Road Bike, Cycling Forums banner
1 - 11 of 11 Posts

·
Registered
Joined
·
257 Posts
Discussion Starter · #1 ·
I'm trying to figure out how far I actually ride.
On a long ride today, I used a RidewithGPS route of 103.0 miles. My Garmin had me at 104.1, Strava at 102.9.

My Garmin 820 is set to get a reading every second, and I did a measured roll-out for the tire circumference.

Which is the most accurate?
 

·
Registered
Joined
·
3,890 Posts
No idea. Your numbers are within the expected accuracy of each other.

As an aside, you can set your Garmin to auto-calibrate (I think that should be the default), there's no need to do a roll-out test. The auto-calibration of your Garmin can be expected to be more accurate than a manual roll-out test.
 

·
Registered
Joined
·
257 Posts
Discussion Starter · #3 ·
No idea. Your numbers are within the expected accuracy of each other.

As an aside, you can set your Garmin to auto-calibrate (I think that should be the default), there's no need to do a roll-out test. The auto-calibration of your Garmin can be expected to be more accurate than a manual roll-out test.
Yes I've used the auto calibrate setting. But of you change wheels or tires (which I have), the Garmin does not recalculate for the new size. I realized this when I switched from Bontrager R3 25c's on PAradigm Elite wheels (I think they were 2096 mm) to Schwalbe Pro One 25c's on Aeolus 3 D3 rims(they were really 28 mm wide, and measured 2107mm om the roll-out). After the switch, I saw that the Garmin did not change the\ tire circumference. That's close to a 4% difference. I then corrected it, but now have the differences that I wrote about today.
 

·
Registered
Joined
·
3,695 Posts
You need to look at how ridewithgps calculates distance. GPS accuracy is not a garmin issue but a problem with satellites in general. signals bounce etc

Also found the following on a site.

GPS units never know exactly where you are located–they’re only accurate to within a few feet. So each time a GPS checks with the satellite, it’s calculating a slightly different position, even when you’re not moving
 

·
Registered
Joined
·
1,017 Posts
you also have to understand that all 3 of those things you are measuring with are just approximates, they could be accurate to the inch by random luck or they could be out by who knows what, the only way you will know for sure is use a certified measuring wheel thingo they use for course measurement like for marathons etc).
The fact that you got all 3 within 1.2 miles is pretty good.
 

·
Registered
Joined
·
1,039 Posts
I use three wired computers for my bikes. When I map the ride I've done on ridewithgps, it almost always comes very close to matching with all three. From hundreds of rides, I would say that ridewithgps is very accurate with their distance. I know there were some hiccups with their elevation measuring but it's been stable for a while now.
 

·
Registered
Joined
·
16 Posts
It's highly unlikely that they all would ever agree, or that any of them would be accurate in the absolute sense, for a number of reasons.

First of all, you probably aren't making the best use of GPS because virtually all civilian SatNav devices only use a single receiver. Military SatNav devices are more likely to use dual receivers because receiving two frequencies simultaneously allows the device to null out atmospheric-induced errors and produce a more precise and more accurate location. Every time you ride past a tree or any structure that comes between you and any of the (typically eight) satellites your GPS is receiving from, your device receives a multipathed signal, which degrades its accuracy. But you'll rarely see dual-receiver GPS in civilian use because it's too expensive. It's not just the cost of a second receiver, there's also a lot of processing hardware that compares, mixes and matches the two signals to produce enhanced precision.

Second, GPS just isn't that accurate. Everything from ambient temperature to cloud cover to solar flares to the state of discharge of the battery has a tiny impact on GPS accuracy. A GPS is basically a stopwatch measuring the length of time a radio signal takes to reach it from a bunch of transmitters that are 12,500 miles above the surface of the earth. The problem is, it's making those measurements in the billionths of a second. A timing error of a billionth of a second means its location calculation is off by about a foot. So even the tiniest of induced errors will materially affect its calculations.

Third, GPSs don't even measure distance, or speed, they measure your position relative to all the GPS satellites you have line of sight to. Location always was its primary function and everything else is a derivative of a series of locations (and their associated times). In order to produce a record of distance traveled, current velocity or ETA to a destination, it has to interpolate between multiple points of data which in themselves ALL contained errors. And no mathematical solution ever can be any more accurate than the LEAST accurate term used to produce it.

The units in a smartphone typically good to no better than about 16 feet. Hi-end (civilian) units, like those in commercial aircraft, are good to maybe six feet.

Fourth, nobody rides a bicycle in a straight line. You weave when you ride. So how does a GPS that's only accurate to 16 feet (or even six feet) measure the additional ground you cover from your endlessly weaving, even if it's just an inch or two? It might detect some of it but never all of it. Or one GPS might detect part of one particular weave that the indentical GPS right next to it overlooked. And if you weave a couple of inches every 20 seconds over a five hour ride, how many weaving inches have you ridden that at best only were partially detected?

And that's without getting into Heisenberg's uncertainty principle. It's impossible to measure both position and velocity with perfect accuracy at the same time.

The truth is, for measuring distance, a $2 tape measure from Wally World will always be more accurate than the best GPS ever built. A simple bicycle computer that calculates distance based on wheel counting rotations always will be more accurate because it was designed specifically for measuring distance, and it can tell when you weave, even the tiniest bit. Just be sure to calibrate it meticulously, then re-calibrate when any change occurs that affects tire circumference.
 

·
Registered
Joined
·
1,830 Posts
It's highly unlikely that they all would ever agree, or that any of them would be accurate in the absolute sense, for a number of reasons.

First of all, you probably aren't making the best use of GPS because virtually all civilian SatNav devices only use a single receiver. Military SatNav devices are more likely to use dual receivers because receiving two frequencies simultaneously allows the device to null out atmospheric-induced errors and produce a more precise and more accurate location. Every time you ride past a tree or any structure that comes between you and any of the (typically eight) satellites your GPS is receiving from, your device receives a multipathed signal, which degrades its accuracy. But you'll rarely see dual-receiver GPS in civilian use because it's too expensive. It's not just the cost of a second receiver, there's also a lot of processing hardware that compares, mixes and matches the two signals to produce enhanced precision.

Second, GPS just isn't that accurate. Everything from ambient temperature to cloud cover to solar flares to the state of discharge of the battery has a tiny impact on GPS accuracy. A GPS is basically a stopwatch measuring the length of time a radio signal takes to reach it from a bunch of transmitters that are 12,500 miles above the surface of the earth. The problem is, it's making those measurements in the billionths of a second. A timing error of a billionth of a second means its location calculation is off by about a foot. So even the tiniest of induced errors will materially affect its calculations.

Third, GPSs don't even measure distance, or speed, they measure your position relative to all the GPS satellites you have line of sight to. Location always was its primary function and everything else is a derivative of a series of locations (and their associated times). In order to produce a record of distance traveled, current velocity or ETA to a destination, it has to interpolate between multiple points of data which in themselves ALL contained errors. And no mathematical solution ever can be any more accurate than the LEAST accurate term used to produce it.

The units in a smartphone typically good to no better than about 16 feet. Hi-end (civilian) units, like those in commercial aircraft, are good to maybe six feet.

Fourth, nobody rides a bicycle in a straight line. You weave when you ride. So how does a GPS that's only accurate to 16 feet (or even six feet) measure the additional ground you cover from your endlessly weaving, even if it's just an inch or two? It might detect some of it but never all of it. Or one GPS might detect part of one particular weave that the indentical GPS right next to it overlooked. And if you weave a couple of inches every 20 seconds over a five hour ride, how many weaving inches have you ridden that at best only were partially detected?

And that's without getting into Heisenberg's uncertainty principle. It's impossible to measure both position and velocity with perfect accuracy at the same time.

The truth is, for measuring distance, a $2 tape measure from Wally World will always be more accurate than the best GPS ever built. A simple bicycle computer that calculates distance based on wheel counting rotations always will be more accurate because it was designed specifically for measuring distance, and it can tell when you weave, even the tiniest bit. Just be sure to calibrate it meticulously, then re-calibrate when any change occurs that affects tire circumference.
Can somebody make this a sticky ?, great response.

FWIW, I used to use calibrated Cateye wireless bike computers, found my Garmin 810 to be accurate to .6 miles in 100. That was plenty accurate for my purposes.
 

·
Russian Troll Farmer
Joined
·
3,430 Posts
Sunday, I turned Strava on, and while standing still, it said I was going back and forth at a velocity of 1.6 mph....
 

·
Cycling Coach
Joined
·
1,734 Posts
Yes I've used the auto calibrate setting. But of you change wheels or tires (which I have), the Garmin does not recalculate for the new size. I realized this when I switched from Bontrager R3 25c's on PAradigm Elite wheels (I think they were 2096 mm) to Schwalbe Pro One 25c's on Aeolus 3 D3 rims(they were really 28 mm wide, and measured 2107mm om the roll-out). After the switch, I saw that the Garmin did not change the\ tire circumference. That's close to a 4% difference. I then corrected it, but now have the differences that I wrote about today.
2096/2107 = 0.9948

That's a 0.5% difference.
 

·
Cycling Coach
Joined
·
1,734 Posts
2096/2107 = 0.9948

That's a 0.5% difference.
BTW - a wheel magnet sensor properly working is the best for accuracy. Even getting wheel circumference measurement wrong by 5mm is still only a 0.2% error. With care you can get it within 1-2 mm.

Try riding a local crit circuit and compare wheel sensor measurement with a GPS.
 
1 - 11 of 11 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top