What are your GPS/Bike computers reading as distance?

[indymtb]

Not that this has any bearing in terms of the trail but I am curious what people generally measure for 1 lap at town run (From entrance to shelter)?

Just for the record -
My Garmin 705 says it is about 6.9 miles.
My 920XT thinks it is about 6.75 miles
and my Fenix 3 thinks its 6.6 miles.

The listed distance is 6.6.  I am guessing that most would register between 6.75 and 6.9.  Any insight would be appreciated.

[K_W]

I've been wondering the same, I never seem to get accurate distance with a cycle computer, compared to posted trail lengths.  To actually check accuracy I'm going to have to ride from one mile marker sign to another

[Dave Tozer]

"Hardwired" bike computers typically show 7.2 miles, give or take when people put in the correct wheel size in the setup. That's what I consider the full lap to be.

My Garmins (510 and older models) have consistently yielded 6-something. From 6.2 to 6.9.

Not long ago my 510 read 7.2 on the nose. I was floored.  I have had Garmin wheel sensors for years but the Garmin units apparently still totals the mileage from what the GPS can identify. Many tight corners can't be captured and look to the GPS like a straight line.  The same Garmin on an open road with no tree canopy will give a much more accurate reading.  When I ride the Monon and I go from a wide open area with no trees into a section with a dense tree canopy my speed reading immediately goes down.

Getting back to my mentioning my Garmin reading 7.2 miles at Town Run - the Garmin needed a full reset to factory defaults a while ago.  I normally do a wheel roll out to get an accurate wheel size and enter the number in the "custom wheel size" section of the Garmin. But before I left my house to ride to Town Run one time, I heard the Garmin beep as I was riding down 116th street and could see that the Garmin was saying that the automatic wheel size was just calculated. After a lap at Town Run I noticed the distance was 7.2 miles. So, I guess Garmin's automatic wheel size calculation is the way go.

As a side note for those who look at the Town Run Strava segment, keep in mind that the most popular Strava segment (the one pretty much everyone uses) is short. The segment doesn't start until you are well into the woods and on the return the segment ends well before you get to the shelter. I don't care about that, but many do. I'm not a Strava hater, either. I uploaded once to the Town Run Strava segment just to see how it works, going from a Garmin up to Strava.

All I really hope for is a Garmin unit that will accurately measure my distances. I like to know how many miles I ride in a year.

On another note, my 510 "bricked" a few weeks ago. Garmin tech support confirmed that it is shot. That's so frustrating. A little over 2 years old. It was not cheap. They'll give me 20% off a new unit or give me a refurbished 510.

[jasonhilt]

I consistently get 7.4 on my wired odometer at Town run.  Many people I have talked to also get that distance.
GPS units will almost always under report the distance.  They don't work well on lots of curves as mentioned before.
 

[suboptimal]

Interesting article about GPS distance overestimates here:


http://www.i-programmer.info/news/145-mapping-a-gis/9164-gps-always-over-estimates-distances.html

[Mahk]

I understand the conclusion of the article:

The only displacements that reduce the distance are those that move the two points closer together, for example displacements along the line towards each other. The majority of random displacements end up increasing the distance.

This is the reason that unbiased errors end up biasing the distance measurement.

So given that the GPS path is just a sum of distances computed between pairs of points, the total estimated distance is going to be bigger than the true distance because of random errors.

But, in the discussion I don't understand this statement:

...the expected measured length squared between two points is bigger than the true length squared unless the errors at both points are identical.

Why are the lengths squared?

[suboptimal]

I am guessing that coordinates are returned as longitude and latitude which would translate into x/y position coordinates.  So the distance traveled is the square root of the quantity ((x2-x1)^2 + (y2-y1)^2).  That quantity is the distance squared and is probably what they are talking about..


To me, it made better sense when I thought about two points that represent the true location of the points being used to calculate the distance.  Then you can draw circles around those points that have a radius that is the magnitude of the position uncertainty.  If you strike an arc from one true point to the other that is the true length traveled, it will intersect the uncertainty circle at two points.  The fraction of the uncertainty circle that is inside that arc is significantly less than the fraction that is a greater distance from the first point.  So if the uncertainty is truly random, the chance of the calculated distance being greater than the true distance is higher.

[mtbikernate]

Don't forget that the positions are in 3 dimensional space. a^2+b^2=c^2

[SurlyWill]

After the later Garmin auto calibrate, they will use wheel sensor data to calculate speed and distance.

FYI, you can always define your own segments and Strava should default to those.