Instinct 2x inaccurate swimming distances

My 2x almost doubles my open water distance.  Never happened on original much older Instinct. Must be new algorithms that calculate distance when under water? I know about breaststroke limitations but I when I snorkel, or swim leisurely I would simply extend my wrist out of the water regularly until a signal is re-obtained and the points are connected to estimate reasonably accurate distance.  Now  my new 2x watch keeps calculating when under water or staying around one spot. 
My settings: all plus multi band, measure every second, I add1 minute after gps signal is ready to ensure greater connections and have the latest updates (14.26). Nothing helps. Seems it would be wise for Garmin to have two sub categories in swimming: competitive or training (front crawl…) and casual or leisure/snorkeling that doesn’t keep adding distance based on programming assumptions of continuous forward motions.  Pool swims are also inaccurate compared to my spouse who uses an Apple Watch.  

  • Yes, the new algorithm is flawed. It uses accelerometer data to estimate the distance from strokes. If you do not swim in regular pace, or the pace is not yet sufficiently calibrated from previous swims, then the estimated distance can be indeed quite off.

    To work around it, I simply use Strava to correct the distance, then edit the activity in Garmin Connect, and enter the distance from Strava. Strava has a function "Correct Distance" that recalculates the distance from the GPS track (which is mostly rather correct in my case), so the corrected distance is much more accurate than the estimated one.

  • My 2x almost doubles my open water distance

    I swim several times a week with my 2X and have no issues with accuracy when compared with other watches worn at the same time. I swim a course with well known distances and distance markers. Allowing for tide a 2500m swim is recorded as a 2500m swim +/- 100m and usually within 10s of metres of a comparison watch such as an EP47 or F7.

    i do not believe the new algorithm to be flawed if swimmers follow the Garmin guidelines, at least as far as my freestyle stroke is concerned. Average pace is around 2:30 to 2:45/ 100 so nothing flash there. 

    However, in this case I think the problem is more related to this

    know about breaststroke limitations but I when I snorkel, or swim leisurely I would simply extend my wrist out of the water regularly until a signal is re-obtained and the points are connected to estimate reasonably accurate distance.

    The algorithm has not been designed for swimming like this. If you want accuracy in this case perhaps consider towing your watch in a float bag behind you on the surface. 

  • Yes, it works reasonably well, as long as you swim in quiet waters without currents, without waves, always in consistent way and using the same stroke, and the same cadence. In the moment the weather is a bit rough, it counts waves as strokes, adding the distance even when you float on the same spot without moving. As long as there is a current, the covered distance does not match either. If you use fins, pads, or change the technique or stroke, the distance gets skewed as well. Personally, I do not mind much, since I mostly swim on the same spot, so I know well the covered distance. And even when I happen to swim elsewhere, fixing the distance with Strava does the trick well.

  • the moment the weather is a bit rough, it counts waves as strokes

    Take it from me it does nothing of the sort. Yesterday’s swim for instance was as rough as it gets, large swell, white caps etc, 2:30/100 one way, 3:45 the other way. It was a shortened 1000m swim and was recorded as such. The swimming distance is not affected by current, waves etc. I swim in the open sea affected by tide and weather several times each week. On occasion I’ll go inland to a fresh water lake and still get the distance recorded accurately. 
    https://www.strava.com/activities/11031520307 
    Easy to see the pace change due to current and wind over tide just from the graph  

    adding the distance even when you float on the same spot without moving

    That’s a function of GPS in or out of the water. Largely due to satellite precession but then, in the water even if you are floating there is still some movement.

  • That’s a function of GPS in or out of the water. Largely due to satellite precession but then, in the water even if you are floating there is still some movement.

    No, not a problem of GPS drift. The GPS track is just fine, and when the distance is recalculated using the GPS coordinates, it is more accurate. And if you do not believe me that the accelerometer is now used to adjust the GPS distance, you can verify it with the following test:

    Disable GPS in the OW Swim activity profile, and start a dummy OW Swim activity while sitting and beating with the fist on the desk - after some 20 hits, the watch starts counting the distance, despite no GPS drift (since GPS is off).

    The same happens in the waves, or for example when you swim really slow, while keeping fast cadence. Happens frequently to me, when I accompany my wife who is not a good swimmer and swims slow - the watch records 300m, despite that we did barely 50m, just because I do not swim in the usual way. Similarly the use of equipment, or any drills like legs-only, arms-only, single arm swims, or long glides throw the distance measurement completely off. The GPS track is still fine, and the distance after correcting at Strava is accurate.

  • It's quite simple really. Follow Garmin's guidelines for swimming in open water and the distance recorded is accurate irrespective of currents, waves, etc. https://support.garmin.com/en-NZ/?faq=AnrKqtOX8FAzmF9yXaIXe6

    Do otherwise and accuracy decreases.

  • I do know very well how it works, and I do know how to get the right distance even if I do not follow the guidelines (which often I indeed cannot and do not want to follow). I am just explaining others how to work around these limitations, and how to get the right distance when the watch fails doing it.

  • Accepted.

    However, this suggests otherwise and could be viewed as misleading. it simply is not the case that

    as long as you swim in quiet waters without currents, without waves, always in consistent way and using the same stroke, and the same cadence. In the moment the weather is a bit rough, it counts waves as strokes

    The only thing close to accurate in that statement is the need to swim with the same stroke. And even then, that stroke needs to be in accordance with Garmin's guidelines, ie a stroke where the arm regularly comes out of the water such as freestyle.

  • We speak about two different things. While you speak about the accuracy of the GPS track, I speak about the distance accuracy influenced by the accelerometer. You can have two practically identical GPS tracks of excellent quality, and still get dramatically different distance from them, when the accelerometer data differs. It did not used to be the case, previously at swimming, although similar method is used at running already since a longer time. Not sure when exactly it changed, but probably already more than a year ago, if I remember well.