What device is more accurate for distance travelled (Wahoo / Garmin / Samsung)?

[Two-Wheels]

Now the thread will probably end up as a "what device to you choose to use" I imagine, but does anyone reading this actually have something to validate an answer to the thread title question?

Now I'm sure there's more to it than company name vs company name. There's probably a range in and amongst Garmin itself.

I went on a ride today - Garmin on the left wrist, Samsung on the right wrist, Wahoo up front on the handlebars.

Once i finished, there was 1.7 mile between the lowest & highest & when you're looking to hit a set number & your legs are feeling like lead by the end, you just want it done. Now of course I could've gone out with only one device & I wouldn't have known any different, or I could just pick one to go with & that's 'The One'. But anyway...

Wahoo Roam: released May 2019 I believe
Garmin Forerunner 735XT: May 2016 apparently, so a bit old now & maybe inferior tech?
Samsung Galaxy Watch 3: August 2020.

 

[ColinJ]

I plot my new routes on OS digital maps before riding so I know how long they are before I ride them. Over many years with a Garmin Etrex, Garmin 500, and a Garmin 200, I have never found the distances measured by any of the devices to be far off what I expected.

 

[cougie uk]

How was the Samsung? I'd have thought the others would be more accurate?

 

[marzjennings]

Generally they all use similar gps chip technology so it'll be down to how many points each device captures and how they've coded interpolation between points. If you can download the gpx file from each device you could literally count how many points each device captured and the lowest count would indicate the lowest accuracy. But it'll tough to know for sure unless you know the actual distance traveled. We use Garmin for professional GPS mapping and I've used Garmin devices for years, so my bias would tend towards Garmin being the most accurate, even in an older device.

 

[Regular.Cyclist]

Stick the GPS files from each into some mapping software and see how they match up with the route you know you took. The most accurate will map closest.

 

[Deleted member 26715]

I tried this a few times on the same route at the same time with a Garmin 800 & a Samsung 5 (maybe/possibly) when I then compared the 2 files, the Garmin read a longer distance as there were a lot of sharp 90 degree turns around housing developments, where the Garmin would plot every 5 yards the Samsung plotted every 20 which meant it often cut the corners.

On the long straight open roads it didn't make much difference, but in the woods where there was lots of twists & turns it did, also suspect speed would affect it, in the end my unscientific result was the Garmin was doing the job it was designed to do, the Samung was doing something it had been asked to do.

On the Garmin/Wahoo/Whatever front, no idea, suspect they would all be much of a muchness & is it really that important, if it is then maybe you need a 3rd calibrated wheel out back.

 

[cougie uk]

Stick the GPS files from each into some mapping software and see how they match up with the route you know you took. The most accurate will map closest.

Or just go with DC Rainmaker and his test of the latest Samsung. On that basis it's not as good as the wahoo and Garmin.

https://www.dcrainmaker.com/2022/08/samsung-galaxy-watch5-pro-early-sports-review.html/amp#

 

[Dogtrousers]

or I could just pick one to go with & that's 'The One'. But anyway...

That would be my recommended approach. Life's too short.

If you're really interested, get one of those wheels that surveyors use and walk the route with that. Do this several times to get an average reading with an associated error of measurement.

Now ride the route several times with your devices, and get averages and standard deviations for those too. And then you'll know.

 

[biking_fox]

I tried this a few times on the same route at the same time with a Garmin 800 & a Samsung 5 (maybe/possibly) when I then compared the 2 files, the Garmin read a longer distance as there were a lot of sharp 90 degree turns around housing developments, where the Garmin would plot every 5 yards the Samsung plotted every 20 which meant it often cut the corners.

On the long straight open roads it didn't make much difference, but in the woods where there was lots of twists & turns it did, also suspect speed would affect it, in the end my unscientific result was the Garmin was doing the job it was designed to do, the Samung was doing something it had been asked to do.

On the Garmin/Wahoo/Whatever front, no idea, suspect they would all be much of a muchness & is it really that important, if it is then maybe you need a 3rd calibrated wheel out back.

^this.

If you walk a long a sharpish arete you can quite easily tell the difference, those with poor positional fixes and infrequent updates zig zag eitehr side and gain a lot of 'extra' height and distance.

I very much doubt it will be a brand issue, but down to the specifications of particular models - updating more frequently will drain the battery quicker, run the processor hotter, require more memory etc etc hence I'd expect the watch to be lower accuracy.


I've always found real world GPS to give longer distances and more height, cf map planned routes.

 

[Dogtrousers]

Oddly enough I went for a walk at the weekend and recorded it using my phone (RideWithGPS app) and also a GPS (Garmin Oregon). The phone came out about 8% longer, but when I look in detail at the routes I see the phone track dithering all over the place and the GPS track being relatively smooth. The phone recorded six times the number of trackpoints.

I don't think I was staggering around from side to side constantly so I think that in this case the phone was recording faster than was justified by the accuracy of the device. The GPS was in "auto" mode which I think records more points when you change direction and fewer when going straight.

So rather than "cutting corners" and under reporting I suspect the GPS was actually doing a better job by recording a smoother path. I suspect there may be a "sweet spot" which balances the accuracy of the device and the frequency of sampling. Too high a frequency and you just record noise. Too low and you start cutting corners.

GPS:

Phone:

OS:

 

[T4tomo]

if you are primarily comparing different efforts against yourself, as long as its consistent it doesn't matter a jot!

 

[Dogtrousers]

if you are primarily comparing different efforts against yourself, as long as its consistent it doesn't matter a jot!

^ This!

 

[Two-Wheels]

Ahh but see that's where you're wrong. It did matter, in this instance (to me) . Though I'll agree that virtually any other time it wouldn't.

See, I've never covered 50 mile before. It's been something I'd been wanting to hit after doing about 4-5 40-45 mile rides but without getting in to it my earmarked day kind of went out the window & I was pushed back.

At the weekend I didn't even feel like going out but decided I really should get some kind of exercise in, even if it's just round the block.

So I went out with the idea of doing a short loop, which turned in to well I'll just try this extra bit, this extra bit, so on etc.

Before long I was doing my old climbs & 10 mile turned to 20 turned to 30. Then I realised I was a bit off home & I'll end up doing about 45 mile by the time I get home so I could pad it out a few mile to finally hit that 50.

Since I hadn't planned on being out so far/long, I took no food. Around mile 42 my legs were absolutely shot so just kind of auto piloted my way towards 50, but that's when I thought - I wonder what the Garmin says. It was a few mile less.

I was desperate to knock it on the head as I was knackered. I thought - I'll just go off the Wahoo as it's further ahead than the Garmin, but then knowing what I'm like, I'd have got home, the Garmin would've said 49.9 mile and I wouldn't have been satisfied.

So I just kept riding around the local area then to add on those extra few miles to take me over the line. Went past a guy washing his car 3 times. Must've thought I was up to no good lol

Wahoo Roam: 52.16 mile / 14.7mph av / 3940 ft or 1201mtrs
Garmin Forerunner 735XT: 50.46 mile / 14.3mph av / 967mtrs
Samsung Galaxy Watch 3 46mm: 51.05 mile / 14.0mph av / 5.62 mile (allegedly)

 

[Alex321]

Wahoo Roam: 52.16 mile / 14.7mph av / 3940 ft or 1201mtrs
Garmin Forerunner 735XT: 50.46 mile / 14.3mph av / 967mtrs
Samsung Galaxy Watch 3 46mm: 51.05 mile / 14.0mph av / 5.62 mile (allegedly)

If all else was equal, I would tend to expect the larger device to be probably better than the watches.

 

[Ming the Merciless]

It depends on

• The accuracy of each position recorded
• The number of positions recorded

The former depends on how modern the gps chipset is, and the placement and orientation of the device. Something out on the handlebars will get a clearer view of the sky which may mean it’s getting clean signals from more satellites. All good to reducing the error in positioning. You can also turn on EGNOS in settings to receive a correction signal calculated by ground stations.

The latter is often called the sampling rate and can be changed in settings for most devices , though some are dumbed down.

From a practical point of view for you.

Mount devices out on the bars orientated to the sky. Set the sampling rate to its maximum value, usually once a second.

If you want to know a consistent distance for a route then use an online mapping to plot your route. But if you follow the above practice (and don’t keep gps running if you go indoors) you get pretty consistent distances with low errors.