Originally Posted by ProLiteNews

Depends if you have calibrated your computer properly.

If your computer is properly calibrated, I would trust it over a Garmin any day of the week.

To calibrate a computer, mark the ground where the valve is, then ride in a straight line and mark again at the point where the valve meets the ground. Stay weighted on the bike. Now measure the distance between the two points in mm. There should be a setting in the computer where you can tell it the distance.

I always check my computer afterwards on known and marked distances. I usually set up at a track and ride the white line straight around. When properly calibrated, I can do 20 laps with very little deviation. If you do a 1km track with less than 2m of deviation per lap, you are under 0.2% error level.

The computer works very simply. It just adds the distance represented by one revolution every time it receives a signal.

While it is somewhat prone to error over very long distances, if you set it up properly, it should be very accurate, within a fraction of a percent.

The Garmin will be more accurate for very long distances, and might even be able to work with plotted non-linear courses, but won't be as accurate with hilly courses.

My computer is usually within 0.2% accurate. Over 70 km, that's an inaccuracy margin of around 140m. That's around 0.1km. Your discrepancy is 1.92km, which is 20 times greater.

My father is a cartographer and helped develop some of the systems involved with mapping and GPS. We're a bit nutty for accuracy on stuff like this.