tl;dr: 'did science' and learned way too much about the garmin vector. my coworkers also now have proof postive i'm insane

Spent several hours over the last two days 'calibrating' my Garmin Vector power meter after physics suggested it was reading low (312W for something that should be 330-340W, as one example). Using the recommended procedure, I brought my bike to the office and set it up on a trainer between two standing desks (which were conveniently adjustable allowing for me to get everything completely level) and weighed my weight/string/hook on a calibrated scale to figure out what value I was looking for (27.84Nm). The pedals were reading low 27's, which wasn't quite the ~10% I was looking for and were at the low end of the +- 2% that the power meter was supposed to be accurate to. I then tried everything I possibly could to get the pedals to read the 'correct' value (though if I'm being honest, if I got them to instead report at the high end of the 2% range I would still have been happy with that...).

Doing this experiment in a building full of engineers meant I had lots of curious onlookers throughout the process with a number of suggested variables to try. I tried altering:

- torque on the pedals (25-30 ft-lbs is what is recommended by Garmin - I tried at pretty much every ft-lb increment)
- the number of spacers between the pod and the crank
- amount of grease on the pedals
- static calibration angle
- pod angle (not supposed to matter)

and several other things I'm now forgetting. Each setup required dragging the bike outside and putting on my bike shoes to do a few hard sprints to settle the pedals and then allow the garmin to figure out the installation angles, followed by numerous weighings on each pedal with numerous static calibrations in between. At one point I had the right pedal pretty much perfect (27.65-27.95 Nm no matter the static calibration) but nothing I did managed to bring up the left pedal from 27.2 (27.0-27.4). After spending way too much time on it I gave up and decided to do a clean install of both pedals and to just accept whatever values I got from it. Somehow, the ultimatum worked and I ended up getting it pretty much on the nose which was rather amusing... I'm looking forward to 2-3 rides at the correct calibration before something 'slips' and its back to crap values again.

The frustrating part of this 'calibration' process is that its not so much 'calibration' as 'continually testing and modifying until the test reports the right values'. It seems like I should be able to enter what the expected torque value should be to calibrate the meter (like you can with known GPS elevation, for example) which would have made the whole process take like 15 minutes. Either way, I'm going to continue to not recommend the vectors to anyone interested in power meters - the powertap hub on my other bike is much less stressful.