Shift in GPS track depending on wrist

Which satellite systems do you use? For me GPS + Galileo is very poor. GPS only and GPS + GLONASS are much better and very similar in results to each other. I choose GPS + GLONASS in the hope for better performance in more challenging terrain when GPS might be blocked from view.

The default GPS + GLONASS as that still seems to be recommended over Galileo.

I have the same exact issue.

The following two activities were from two consecutive days, I ran on the same path in a park with just a few trees, one day in one direction, the other day in the exact opposite direction.

https://connect.garmin.com/modern/activity/4776906281

https://connect.garmin.com/modern/activity/4781521547

Also, these are other activities, all of them are done on the same path(different path from the previous one), in the park mentioned above. I've tried all sorts of combinations(GPS, GPS + GLONASS, GPS + GALLILEO, Right Wrist, Left Wrist).

https://connect.garmin.com/modern/activity/4731695692 (Watch worn on the outside of the loop)

https://connect.garmin.com/modern/activity/4727952298 (Watch worn on the outside of the loop)

https://connect.garmin.com/modern/activity/4725371491 (Watch worn on the inside of the loop)

https://connect.garmin.com/modern/activity/4714617856 (Watch worn on the inside of the loop)

It always seems to have the same outcome: there's a shift towards the side on which the GPS is worn on. I think the shift is upwards, because the watch is always sideways when I run.

My 5 year old Fenix 3 and my 6 year old Forerunner 220 both do a better job from an accuracy POV.

I must admit I was sceptical but decided to give it a try this afternoon. I went to a usual route: 2 km with trees and some buildings followed by 5 km next to a river and back. Usually first (and last) 2 km are wobbly and shifted to the left from 5-30 m so when I look at the tracks (back and forth) they part for up to 50m. Other 5km are never a problem with nothing in site to block the signal, this part is always excellent. 

This time I rotated the watch to the inside of the wrist (left wrist both times) for the last 2 km and a mirracle happend! It's SPOT ON even next to a buildings and under trees it's clean, straight and spot on!!!

How is this possible? Position of the watch in space is the same, only antenna position changes?

I need to confirm this on a couple more runs, but if it proves correct:

1. my Fenix IS capable of recording excellent track even next to tall buildings and under trees- that's a huge releif because I love the watch otherwise

2. all those excuses about metal bezel blocking signal and hand moving making signal recording difficult are jibber jabber

3. all we need now is Garmin team to fix this :)

One more challenging run with the F6 on my inner left wrist. Running a little 400m loop around a building, anti clockwise (so the building in on the side of the watch). Some little sharp turns. So it's clearly not that easy for a GPS recording. A Suunto 9 on my right outer wrist.

I have done this loop before, clock wise, anti clock wise, left and right wrist with the F6. Today was clearly the best recording, GPS wise. I was even skeptical looking at the distance on the watch ! I need to do that a bit more to conclude, but it's not just the satellites today :

• Garmin distance, 6.64km
• Pure GPS distance : 6.70km
• Suunto distance : 5.78km (!!!)
• Real distance : around 6.8km

I didn't have my footpod today, but it measures this loop around 400m. 

I need to repeat this test to confirm, but the difference with the Suunto is significant. I had days with cleaner GPS recordings, but then the S9 and the F6 both have better results. 

I looked at my posture, and the watch is vertical, facing my body, quite close to it. I would say 4 inches maybe ? So in theory, it's bad, my body is blocking the signal significantly. But could a GPS expert share some thoughts? could it be that not facing the outside, the watch doesn't receive a lot of rebounds from the building, and only what is rebounding on my own body ? Rebounds of GPS signal against buildings (or any structure) are usually one of the reasons why it sucks when running in a city or a canyon, am I correct ? Maybe this position helps to receive significantly less of them, making the result cleaner ? 

Well, it's just a few recordings we have done. I will record more and keep comparing with the other watch (same chipset after all !).

I think before complaining your small wrist based device (of any brand) that is perpendicular to the sky most of the time is showing you are on the 'wrong' side of the pavement..... it is triangulating with many things twelve and a half thousand miles away to try to determine where you are

I wouldn't buy a Tesla but that is a different conversation With regards the accuracy I wonder what has fundamentally changed in the last five years then. Are they using different materials to the Fenix 3 ? I can't imagine the areal is much different . Is there something about the new CubeSat's and the way they perform in space compared to more traditional ones. I know they have now changed to Sony chips in the watch for power benefits and maybe there is a trade off because of it. Perhaps the Fenix 3 was a traditional bulb and the 6 is more of an energy saver 

The Sony chipset is clearly giving some headaches compared to the best chipsets we were using before (like the SirfStar). What is the technical difference explaining it, I don't know, but it's quite obvious that it's a bit less accurate. Maybe since it's using much less energy, it's less sensitive ? i don't know, I'm not an electronic expert, nor a GPS expert. Materials could be a difference, but even the plastic watches aren't really better with this chipset. It's still very acceptable. in fact, to me, the biggest issue is rather Garmin tweaking the distance, rather than the GPS accuracy. If I had the choice to fix ONE thing, I would surely ask for an option to deactivate distance adjustment. The GPS is good enough for me. And in a couple of year, maybe maybe maybe dual frequency will give the watch a terrific accuracy. And we will still come here to complain about something else ;)

PS : woudn't buy a Tesla myself either, but you see the idea ;)