Difference between run time and moving time

Hi, I would like to understand this statistics because they doesn't add up. My running time is correct because I stopped at a street light for 4 seconds, so 1:07:12 - 4 sec = 1:07:08. But I don't understand the moving time. 8 seconds less than run time. How is it possible that I'm running for 1:07:08, but I'm moving only for 1:07:00? As you can see from the walk/run graph, the watch already calculated the inactive time (yellow lines which equals to 4 secs), but still it says that my moving time is 8 seconds less than my run time (1:07:08), which isn't possible.

Top Replies

All Replies

  • From the documentation

    So the pauses (both auto-pause and manual pauses) are not included in Time and in Moving Time. Additionally, if you are inactive (not moving) during the activity, you will see the yellow gaps in the graph, but unlike the pauses, the time is included in the Time (not in the Moving Time though)

    I understand all of that. I am saying that auto-pause (or manual pauses) cannot explain the following inconsistency:

    1. The activity time is 1:07:12

    2. Moving time is 1:07:00 (which means that the moving time algorithm determined OP was "not moving" for 12 seconds).

    3. According to the run/walk algorithm, OP was running for 1:07:08 and idle for 0:04.

    Note that the total time covered by the run/walk algorithm - 1:07:08 running, 0 seconds walking and 4 seconds idle - is equal to the activity time (1:07:12), which is ofc absolutely expected.

    As you say, paused time is excluding from moving time calculations. Well, paused time is also excluded from run/walk calculations, so again it cannot explain an inconsistency between the two sets of calculations.

    To be clear, the perceived inconsistency is that:

    - According to the moving time calculations: OP spent 1:07:00 moving and 12 seconds not moving

    - According to run/walk (and idle) calculations: OP spent 1:07:08 running, 0 seconds walking and 4 seconds inactive

    I think in a perfect world, the "not moving time" would be equal to the "inactive time". Or at the very least (assuming different pace thresholds for the two sets of calculations) "not moving time" should be equal to "inactive time" plus "walking time."

    And again, the activity time is equal to the elapsed time (both are 1:07:12) which means that even if pauses could be a factor in this sort of discrepancy (I don't believe they can), OP did not pause their activity. If the activity had been paused, I would expect elapsed time to be greater than activity time.

    I think you are approaching this with the assumption/understanding that OP paused their activity and they expect the paused time to be included in the idle time (under the run/walk info), but that is not the discrepancy they are referring to.

    What they are expecting is their moving time to line up with their running time (under run/walk).

  • Thank you for the reply. What's the difference between "not moving" and "inactive"?

    There's two separate calculations/features/algorithms here:

    - Moving time (which is a feature/calculation that has existed on Garmin watches long before run/walk)

    - Run/walk (and idle) time, which is a newer feature

    When I say "not moving" (time), I mean the periods of the time that the moving time feature/algorithm determined you were not moving. Like I said, the Strava moving time algorithm for running decides you are moving when your pace is faster than 30:00/mile. My guess is that Garmin also uses a pace threshold for moving.

    When I said "inactive", I should've said "idle". I mean the periods of time that the run/walk feature/algorithm decides you aren't running or walking. According to Garmin, the run/walk feature uses both pace and cadence to categorize your movement into running, walking and idle time.

    Garmin doesn't really go into details about how either feature works.

    I agree that what you're seeing (the 8-second discrepancy) doesn't make much sense from a practical POV. But given that there are 2 separate features and we don't know exactly how they work, it seems inevitable that their results won't match exactly.

    To be honest, I never expect moving time to line up with run/walk/idle time. (Yes, in a perfect world all the numbers would line up.)

    IOW these calculations don't necessarily reflect the "reality" of your movement, just 2 different (and imperfect) methods of categorizing it. And I don't think the methods "talk to each other" - e.g. there's no attempt by Garmin Connect to reconcile the numbers from the two methods so everything lines up.

  • this doesn't explain why I have a moving time inferior compared to the run time

    The explanation is in another document: What Is the Run/Walk Graph in My Run Activity Details? | Garmin Customer Support :

    The Run/Walk graph is a representation of the watch detecting you running, walking, or idle during a Run profile activity based on your steps per minute (spm) and your pace. Below is a definition of the requirements for each representation:

    • Idle (inattività): shows you had 0 spm AND had a pace of slower than 0.03m/s (meters per second) for more than 6 seconds.
    • Run (corso): shows you had greater than 140 spm and/or a pace of 12 min/mi  or faster for more than 6 seconds.

    So during those 6 seconds, before the Idle (Tempo di inattività) timer starts incrementing, it is the Run timer (Tempo di corsa) that continues to increment, although you no more move, and the Moving Time (Tempo in movimento) does not increment either

  • The explanation is in another document

    Great find! It's nice to see a detailed explanation of how it works.

    Just to complete the explanation:

    Walk: shows you had less than 140 spm AND a pace of slower than 12min/mi (Minutes per mil) for more than 6 seconds.

    imo the "for more than 6 seconds" qualifier is open to interpretation [*], but I don't think it really matters. The point is that the run/walk/idle time calculation is clearly different than the moving time calculation (which is likely based on a simple pace threshold, as with Strava).

    [*] i.e. just because you need to be idle for more than 6 seconds in order to *trigger* idle detection, does not mean that the 6 first seconds is necessarily excluded from the idle time. It would be easy for that 6 seconds to be retroactively subtracted from the run timer and added to the idle timer.

    As a bit of a tangent, I will also point out that aside from being two different algorithms, moving time and run/walk/idle time calculations are performed in different places:

    - moving time is apparently calculated in Connect, after the activity is synced. At least, that's what I always inferred because: it's not available in the activity summary on the watch, and it's not saved to the FIT file

    - run/walk/idle time is calculated on the watch itself, as it's available in the activity summary on the watch

    This may be relevant because the data that's available to Connect may not be as detailed (high-resolution) as the data available to the watch. This is because, by default, Garmin watches record FIT files in "smart recording mode" (which means they only save data points when they deem it to be necessary, so there could be several seconds between two consecutive data points, if you're standing still.). I personally prefer to use "per second recording mode", as this avoids any potential data loss associated with smart recording..

    You can enable per-second recording as follows: hold UP for the menu > System > Data Recording > Frequency > Every Second

    I doubt it will make your moving time line up any closer with run/walk/idle time, but it doesn't hurt to enable this setting.

  • Great find! It's nice to see a detailed explanation of how it works.

    Just to complete the explanation:

    Walk: shows you had less than 140 spm AND a pace of slower than 12min/mi (Minutes per mil) for more than 6 seconds.

    imo the "for more than 6 seconds" qualifier is open to interpretation [*], but I don't think it really matters. The point is that the run/walk/idle time calculation is clearly different than the moving time calculation (which is likely based on a simple pace threshold, as with Strava).

    [*] i.e. just because you need to be idle for more than 6 seconds in order to *trigger* idle detection, does not mean that the 6 first seconds is necessarily excluded from the idle time. It would be easy for that 6 seconds to be retroactively subtracted from the run timer and added to the idle timer.

    As a bit of a tangent, I will also point out that aside from being two different algorithms, moving time and run/walk/idle time calculations are performed in different places:

    - moving time is apparently calculated in Connect, after the activity is synced. At least, that's what I always inferred because: it's not available in the activity summary on the watch, and it's not saved to the FIT file

    - run/walk/idle time is calculated on the watch itself, as it's available in the activity summary on the watch

    This may be relevant because the data that's available to Connect may not be as detailed (high-resolution) as the data available to the watch. This is because, by default, Garmin watches record FIT files in "smart recording mode" (which means they only save data points when they deem it to be necessary, so there could be several seconds between two consecutive data points, if you're standing still.). I personally prefer to use "per second recording mode", as this avoids any potential data loss associated with smart recording..

    You can enable per-second recording as follows: hold UP for the menu > System > Data Recording > Frequency > Every Second

    I doubt it will make your moving time line up any closer with run/walk/idle time, but it doesn't hurt to enable this setting

    Thank you all for trying to explain this, I've learned new things. At the moment I think that the stats about the time in Connect are more precise compared to the run/walk ones. At least this is what I see, but anyway on my Fenix 8 I can't find that Data Recording settings.

  • Data Recording : Frequency, can be found via the following…
    Hold middle button > Watch Settings > System > Advanced > Data Recording > Frequency > Smart | Every Second

  • Data Recording : Frequency, can be found via the following…
    Hold middle button > Watch Settings > System > Advanced > Data Recording > Frequency > Smart | Every Second

    I've found it, thanks. 

  • This may be relevant because the data that's available to Connect may not be as detailed (high-resolution) as the data available to the watch. This is because, by default, Garmin watches record FIT files in "smart recording mode" (which means they only save data points when they deem it to be necessary, so there could be several seconds between two consecutive data points, if you're standing still.). I personally prefer to use "per second recording mode", as this avoids any potential data loss associated with smart recording.

    But the manual says that the per second recording mode could not register longer activities correctly.

  • But the manual says that the per second recording mode could not register longer activities correctly.

    As I said, I think this is a mostly outdated concern. It may have applied to Garmin watches from 10+ years ago which had a tiny amount of storage, but modern Garmins should have more than enough storage for lots of large activities, especially if they support maps and or music.

    My FR955 (which supports maps and music) has 32 GB of storage. A 30-minute running activity recorded with per-second recording takes about 230 KB. (This includes data from an external Stryd sensor and HRV logging.) So let’s say per-second recording takes about 560 KB per hour for me. I could record 1700 hours of activities and it would use slightly under 1 GB of storage. Even with all my maps and music, my FR955 still has about 13 GB free, so I could theoretically record over about 20,000 hours of activities without ever syncing.

    I did notice that newer watches seem to only keep so many months of activities, as opposed to saving all of them indefinitely, but I doubt they would start auto-deleting activities that had not yet been synced.

    The related garmin support article has this to say:

    https://support.garmin.com/en-CA/?faq=s4w6kZmbmK0P6l20SgpW28

    Why Should I Choose One Option Over the Other?

    When choosing which option to set your watch to, you will have a few things to consider:

    • Battery Life

    • File Size

    • GPS Accuracy

    The more data points your watch records, the smoother and more accurate your track will be. The trade off for this is that it will consume more battery life, as well as more space in the device's memory. If you need a longer battery life or need to make sure you have more space for a longer activity, you will want to choose Smart Recording. If you want more accuracy and a track that follows your movement based on points that are recorded more often, you will want to choose Every Second Recording. 

    Yes, it does mention the file size concerns, and it mentions battery life concerns. But it also mentions a more accurate GPS track. It might be *possible* that this would affect moving time calculations in Connect, but I doubt the difference would be dramatic. And I doubt it would guarantee that moving time would be equivalent to run/walk/idle time, since those are 2 different algorithms. 

    Personally I would just enable it for the more accurate track.

  • At the moment I think that the stats about the time in Connect are more precise compared to the run/walk one

    I’m not sure about that, but it’s hard to say.

    It seems that the run/walk/idle algorithm is more complicated than the moving time algorithm, for sure. Run/walk/idle looks at both pace and cadence, but it does have a 6 second threshold for detecting a new “phase” as mentioned above. I think moving time just looks at pace. (e.g. the way Strava does it for running, it decides you’re moving if you go faster than 30:00 per mile.)

    As I said, I never expected the numbers from the two algorithms to line up (although I can understand why it could be confusing that they don’t).

    If you think that moving time is already more precise than run/walk/idle time and you’re looking to “improve” the run/walk/idle time, enabling per-second recording won’t help, since run/walk/idle time is calculated on the watch itself.