Misleading elevation gain on MapMyRide

Yes, every time a route I know is posted it seems to be well under. It's good enough to give a decent idea of the more major climbing, but it misses so much of the smaller ones. In that regard it's not useless, but it is important to understand its limitations.

What did you use to measure? Garmins tend to do a lot better, but what device is used makes a difference. GPS based ones are noticeably less accurate than the barometric ones.

Yes, i've seen it wildly exaggerate the vert also.

Seems ok as I click the waypoints, but then something goes awry when i am done

Straight garmin numbers are just as inaccurate.  I find the only way i get an accurate picture is by using the Elevation Correction plugin for Sporttracks.  It uses 3-4 sources for the elevation, including the NED (national elevation database) if you choose to download it.  You can then compare the 3-4 sources of elevation and the NED database always cooresponds the closest to the course I just ran/rode.

Of course this doesn't help when using mapmyride but is good post ride.  I have found that the profile is "generally" accurate for mapmyride, the total accent/decent is not accurate.

MMR is the worst of all the programs IMO. It has one hill that I climb where if I get off the bike I would slide down a good ways....on by butt too. MMR has it at 5%

After this I never used it again. Also keep in mind that you will never know the correct elevation for your ride. Garmin's w elevation correction (no barometer) you would think are good as they are supposedly comapring them to some topographical survey. With baramoter they can also be off as weather changes. Strava, ride with GPS, etc

Just take them as a ball park and fdont get too caught up in them but definetly dont bother with elevation from MMR

mapmyride is all about sampling size/frequency.   If you map out a longer ride, it averages out the elevation a lot, so you don't see the climbs as steep or tall as they might be.  You can see it for yourself if you go map out a big hill that you know of, and just map the hill.   You'll get a more accurate assessment of gradients over just the hill, rather than if you average in a few miles ahead of and/or after the climb.  A 250 foot section of 20% grade disappears quite quickly if your slope includes 10 miles of road.

On days I have way too much free time, I've mapped things out mile by mile (or only a few miles at a time) to get a more accurate idea of total climbing, but that's sort of a pain, and then you don't get to see the big picture quite as well.

We had two Garmin Edge 800s on this ride, both of which measured 3600ft.

I find garmin's barometric altimiter to be pretty accurate.  I've done the same ride many times, on different days, with different weather conditions, and the elevation gain always seems to be within 1-3% of each other.  That's good enough for me.  I'm not exactly sure why I would need a super precise measurment of elevation gain.  I don't see how that would improve my riding.

All of the uphill goes under gain. All of the downhill goes under loss. Added together it should be near zero for a route that ends where it starts, this is net elevation.