Instinct 2x inaccurate swimming distances

trux said:

We speak about two different things

No we do not. The distance measured during swimming is, and always has been since the 310XT (the first watch I had that could be used to record open water swimming), a combination of GPS and accelerometer. Nothing's changed in that regard except that distance recorded and track displayed has got better and better with each iteration. This is not affected by waves or current. The biggest factor affecting the distance is the swimming style. Swim according to Garmin's guidelines and all will generally be well. Do anything other and accuracy of distance and/or track will be compromised.

Yes, the accelerometer was always used, but the algorithm changed. And not only at swimming. The accelerometer is being used much more aggressively than it used to do. I know it very well, since I analyzed it many times, using also own CIQ tools to dig the internal data, and to compare diverse setups. Once I have a bit more time for it, I will compile a more detail report and post it.

trux said:

the accelerometer was always used, but the algorithm changed.

I should hope so too. But there's no surprises there since Garmin (despite the nay sayers) is always looking to improve recorded data. I've seen great improvements over the years with different watches in terms of distance recorded and the quality of the tracks displayed.

Here's the GPX upload compared to FIT file for that rough as boots swim the other day. For me the most interesting thing here is that Strava increased the distance by a 23m but shortened the moving time. The elapsed time remains the same.

The differences are minor and have been minor irrespective of the distance covered.

I've done this comparison on a few occasions when people comment about the accuracy of Garmin's swim algorithm. For me it works. And I cannot believe I am the only person swimming with a Garmin watch that does not have any complaints about the accuracy.

Does not seem to be that rough sea, if there are only 2 minutes of difference of time.

I don't care what you think. You weren't in the water.

I've had a suspicion for some time that my new Instinct is inaccurate.  It is.  Today, I just swam (open water) with Instinct 2 on 1 wrist and 745 on the other.  Instinct cuts each stroke length by about 8% such that a 2km swim (which I have done dozens of times before so I know it is 2kms), recorded as 2kms on the 745 and 1839m on the Instinct.  My average pace was therefore slower on the Instinct.  Only 1 of me got out of the water at exactly the same time of course!!  Personally, as a regular swimmer this is irritating.  I love my 745 but the battery life is now rubbish so I needed to replace.  I wont be recommending an Instinct for swimmers.

I wish I had seen this before buying the Instinct (2x solar). I did the exact same test, with my old Fenix 3 on one wrist and Instinct on the other. On a known course of 500m I was 2:05 per 100m on the Instinct and 1:46 per 100m on the Fenix. Very disappointing to not be able to compare new to old times.

I believe that the more u use the watch the more it calibrates accurately, as for me, comparing to my old FR 245, I feel that my 2X is more believable in terms of distance and pace. 

Thanks for your comment. Someone else said this to me. I'd hope it would be true, but I have swum a lot so far with little change. I also downloaded a third party swimming app by f3b that recorded my swim much more accurately, so that might be my workaround for now.

Swim this morning with instinct 2 X Solar and Epix Pro 47. Water was 12C with slight ripples. As you can see, very little difference between the two. This is not unusual for my swims and I'm swimming several times a week in the sea.