Instinct mileage always short of actual mileage

Whenever I do a group run with people that have a different type of Garmin from my Instinct or they have the Apple Watch my mileage always comes up .25 - .35 mi. shorter. The same thing happens in road races  that are wheel measured and certified courses. 

  • At what distance?  There will always be a degree of accuracy.  A 1/4 mile discrepancy on a two mile track is terrible, but very reasonable for a marathon. 

  • It happens at all distances, training runs and races. My watch is always off from everyone’s so when the group finishes I always keep going till my watch hits whatever mile we have set to run, for example we did a 10 miler they stopped at 10 but my watch only said 9.65 so I kept going to hit the 10. 

  • There are many post in these forums where people have the same issue. Many believe it's some sort of filtering Garmin is now using, but I don't know if that's actually the case. There is a solution if the exact distance is a deal breaker for you, but it comes at a cost.  A Styrd footpod.  I've been using one for a couple of years and nothing comes closer to recording distance and pace as a Stryd. It will also record beneficial metrics similar to the HRM-Tri which I believe doesn't record running dynamics on the Instinct. 

  • There's a misconception between smart and every second recording.  If you want a clearer plot on your recorded track, every second recording is best, but it doesn't change distance. See the link below:

    support.garmin.com/.../

  • but it doesn't change distance

    Unfortunatelly, that's just the theoretical assumption. In reality, the "smart" recording is much more cloggy and skippy, cutting off corners, turns, and other abrupt changes, because when the watch detects the change in accelartion/elevation/HR/... when it then turns on the GPS and acquires the position, valueable time passes away, and distance data is lost. I compared many activities with smart and 1s recording, and the differencies are quite obvious.

  • ... there is another issue with the distance calculation. It is general for all Garmin devices (or for most of them), but the effect is more apparent at Instinct, which clearly does not use the most performant GPS hardware or design. The problem is that, when there is a drop of GPS signal, such as in underpasses, or near walls, buildings, or under dense vegetation, the watch either completely excludes the distance made during the period without the signal, or it just uses the average stride length to calculate your pace. If you are running or cycling, the distance will be always underestimated.

    For some uncomprehensible reason, Garmin does not interpolate the distance from the points before and after the signal drop, and also it does not calculate the average stride length during that given activity (although it could easily do it), but uses either the default hardwired value, or the value in your settings.

    You will almost always get a better estimate of the distance when you use the distance correction option of Garmin activities in Strava, because unlike Garmin, it does interpolate the missing distance from the available GPS points. Well, there are situations, when it would lead to wrong results too (i.e quite long drops of signal with a complicate path lost), but those are rather exeptional cases. In very most cases the corrected distance is much more accurate, than the one reported by Garmin.

  • "smart" recording is much more cloggy and skippy, cutting off corners, turns, and other abrupt changes, because when the watch detects the change in accelartion/elevation/HR/... when it then turns on the GPS and acquires the position, valueable time passes away, and distance data is lost.

    It does look more "cloggy and skippy" in the recorded map, but in the recorded map only.  GPS is not turned off and back on again at any point of the recording, so there is no loss of valuable time reacquiring it's position resulting in distance data being lost.

    I compared many activities with smart and 1s recording, and the differencies are quite obvious.

    I too have compared many activities using both smart and 1 sec recording at the same time using a 945 on one wrist and a 935 on the other.  My findings are completely different than what you said you found to be quite obvious.  The 1 sec recording shows a track with clean crisp distinct lines which are very apparent in 90 degree turns while the smart recording looks sloppy.  But, pace and distance were equal and and confirms what Garmin has posted.  Both are recording the exact same data, the only difference is smart recording eliminates irrelevant data points to make the file smaller.  I don't know how you claim that it's just a theoretical theory when it's something that can be simply tested and proven.

  • Both are recording the exact same data, the only difference is smart recording eliminates irrelevant data points to make the file smaller.

    This is your interpretation. What Garmin writes is "Smart Recording does not affect the data being recorded such as pace and distance or heart rate" - and that is in fact just the wishfull thinking, that recording only at the change of direction/speed/... won't have any impact on the data. The reality differs.

    And even Garmin self admits it in another document:

    Settings that Improve the Distance and Speed Accuracy 

    "Every Second recording will produce highly detailed tracks of your activities. A more detailed track will lead to more accurate speed and distance data."

  • GPS is not turned off and back on again at any point of the recordin

    This is also not true. The GPS chip is being powered on and off all the time, and that's exactly why Garmin recommends using the Smart Recording for longer battery life. At the Ultra Trac mode the GPS position is being queried only once a minute, so the battery saving is much bigger, but the accuracy renders the mode practically unusable.

    Tips to Extend the Battery Life on a Garmin Fitness Watch