If one more person complains...

PhilPub

EG Graham wrote (see)

 It is a reasonable stab at estimation but there is no real accuracy, the minor errors in the many variables will produce a compound error that could amount to 5-10%, similar to the use of two Garmins by the same person, two people using identical bikes and Jones set-ups will produce varying results.

I find this figure entirely implausible.  I've been using a basic, non-GPS computer on my bike(s) for donkeys' years to measure trip distance.  This is effectively a Jones counter because it simply counts the revolutions of the wheel and you input the wheel/tyre size.  (So actually it's probably slightly less accurate than a Jones counter becaue I guess even standard size tyres can vary slightly, with temperature, etc.)  I've not sat down and worked out the stats but from general observation, the figures I get from riding the same routes invariably fall within a very, very similar range, e.g. I know that such-and-such route is between 44 and 45 miles; I would be completely dumbfounded if I came back from a ride I've done possibly dozens of times before, to see that my computer says I cycled 48 miles.  It just doesn't happen.

If two course measurers come up with anything like significant differences, this would suggest to me measurer error (one person cut off the race line going round a corner, etc.) rather than a systematic error with the Jones counter itself.  This is unlike the additional errors that GPS units are prone to, including random errors of lost signal, and systematic errors due to interpolation, e.g. I'm fairly convinced after using a Garmin for numerous interval sessions that it systematically overestimates how far you travel when running around a track.  Presumably this is something to do with the adjustment it makes in estimating repeated curves of the track from straight-line observations.

2Old

I want to complain as when I ran VLM in 2010 I missed GFA by 26 seconds and the course according to my Garmin was a good bit longer than 26.2.So technically I ran 26.2 in the GFA time.Can I apply to a higher authority to have this reviewed? Really I accept I missed it but this does show how close things can be even in a marathon .Maybe I didnt cut enough corners and shouldnt have stopped to hug my wife ....oh well..must try harder this year

Toucs

So do we all agree then that GPS systems are more accurate than UKA measured courses? 

Kryten

Agreed.

That is probably why UKA officially don't allow GPS systems in races; they don't want us to know they can't measure the course properly.

RicF

We need to spell it out. Y O U R G A R M I N I S N O T A S A C C U R A T E A S Y O U T H I N K.

Peter Collins

Yuri Gagarin is not as accurate as I think?

Kryten

I don't quite understand RicF's post but I think he is agreeing with Toucs and me.

Was Yuri Gagarin in the Reading half then? I bet he had to run further than 13.1 miles.

Peter Collins

He'd have difficulty running at all these days

Wilkie

He went considerably further than 13.1 miles - but he didn't have a Garmin so didn't know when he should stop.

PhilPub

Is he the bloke who plays the fiddle?

Toucs

Perhaps he's the one to blame for putting a faulty tracker system in the sky? 

I like my Garmin because it always says I've run longer than the actual event I've been in.......for half marathons and above at least half a mile.  London Marathon two years ago it was just over 27 miles.  I always think I get my money's worth though!

Booo

So then what you're saying is that Garmins could be out by up to 5% ?

[Deleted User]

Wow!! Bruce, that's some speedy movement there. What were you doing? I assume not having a shower! Well, not unless the earth moved for you

No, we said that the manufacturer claims a tolerance of a few percent. If you have reason to believe that you've used it in accordance wth instructions and it is showing something more than that, you have a case that it's not fit for purpose. It's not an upper limit of tolerance, that's a different thing entirely.

And, while we're at it, and putting my scientist coat on, there is a tendancy to confuse accuracy and precision. They're not the same thing. It's is possible to be accurate and imprecise, as well as precise and inaccurate. The holy grail of measurement systems is accurate AND precise.

RicF

And a Garmin GPS is thereby not the Holy Grail of measurement systems. It is however an excellent gadget. For something.

dhale750

Happychap wrote (see)

In that case Dhale I must have deviated a helluva long way in VLM last year. The watch clocked 27.59. Running bench it would make you faster though as most measurers jones counters are attached to bikes <insert smily>

Plausible enough, especially in a race as crowded as VLM.

 http://www.dcrainmaker.com/2009/03/racing-line-understanding-how-courses.html

[Deleted User]

When I'm sat on my sofa with my Garmin in my hand displaying the map the marker moves all over the place. Mine and Wilkie's must be from the same batch!

Cheerful Dave

dhale750 wrote (see)

Happychap wrote (see)

In that case Dhale I must have deviated a helluva long way in VLM last year. The watch clocked 27.59. Running bench it would make you faster though as most measurers jones counters are attached to bikes <insert smily>

Plausible enough, especially in a race as crowded as VLM.

 http://www.dcrainmaker.com/2009/03/racing-line-understanding-how-courses.html

Interesting article, but he never mentions that the GPS data points themselves have an in-built error, before any of the errors in interpolation (which he does mention at the end of the article, to be fair). 

That course he uses has a lot more turns than the London course, and even taking every one of those turns on the outside of the bend he only gets an extra half mile.

If someone's getting an extra mile and a half at VLM, it's mostly down to GPS errors & lost signals rather than deviating from the blue line.

PhilPub

I agree with Dave.  You'd have to be running like a drunken idiot to add more than a mile to a marathon route.  I'm sure I said it earlier (or it might have been in another thread, CNBA checking) but the Garmin invariably and systematically overestimates distance when running round a track and I don't remember weaving in and out of runners there, or diverting to take on water.

Cheerful Dave

You've not seen HC run then?

RicF

How many would swear blind that the running track must be wrong because their Garmin says something different?

RicF

To be honest, I really like the principle behind these devices, but having field tested a Garmin at least ten years ago against a Nike SDM, the Nike won hands down. The signals sent to the Garmin are simply not up to the job and to the nearest 10m is also nowhere good enough. When the signal is pinned down to the nearest 500mm, the signal transmitted at 50Hz plus, and strong enough to penetrate buildings and trees, then it might be worthwhile.

Happychap

LOL CD

Actually because I literally started at the back last year and pretty much stayed there. I was able to follow the line more closely than ever before. And it was the longest distance the Garmin clocked over that course.

Screamapillar

I think the bottom line is that if elite runners (whose livelihoods depend on their times) trust that course distances are correct then so should everyone else.

BirminghamOwl

Sorry if this has been posted before.

Have a read of this article - http://www.goodrunguide.co.uk/MeasuringRoutes.asp

Great example of how inaccurate GPS can be just around a track.

Owl