Is it accurate?

What appears to be the case is that while the GPS in these units "has" WAAS corrections they're ineffective under moving-target use, where the chipset in the 910XT and 310XT did not have that limitation.

The "advertised" GPS accuracy (SA Off, which it has been since the end of the Clinton timeframe) is ~15 meters with a 95% confidence level. With WAAS it's 3 meters with the same confidence level. 3 meters is good enough to land an aircraft (15 meters is NOT!) and it's also good enough to nail, repeatedly, the side of the road you're running on (or whether you're on the sidewalk.)

This assumes you have an actual reasonably-clear view of the sky; multipath will still get you (e.g. running in a downtown high-rise area where reflected signals and attenuation simply cannot be overcome) of course.

The improvements Garmin has made of late to the software are apparent; if you go into mapping mode and walk out to the center of the road in front of your house, zoomed in, you'll see that when at rest it's now getting right in that 3m range. In other words, the corrections are working. But as soon as you start to run that accuracy degrades -- and looks to be in the 15m range, which is base GPS accuracy.
In other words, WAAS augmentation under that use case is ineffective.

I have no idea if this is a function of the filtering in the chipset and thus can't be overcome with firmware on the watch or whether it's something that Garmin can address. I assume it's the former because I find it hard to believe Garmin wouldn't have fixed it by now in the F3 series (since it was easily repeatable and people were complaining about it) if it was something they could take care of in firmware. Thus, the assumption I think we have to make is that it's a function of the GPS chipset and has something to do with how it uses (or doesn't) WAAS corrections when the unit is moving (and, presumably, moving on your wrist as well.)

Look at this track: https://connect.garmin.com/modern/activity/1685556681

It's consistently on the wrong side of the road on the north end of it, and the "wrong" swaps sides when I come back down Bay. No, I was not running with traffic (wrong way) in BOTH segments; the track error is fairly consistent in the 10-15m range. That strongly implies that WAAS is ineffective.

There *are* tracks I have from the 910XT with the same sort of error, by the way. But there are also tracks where it's bang-on to the point that it brings me right down the middle of a sidewalk I was running on. I've never seen that accuracy from the F3 or F5x and it wasn't *all* the time with the 910XT, but it was frequent enough for me to immediately notice the difference when I upgraded.

YMMV and all, and it's "good enough" for what I use it for. Yes, I'd like my 910XT style performance back (although it wasn't *all* the time better) but it may just be an unrealistic ask for a wrist-mounted device with a small antenna in the bezel and serious power-consumption, pre-amp and antenna-size constraints.

All in I'm happy enough and the distance recorded (and thus pace) are, in my uses, extremely close. That I can quantify as I've run a couple of professionally timed and wheeled-distance measured races since getting the F5x and those have all been within spitting distance (as was the F3 through my time of owning it.)

I have to say I am now overall happy with the watch. Do get the odd set of data - last night it recorded a max speed of 131mph on my mountain bike! Seen those weird errors on every device I have own. I like that I could go edit and cap it correctly in the software though. That has only happened twice though. Had a couple of bad GPS maps where the route was all over the place - again seen that on all devices. Had some where it has been spot on. Being able to ride several times a week without having to charge after every ride, not having my rides capped by a battery is a big bonus for me.

I'm going with the thoughts that its a general sports watch that'll record data reasonably accurately for whatever I want to do. If I wanted to nail the data perfectly in an attempt to use it to improve certain sports metrics, with a much higher level of confidence, I would have to add on devices that are specific to what I want to achieve. Such as a HR strap for my accurate HR, etc.

Apologies if my first comment came across as serious. My intention is to poke fun at those individuals who do just what you discuss in terms of pixel peeping. I agreed, your track looks good as it is displayed, just like the majority of mine have done over the years. But people will insist on zooming in to unrealistic levels to illustrate their point. Here's some more fuel for their fire from your same track. My god that overshoot on the corner is monstrously massive!

Totally got the joke, don't worry :)

Btw that isn't an overshoot but a navigational error :D

" />">



" />">

Can't see any watch picking up WAAS other than the the most perfect conditions

" />">



" />">

Can't see any watch picking up WAAS other than the the most perfect conditions

The first Fenix and the F2 has WAAS and you could actually see it working and turn it on or off. Now the F3 claimed it has it also, but without anyway to verify it's existence. Just by that fact alone, I'm saying it never had it and was just a marketing sell point. My F1 is always gives a better distance reading than my F3, and I have had 2 different ones.


Check out the pictures here https://forums.garmin.com/showthread.php?374137-Is-it-accurate&p=972352#post972352

Raced a 50 miler this weekend. Course appears to be right at 49.1 miles looking at most people's data. Fenix 5x recorded 48.3 miles. I ran with Act Tracking ON, GLONASS On, 1-sec recording and navigation of the route active.... Had my Vivoactive HR on and it recorded 49.2 miles, no GLONASS, no HR, smart recording. It had 30% battery left... 5x had 39% battery left. 10hrs 30 min total....5x track wasn't to great, vivoactive HR was better! Still love features of 5x but I think it's algorhythms need adjusting a bit. Definitely under reads distance.

5X data...

https://connect.garmin.com/modern/activity/1692897223

The first Fenix and the F2 has WAAS and you could actually see it working and turn it on or off. Now the F3 claimed it has it also, but without anyway to verify it's existence. Just by that fact alone, I'm saying it never had it and was just a marketing sell point. My F1 is always gives a better distance reading than my F3, and I have had 2 different ones.


Check out the pictures here https://forums.garmin.com/showthread.php?374137-Is-it-accurate&p=972352#post972352

675 euros spent, and this is the result...

That's quite unfortunate.

Maybe one-day, hopefully with the F6 series, they actually produce a watch that is accurate. Especially when they are charging outrageous prices for these high tech "toys".

Maybe I'm just lucky, but I think the tracks and pace that I've gotten from the 5X have been much improved over the last few weeks. I've been directly comparing it to the 935 on my last several runs. While the 935 is a little better, they are actually pretty close.