon/evening, and I feel like activity should affect that. I did a 2 hour mountain bike ride and a 2 mile run, so I'm thinking that is somehow affecting my resting calories even though nothing I do should change that each day other than weight fluctuations.

I'm wondering if that is playing in to the calorie discrepancy. Right now, GC has me at 2745 for the day. My watch has me at 1508. My Fitbit says 1544, so I know the watch must be the one that is right.

Anyone else have higher resting calories after an activity?

I'm starting to think my issue is more with the resting calories. It's 1:21 pm, and my resting calories are at 1591. I'm 5'6" and 132 and 31 years old. According to most BMR calculators, that is more than I burn in a day from just being alive. It usually isn't that high this early in the day due to my workouts usually being in the afternoon/evening, and I feel like activity should affect that. I did a 2 hour mountain bike ride and a 2 mile run, so I'm thinking that is somehow affecting my resting calories even though nothing I do should change that each day other than weight fluctuations.

I'm wondering if that is playing in to the calorie discrepancy. Right now, GC has me at 2745 for the day. My watch has me at 1508. My Fitbit says 1544, so I know the watch must be the one that is right.

Anyone else have higher resting calories after an activity?

Looks like the calories for your activity may have duplicated itself in the resting calories? Did you record the Mtb ride with the watch or import it from another device?

I only record my activities on the watch itself. Nothing feeds into it except MyFitnessPal, but only the Garmin writes to MyFitnessPal. It's now 5:20 and my resting calories are still at 1591 which means I've burned nothing just by living since my activities. It seems to do that - jump up with activities and then stay steady the rest of the day. On days before activities, it goes up slowly, more like I think it should.

Two similar days -- different calories

This past weekend I was on a trip and did a lot of walking with my family (and no runs or other activities). About the same amount both days. I have a fitbit one, and it gave me almost identical calorie totals for the two days, but I had a 600 calorie difference on Garmin. The second day seemed to spike at the end. There was no double counting of steps or distance -- they both seemed reasonable. It was just with the calories.

day one: Daily Steps 22,506 689 -- 11.33 mi -- --
day two: Daily Steps 24,870 1,323 -- 12.52 mi -- --

And total calories day one: 2,669
total calories day two: 3,303

The various garmin dashboards are consistent, just not reasonable. Fitbit gave me approximately 3080 both days.
The second day does shoe three hours with a HR of 105 that is clearly one sample. The first day doesn't show any reading over 81, which also doesn't seem reasonable as I did walk 11 miles or so, including some up and down. Seems like I would have spiked up for at least a few minutes.

Just noticed this as well and I think it's related to the 3.30 update which I did Tues evening. Oddly enough though while Google Play claims I have 3.3, the app itself under Settings says 3.2! Running Android Lollipop.

In any case, my calories which used to track fairly closely to the amount of walking I did (I don't use the Vivosmart to track activities, I use other Garmins for that) now suddenly are way overcounted.

Tues data (version 3.20) -- bit of pilates, moderate amount of walking




Thurs data (version 3.30) -- mostly sat on couch, small amount of walking


So do I try to uninstall the app and reinstall?

Yes, apologies. I searched the forum for a similar problem and found your thread. It's from a Vivosmart but the data is on Garmin Connect app.

Will repost over at the Android app and/or Vivosmart part of the forum.

I've been using the Vivosmart and app for nearly a month now and only recently have the calories and distance/steps not shown the same pattern -- coincidentally since I updated the app. So I don't think it's a misinterpretation of the data. Though I was wrong, it is 3.2.1 now, not 3.3.

Agree that it would be nice to see a breakdown of how the data is generated wrt calorie burning. I believe it's HR-based, but my HR trace didn't look much different yesterday to any other normal day.