Forerunner 945 giving wrong distance readings during a run, always short.

Rule of thumb of GPS distance measurement, those which measure more distance are most often more wrong.

So based on that evidence I would need more facts to believe that 945 is measuring wrong distance.

How about drawing the route as a course and measuring that? What does that give you. Granted 0.2 miles of error on 2 miles is pretty big 10%, but still, measuring too short isn't usually the case.

jmto said:

Rule of thumb of GPS distance measurement, those which measure more distance are most often more wrong.

Never heard of such a 'rule of thumb'.  Where did you hear this?  I find it odd because 9 of every 10 postings about GPS distance is it came up short.  Short or long, wrong is wrong.  More wrong?  There's no such thing. Does that make me less wrong or more right for pointing that out? :) 

Compare the tracks from different devices. What are the differences?

tess said:

Never heard of such a 'rule of thumb'. 

Now you have. You're welcome.

Yes, if we make people run eg. marathon which is officially measured the more your GPS is saying, the more wrong it is. It's much much rare that you would get GPS distance to be under the real marathon distance. If that happens you can think was the distance measured and marked correctly and did you run it right.

And most telling they are getting it short is comparing it to some other devices and that compare really can't tell that your distance is too short. 

Logic is how GPS measurement works and the distance comes from there. If you are now thinking something if I got 1,8 miles, real world is 2 miles and someone got 2,1 miles, how 1,8 miles is less wrong? Well it isn't but that really doesn't much happen with GPS. You get extra meters because it's not accurate and jumps around, it's really rate it jumps so that it will shorten your distance.

Like one time the local half marathon was marked wrongly. people were "wtf, my GPS showed under the real distance", they started giving excuses GPS is not to be trusted it is measured correctly.. Well people were like it doesn't really usually work like that and the reports of under distance was a lot.. Then later they noticed they had marked the course wrongly and it really wasn't the real distance.

So even if there's no official measurement, that rule of thumb just works. Most accurate GPS has less errors and gets less meters.

And if you want some source, quick Google gives you articles like this: https://spectrum.ieee.org/tech-talk/at-work/test-and-measurement/why_every_gps_overestimates_distance_traveled 

Waddyuk said:

What can we do?

Measure the distance with a calibrated wheel and see which device is closest?