Originally Posted by
'nother
Note that measuring "total" or "accumulated" climbing is non-deterministic. That is to say, it is impossible to do "accurately".
You will rarely find exact agreement on total climbing numbers, even with two identical devices on the same course. The problem is sampling rate versus position & velocity. Think about a course with a 10-foot hill. If your altimeter happens to take a sample when you are exactly at the base of the hill, and another one exactly at the crest of the hill, it will register a 10-foot gain...great, that's what we want (sort of).
That is true, but two identical devices will usually give you cumulative elevation results within a couple of percent of each other. Even altimeters from different brands will usually give numbers within 10% of each other on a long bike ride.
On the other hand, numbers from web sites can vary by 100% or more, making them pretty useless. I wouldn't even trust web site numbers to compare 2 different rides on the same web site.