Forerunner 955 GPS Problem

You are still missing the main point. You cannot compare two different routes with two different settings and claim that one are better than the other and you cannot either compare it only by looking at garmin connect.

For comparisations I would suggest that you use DCRain Makers anlyzer tool at https://www.dcrainmaker.com/analyzer

Also notice there are differences in the "quality" of the satellite signal depending on where in the world you are located. If you really want to do a compare of the two, then you have to run the same route more with the two different settings.

Especially on "short Curves" the distance will be shorter than one sec recordings as smart-recordings typically are 3-7 seconds between ever recording. You could probably see a difference it your take a run at your local running track with the two settings.

True, the main difference of gps trackings are easy to see on a straight line running. 

To really see the difference - Go for a bike-trip of 13-16 miles (twice)  and compare the routes in DCR Analyzer or just take a look at garmin connect. With 100% certainty the smart recording route is shorter than the 1 sec. route.

Well I gave up riding a long time ago, so I’ll take your word for it as far as cycling is concerned. My point was that, for trail running at least, Smart Recording makes little to no difference in total distance. 

I have a local trail route I use for testing different watches and settings. I’ve tested Smart vs Every Second multiple times, running laps forwards and backwards, comparing lap distances and have never seen a measurable difference between the two. Further, when comparing GPS tracks, Smart Recording has nearly the same number of points as every second. This can be checked using services that give a point count for a given GPX route.

Anyway, just my experience. Like I said, I still use every second 

i too record every second "just because." out of curiosity, i went to my watch and see that with it's internal memory, i have runs dating back to November. very roughly, that's means i have room to store 5 months of runs. i typically run about 5 hours a week (although i have been injured so the last few weeks are closer to 3.5-4 hrs per dr's orders)... so there is still a lot of space to store 1 second data. yet, if one were doing something like the PCT or App Trail and not having access to internet, it might be wiser to do smart recording to make sure everything gets stored internally until one has access to the internet.

Hi everyone, I finished with 1:46, the GPS was very consistent, but it was 500 meters off in 1 place and the average was 4:55 pace instead of 5:02 pace

GPS doesn't work when there's a concrete car park over the top of you!

goo.gl/.../XMJ5YsyYAfLbxKgf6

How did the employees work, for example, my wife's watch is forerunner 245, my friend's watch is fenix 5x+ and others, I think this is what he is talking aboutUltraKapitanevery second can be an important option to prevent such situations

In this sort of situation, one of two things can happen. Either the GPS gets no signal at all and stops recording, producing the straight line as seen on the left track on your picture, or it picks up reflections and maybe two or three close-to-the-horizon satellites and gets confused about its location, as in the right hand track on your picture.

The watch is not at fault here. It's trying its best. GPS does not work through a metre of concrete.

The answer to this error - depends of set og different reasons.

If the setting is "smart recording" and the runner is turning slightly left or right in less than a second afterwards then the prediction is that the user is staying in that direction until the gps-tracking is back. When that happens the gps will simply make a sharp turn just like the above example. With-out the actual file this is impossible to verify if it is the actual reason.

If the setting is "smart recording" and one gps-dot is wrong when going "offine" due to a tunnel/bridge then it cannot ignore that gps-dot and it will use that dot as a true positition and not a faulty gps-tracking-dot. (every GPS-company tries to do gps-smoothing to avoid these faulty dots.

If the setting is "every second". then the tracking should be changed because when the next gps-position is located it will draw a straight line, just like the one above, BUT it should ignore the some of the dots because the become invalid due to no gps-tracking. As a result the the deviation from the original route should be minimized.

Unfortunately garmin removed the gps-smoothing part.. (alignment to roads in connect). So in order to fix it, the user  has to remove the faulty gps-positions - that can either be done in Garmin BaseCamp or an external gps-tool.

The interesting thing is that the line on the left side is at the same crossing point and the GPS drew a straight line there, but on the way back it faltered when passing through the same place.