Skip to main content

The following note exchange discusses a growing problem with runners who don't think a course is accurately measured, because their GPS indicates a total distance that different than the race distance.

My response:

There is going to have to be some learning on the part of GPS users, because unfortunately they are not all that accurate, especially if you get into trees or areas with tall building that block the satellites. The signals from the satellites are even blocked by ones own body, which is why some of the manufacturers, not Garmin, have a separate receiver that you wear on the upper arm. I have the Garmin Forerunner which is the model I assume this person is using. My distance at the end of the Royal Victoria Marathon was something over 43km, not accurate but not bad if you are just out for a training run. Today I ran the Richmond Flatlands 10km and got 10.2 and I know it is 10km; I measurered it.

The Garmin, in generally open conditions, is usually 98% accurate, which for a marathon means 42.2+/-0.8km. The accuracy can be much worse than 98% if one is running through heavy tree cover or in the urban jungle. The person who wrote this note is well within the 98% accuracy at 26.6 miles. She is also mixing up her accuracies as the 15m accuracy refers to the position accuracy, not the total distance.

Sylvan, I am going to post this on the Measurement News Forum as I am sure it is a problem we will have to deal with more and more as the popularity of the these GPS units grows.

-----Original Message-----
From: Sylvan Smyth []
Sent: Sunday, October 16, 2005 2:16 AM
To: 'Laurie'
Subject: GPS Measurement

Laurie, what do you think of this blast RVM took on the comments page for RVM? I think she's nuts and just ran sloppy, off the tangents.

"Victoria is an amazing city and race. I had a blast visiting the city and running the marathon. Unfortunately, the course is over by .40 miles. I was wearing a Garmin GPS that is 98% accurate within 15 meters. A friend of mine was wearing the same kind of GPS but came in 14 min. ahead of me with the same distance. 26.65. I only saw one other person in the race wearing one.
My time at mile 26 was a 3:41:12, now why in the world would it take me another 5 min. to go .20 if the course was accurate. It wouldn't be such a big deal, but unfortunately I missed Boston by 40 sec. Had the course been accurate I would have been 2 min under my qualification, but as it stands
26.61 miles, 3:46:40 time, 8:31 pace. I've e-mail the Race and course director about this, doesn't sound like the care too much. I'm one person in thousands. I will continue on my quest for Boston."

Original Post

Replies sorted oldest to newest

Yes Laurie it is .1% long--but at her pace if my calculations are correct that only gives her back about 14-15 seconds for a marathon--still came in too late--she needs to qualify on certified courses and will have to try again.

I am glad to see others post this topic. It underscores the painstaking accuracy with which we take this task when people trust GPS more than our measurements.

I frequently train on courses that I have measured and love my Garmin Forerunner but it is rare to get the same distance twice even in short distances. I trust it less than a guy who says "would you like to play poker for money? I have never played before."

I use my Garmin during races and qualified for Boston wearing one. They are great but only when used correctly. I suggested turning off the auto mile lap notification during a race and physically hitting each split. It doesn't matter what the Garmin says a mile is in the race--the mile is where the marker is and on a certified course it is likely more accurate than the GPS. Those splits should tell you your real pace for the given event and help get you in on time if you trained well and have a good day.
Comparing a GPS reading with a certified course is comparing apples and oranges. The methodology of getting a course certified has been written down and can be followed by anybody with modest arithmetical skill.

To compare this documented technology with a label on a product that makes a claim to accuracy is ridiculous. The only thing that can be concluded from these GPS readings is that they are not accurate within the limits claimed.
I tend to take the claims of a GPS user after a race (I hear it all of the time locally) with the appropriate amount of salt - grain, shake or entire shaker - especially when they are running on a course that I have personally measured.

I can vouch for the protocol I followed for calibrating my JO and making certain to ride the shortest possible distance, compared to the typical GPS user who may or may not have taken the effort to ensure their calibration is correct, probably did not start at the starting line and most likely did not run the shortest possible distance for the entire course.

In fact, since I started measuring courses locally I've heard more people complain about courses being 'long.'
The best argument I have read to date regarding the differences in measurements using GPS and the Jones Counter is the testing results Stuart Gordon had published here in January, see

Stuart’s results simply takes the air out of those suggesting that they’re newly purchased GPS wrist watch can determine the “true” length of the road course they just ran. New technologies really can be a little dangerous in the hands of a few.

It will be further interesting reading when Stuart publishes his results on aerial photogrammetry.

Add Reply

Link copied to your clipboard.