Go Back  Bike Forums > Bike Forums > General Cycling Discussion
Reload this Page >

Bike Computer Accuracy

Notices
General Cycling Discussion Have a cycling related question or comment that doesn't fit in one of the other specialty forums? Drop on in and post in here! When possible, please select the forum above that most fits your post!

Bike Computer Accuracy

Old 01-21-21, 07:35 AM
  #26  
Flip Flop Rider
Senior Member
 
Flip Flop Rider's Avatar
 
Join Date: Apr 2016
Location: South Carolina Upstate
Posts: 2,103

Bikes: 2010 Fuji Absolute 3.0 1994 Trek 850

Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 761 Post(s)
Liked 553 Times in 320 Posts
it's close enough. don't obsess
Flip Flop Rider is offline  
Likes For Flip Flop Rider:
Old 01-21-21, 07:41 AM
  #27  
Paul Barnard
For The Fun of It
 
Paul Barnard's Avatar
 
Join Date: May 2007
Location: Louisissippi Coast
Posts: 5,843

Bikes: Lynskey GR300, Lynskey Backroad, Litespeed T6, Lynskey MT29, Burley Duet

Mentioned: 12 Post(s)
Tagged: 0 Thread(s)
Quoted: 2131 Post(s)
Liked 1,639 Times in 822 Posts
Originally Posted by bikehoco
Bugger off smartass.
Hey, that was unpolite.
Paul Barnard is offline  
Likes For Paul Barnard:
Old 01-21-21, 11:01 AM
  #28  
Phil_gretz
Zip tie Karen
 
Phil_gretz's Avatar
 
Join Date: Aug 2009
Location: Fair Oaks Ranch, TX
Posts: 7,006

Bikes: '13 Motobecane Fantom29 HT, '16 Motobecane Turino Pro Disc, '18 Velobuild VB-R-022, '21 Tsunami SNM-100

Mentioned: 26 Post(s)
Tagged: 0 Thread(s)
Quoted: 1465 Post(s)
Liked 1,542 Times in 806 Posts
I'm going to agree with the general sentiment expressed here, that the data display from a simple bike computer will have inherent inaccuracies. And this is okay. One can calibrate out errors from the wheel's circumference under load (roll-out 3x and then average isn't a bad method). But what are the other sources? Sampling rate error from the sensor is most likely the next largest. Next will be rounding/truncation errors in the algorithm's math stack. The filtering algorithms will contribute negligible errors. Last will be display resolution. Does any of this matter to the average bicycle rider?

GPS-based systems have their own sources of error. The math's more complicated, too. Does this matter?

It's not the mission to Mars. It's a bike ride.
Phil_gretz is offline  
Likes For Phil_gretz:
Old 01-21-21, 11:24 AM
  #29  
Steve B.
Senior Member
 
Join Date: Jul 2007
Location: South shore, L.I., NY
Posts: 6,825

Bikes: Flyxii FR322, Cannondale Topstone, Miyata City Liner, Specialized Chisel, Specialized Epic Evo

Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 3185 Post(s)
Liked 2,020 Times in 1,158 Posts
Originally Posted by Phil_gretz
I'm going to agree with the general sentiment expressed here, that the data display from a simple bike computer will have inherent inaccuracies. And this is okay. One can calibrate out errors from the wheel's circumference under load (roll-out 3x and then average isn't a bad method). But what are the other sources? Sampling rate error from the sensor is most likely the next largest. Next will be rounding/truncation errors in the algorithm's math stack. The filtering algorithms will contribute negligible errors. Last will be display resolution. Does any of this matter to the average bicycle rider?

GPS-based systems have their own sources of error. The math's more complicated, too. Does this matter?

It's not the mission to Mars. It's a bike ride.
What folks forget is that GPS records a bread crumb (every second maybe ?). There's inherent distance errors when the breadcrumb is rounding off things like a curvy road, single track on a mt. bike, etc.... Is why a speed sensor provides for a better distance record.
Steve B. is offline  
Old 01-21-21, 11:37 AM
  #30  
Hiro11
Senior Member
 
Join Date: Jun 2010
Posts: 2,608

Bikes: 2022 Specialized Allez Sprint custom build, 2019 Giant Defy Advanced Pro 0, 2018 Seven Mudhoney Pro custom build, 2017 Raleigh Stuntman, various others

Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 782 Post(s)
Liked 475 Times in 238 Posts
I've found this varies by the type of ride and conditions of the ride. My Garmin is generally pretty consistent in speed and distance on the road but becomes less accurate with more tree cover and more twisty routes. The Garmin seems particularly inaccurate on singletrack in the woods when compared to dead reckoning based methods like wheel sensors, it can be off by 10%-15% on both distance and average speed (always underestimating both, of course). It also bugs my neuroses and distracts me to see the speed bounce up and down by three MPH when I'm riding steadily. I gave up and installed wireless speed sensors on all of my bikes. They're cheap and they solve the problem. Of course, none of this actually matters...
Hiro11 is offline  
Old 01-21-21, 05:34 PM
  #31  
Camilo
Senior Member
 
Join Date: Apr 2007
Posts: 6,760
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 1106 Post(s)
Liked 1,197 Times in 758 Posts
Originally Posted by aplcr0331
Go with whichever is higher.

I actually use a Wahoo speed sensor on my front hub with my Bolt. Is it accurate? Meh...it’s consistent so that’s what I go with.

We’re not in danger of setting any land speed records.
And that goes for a group ride - use whichever rider has the longest distance and fastest average speed.
Camilo is offline  
Likes For Camilo:
Old 01-21-21, 05:43 PM
  #32  
bikehoco
Full Member
Thread Starter
 
Join Date: Jun 2017
Location: Virginia
Posts: 218
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 180 Post(s)
Liked 85 Times in 53 Posts
I’ve concluded my wired computer is accurate and the problem is with the Ride With GPS app.

The apps is great for keeping a log of all rides; I just wish it was more accurate. I may try Map My Ride.

Last edited by bikehoco; 01-21-21 at 05:47 PM.
bikehoco is offline  
Old 01-22-21, 08:18 AM
  #33  
jadocs
Senior Member
 
jadocs's Avatar
 
Join Date: Jul 2018
Location: USA
Posts: 2,192

Bikes: Ti, Mn Cr Ni Mo Nb, Al, C

Mentioned: 22 Post(s)
Tagged: 0 Thread(s)
Quoted: 942 Post(s)
Liked 526 Times in 349 Posts
I get what people are saying with regard to the difference being so small it doesn't matter. However, lets say one average says 19.9mph and the other says 20.8mph...although the margin of error is the same...the difference (all things being equal) between averaging 19.9 and 20.8 is pretty significant.
jadocs is offline  
Likes For jadocs:
Old 01-22-21, 09:41 AM
  #34  
pdlamb
Senior Member
 
Join Date: Dec 2010
Location: northern Deep South
Posts: 8,842

Bikes: Fuji Touring, Novara Randonee

Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 2575 Post(s)
Liked 1,900 Times in 1,192 Posts
Originally Posted by bikehoco
I’ve concluded my wired computer is accurate and the problem is with the Ride With GPS app.

The apps is great for keeping a log of all rides; I just wish it was more accurate. I may try Map My Ride.
Don't hold your breath. There may be a slight change in how the apps process it, but the major inaccuracies are going to be in the GPS hardware and software embedded in your phone (as noted above, multiple times).
pdlamb is offline  
Old 01-22-21, 10:06 AM
  #35  
berner
Senior Member
 
Join Date: May 2010
Location: Bristol, R. I.
Posts: 4,340

Bikes: Specialized Secteur, old Peugeot

Mentioned: 20 Post(s)
Tagged: 0 Thread(s)
Quoted: 663 Post(s)
Liked 496 Times in 299 Posts
A statute mile is 5280 feet so one percent is a mere 52.8 feet. A two percent error would be 105.6 feet. On a century ride the error would be tiny.
berner is offline  
Old 01-22-21, 12:07 PM
  #36  
wphamilton
Senior Member
 
wphamilton's Avatar
 
Join Date: Apr 2011
Location: Alpharetta, GA
Posts: 15,280

Bikes: Nashbar Road

Mentioned: 71 Post(s)
Tagged: 0 Thread(s)
Quoted: 2934 Post(s)
Liked 341 Times in 228 Posts
The wired bike computer is the more authoritative source for speed and distance. GPS wins for convenience and additional functionality, which is the only good reason to use GPS at all IMO. Keeping a record. Convenient for analysis. Integration with maps.
wphamilton is offline  
Likes For wphamilton:
Old 01-22-21, 01:58 PM
  #37  
Steve B.
Senior Member
 
Join Date: Jul 2007
Location: South shore, L.I., NY
Posts: 6,825

Bikes: Flyxii FR322, Cannondale Topstone, Miyata City Liner, Specialized Chisel, Specialized Epic Evo

Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 3185 Post(s)
Liked 2,020 Times in 1,158 Posts
Originally Posted by bikehoco
I’ve concluded my wired computer is accurate and the problem is with the Ride With GPS app.

The apps is great for keeping a log of all rides; I just wish it was more accurate. I may try Map My Ride.
Any inaccuracies are likely from the phone. When I plot out a route on RWGPS online, it's typically within 1/10 a mile in 20 accurate to what I end up riding.
Steve B. is offline  
Likes For Steve B.:
Old 01-22-21, 03:00 PM
  #38  
tomato coupe
Senior Member
 
Join Date: Jul 2009
Posts: 5,879

Bikes: Colnago, Van Dessel, Factor, Cervelo, Ritchey

Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 3905 Post(s)
Liked 7,181 Times in 2,905 Posts
Originally Posted by wphamilton
The wired bike computer is the more authoritative source for speed and distance. GPS wins for convenience and additional functionality, which is the only good reason to use GPS at all IMO. Keeping a record.
A GPS computer, when used with a speed sensor, is generally much more accurate than a wired bike computer, for both speed and distance.
tomato coupe is offline  
Likes For tomato coupe:
Old 01-22-21, 04:20 PM
  #39  
alcjphil
Senior Member
 
alcjphil's Avatar
 
Join Date: Jun 2002
Location: Montreal, Quebec
Posts: 5,869
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 1792 Post(s)
Liked 1,671 Times in 955 Posts
Originally Posted by Phil_gretz
It's not the mission to Mars. It's a bike ride.
I agree wholeheartedly.
alcjphil is offline  
Old 01-22-21, 04:48 PM
  #40  
pdlamb
Senior Member
 
Join Date: Dec 2010
Location: northern Deep South
Posts: 8,842

Bikes: Fuji Touring, Novara Randonee

Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 2575 Post(s)
Liked 1,900 Times in 1,192 Posts
Originally Posted by tomato coupe
A GPS computer, when used with a speed sensor, is generally much more accurate than a wired bike computer, for both speed and distance.
IME that's true for a straight road without any nearby vertical obstacles, like trees, hillsides or cliffs, or buildings. IOW not where I normally ride.
pdlamb is offline  
Old 01-25-21, 02:56 PM
  #41  
bikehoco
Full Member
Thread Starter
 
Join Date: Jun 2017
Location: Virginia
Posts: 218
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 180 Post(s)
Liked 85 Times in 53 Posts
I suspect the primary difference in mph comes down to how the two methods deal with pauses (e.g. red lights).
bikehoco is offline  
Old 01-26-21, 06:37 AM
  #42  
ZHVelo
Senior Member
 
Join Date: Apr 2020
Location: Zurich, Switzerland
Posts: 877
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 528 Post(s)
Liked 230 Times in 161 Posts
If the distance between the two is almost identical but the average speed is not, that must mean they have a difference in time. Have you compared those? Does the app have an auto start stop functionality?

I used to ride with a bike computer with sensor though wireless on the wheel not wired and wahoo on phone. When I didn't stop, both time and distance matched close enough for my liking. When I did the wahoo always read less time given the auto stop is what 3kmh or maybe 5 idk what I set. Instantaneous speed displayed differed too though in most cases was fairly close and the more even i kept going the more aligned they were. Similar to my two HR monitors. My polar band reacts a lot faster than my fitbit watch, so one might be 150 the other still 140, but a few seconds later the fitbit would catch up. Similar going down and overall the average HR was never more than 1, a few times 2 bpm different.
ZHVelo is offline  
Old 01-27-21, 11:41 PM
  #43  
wphamilton
Senior Member
 
wphamilton's Avatar
 
Join Date: Apr 2011
Location: Alpharetta, GA
Posts: 15,280

Bikes: Nashbar Road

Mentioned: 71 Post(s)
Tagged: 0 Thread(s)
Quoted: 2934 Post(s)
Liked 341 Times in 228 Posts
Originally Posted by tomato coupe
A GPS computer, when used with a speed sensor, is generally much more accurate than a wired bike
computer, for both speed and distance.
Not true IMO. The speed sensor is the same hardware as a wired bike computer, and you're probably thinking that it's only used for measuring instantaneous speed and the GPS is used for overall distance.

The wired bike computer properly calibrated is much more accurate (in the case that the GPS ignores the wheel sensor for distance) because of the numerous sources of error of the GPS. Primarily among them, the transient variance of measured position (5 meter accuracy) by the GPS means that it only approximates the actual path travelled.
wphamilton is offline  
Old 01-28-21, 01:27 AM
  #44  
tomato coupe
Senior Member
 
Join Date: Jul 2009
Posts: 5,879

Bikes: Colnago, Van Dessel, Factor, Cervelo, Ritchey

Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 3905 Post(s)
Liked 7,181 Times in 2,905 Posts
Originally Posted by wphamilton
Not true IMO. The speed sensor is the same hardware as a wired bike computer, and you're probably thinking that it's only used for measuring instantaneous speed and the GPS is used for overall distance.

The wired bike computer properly calibrated is much more accurate (in the case that the GPS ignores the wheel sensor for distance) because of the numerous sources of error of the GPS. Primarily among them, the transient variance of measured position (5 meter accuracy) by the GPS means that it only approximates the actual path travelled.
1) Rollout calibrations for a GPS-based computer can actually be a bit better than a wired computer, because the rollout measurement can be performed automatically at the beginning of every ride over a much longer distance.

2) 5m is the typical absolute accuracy of a GPS-based bike computer, but you don't need high absolute accuracy for distance measurement, just high precision. (Accuracy and precision here are used in the strict metrological sense.)

3) Wired computers will measure distance better at short distances than a GPS-based computer, if the GPS computer is not used with a speed sensor. The distance error of a wired computer accumulates (linearly) as the distance increases, however, giving a GPS-based computer the advantage at long distances because GPS positional errors do not accumulate linearly.

4) If a GPS-based computer is used with a speed sensor, there's no reason it couldn't measure distance as well as a wired computer at short distances, but I don't know if they use speed sensor data when they calculate distance.
tomato coupe is offline  
Old 01-28-21, 07:15 AM
  #45  
Steve B.
Senior Member
 
Join Date: Jul 2007
Location: South shore, L.I., NY
Posts: 6,825

Bikes: Flyxii FR322, Cannondale Topstone, Miyata City Liner, Specialized Chisel, Specialized Epic Evo

Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 3185 Post(s)
Liked 2,020 Times in 1,158 Posts
Originally Posted by tomato coupe
1) Rollout calibrations for a GPS-based computer can actually be a bit better than a wired computer, because the rollout measurement can be performed automatically at the beginning of every ride over a much longer distance.

2) 5m is the typical absolute accuracy of a GPS-based bike computer, but you don't need high absolute accuracy for distance measurement, just high precision. (Accuracy and precision here are used in the strict metrological sense.)

3) Wired computers will measure distance better at short distances than a GPS-based computer, if the GPS computer is not used with a speed sensor. The distance error of a wired computer accumulates (linearly) as the distance increases, however, giving a GPS-based computer the advantage at long distances because GPS positional errors do not accumulate linearly.

4) If a GPS-based computer is used with a speed sensor, there's no reason it couldn't measure distance as well as a wired computer at short distances, but I don't know if they use speed sensor data when they calculate distance.
A GPS (at least a Garmin) will always use the data from a speed sensor for speed and distance over the GPS data.

I would expect that a standard (non GPS) bike computer that uses a wheel magnet, and when calibrated for tire size, would be more accurate than a GPS without a speed sensor, and as accurate as a GPS using a speed sensor. But without using a surveyors roll out measurement, I've no clue which might be MORE accurate and doubt it matters real world. I mean how accurate does all this need to be ?.
Steve B. is offline  
Old 01-28-21, 09:47 AM
  #46  
njkayaker
Senior Member
 
Join Date: Sep 2007
Location: Far beyond the pale horizon.
Posts: 14,240
Mentioned: 31 Post(s)
Tagged: 0 Thread(s)
Quoted: 4221 Post(s)
Liked 1,318 Times in 914 Posts
Originally Posted by tomato coupe
1) Rollout calibrations for a GPS-based computer can actually be a bit better than a wired computer, because the rollout measurement can be performed automatically at the beginning of every ride over a much longer distance.
This is about accurately measuring the circumference. That is, whether GPS is more accurate for measuring circumference is a different question from whether it's more accurate for distance.(Obviously, the wired computer needs an accurate circumference.)

Originally Posted by tomato coupe
4) If a GPS-based computer is used with a speed sensor, there's no reason it couldn't measure distance as well as a wired computer at short distances, but I don't know if they use speed sensor data when they calculate distance.
It's not really a "speed sensor". It's a wheel rotation sensor. A wired computer also counts wheel rotations (it's the same thing as a "speed sensor"). It's not clear how you think having both ("GPS-based computer is used with a speed sensor") makes GPS better.

========================

One issue with GPS is loss of signal.

https://support.garmin.com/en-US/?fa...623Z535geTx2e9

The wheel rotation counter is taking measurements 3 times a second. The GPS is typically only taking measurements once every second.

That means the counter has a higher resolution than GPS (which has a straight line between each one second measurement).

The GPS measurement has an error of 10 feet or more. The wheel counter has error on the order of centimeters. This really matters for accurate "instantaneous" speed.

If you have an accurate circumference, the distance based on wheel rotations will be very accurate. There's no reason to expect that GPS (at the low speed of bicycles) would be better.

Last edited by njkayaker; 01-28-21 at 09:59 AM.
njkayaker is offline  
Old 01-28-21, 10:01 AM
  #47  
njkayaker
Senior Member
 
Join Date: Sep 2007
Location: Far beyond the pale horizon.
Posts: 14,240
Mentioned: 31 Post(s)
Tagged: 0 Thread(s)
Quoted: 4221 Post(s)
Liked 1,318 Times in 914 Posts
Originally Posted by Phil_gretz
I'm going to agree with the general sentiment expressed here, that the data display from a simple bike computer will have inherent inaccuracies. And this is okay. One can calibrate out errors from the wheel's circumference under load (roll-out 3x and then average isn't a bad method). But what are the other sources? Sampling rate error from the sensor is most likely the next largest. Next will be rounding/truncation errors in the algorithm's math stack. The filtering algorithms will contribute negligible errors. Last will be display resolution. Does any of this matter to the average bicycle rider?

GPS-based systems have their own sources of error. The math's more complicated, too. Does this matter?

It's not the mission to Mars. It's a bike ride.
Incurious.
njkayaker is offline  
Old 01-28-21, 10:04 AM
  #48  
stevoo
Stevoo
 
Join Date: Jun 2013
Location: So Cal
Posts: 220

Bikes: Road and mountain tandems, single bikes too.

Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 34 Post(s)
Liked 4 Times in 4 Posts
Yup, agree roll out method is quite accurate IF it is calibrated properly.
They measure and certify running race courses using the roll out method. Look up Jones counter. Accurate to 1 part to 1000.
stevoo is offline  
Old 01-28-21, 12:54 PM
  #49  
tomato coupe
Senior Member
 
Join Date: Jul 2009
Posts: 5,879

Bikes: Colnago, Van Dessel, Factor, Cervelo, Ritchey

Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 3905 Post(s)
Liked 7,181 Times in 2,905 Posts
Originally Posted by Steve B.
I would expect that a standard (non GPS) bike computer that uses a wheel magnet, and when calibrated for tire size, would be more accurate than a GPS without a speed sensor ...
For short distances and (instantaneous) speed, yes.
... and as accurate as a GPS using a speed sensor.
Yes, assuming the rollout calibration is the same for both.
But without using a surveyors roll out measurement, I've no clue which might be MORE accurate...
The GPS can be more accurate because it can calibrate the rollout over a longer distance.
... and doubt it matters real world. I mean, how accurate does all this need to be ?.
Agreed. Both wired computers and GPS computers without speed sensors are accurate enough for the vast majority of riders.
tomato coupe is offline  
Old 01-28-21, 01:17 PM
  #50  
bikehoco
Full Member
Thread Starter
 
Join Date: Jun 2017
Location: Virginia
Posts: 218
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 180 Post(s)
Liked 85 Times in 53 Posts
“It’s close enough, don’t obsess”
”It’s not the mission to Mars. It’s a bike ride”

Thanks for your input, I always find it valuable when anonymous people on the internet tell me what I should-shouldn’t care about. /s
bikehoco is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service -

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.