I’m seeing large differences between Garmin & Trainerroad’s average power data.
Garmin Data: 229W Average
Trainerroad Data: 190W Average
That’s a 39W difference. The data file is basically useless if it isn’t accurate. Trainerroad is following the Strava data and therefore and suggestions pulled from the data are based on inaccurate data.
Has anyone come across this? It seems like it has to do with how garmin is able to make use of the accumulated power data in the FIT file but Trainerroad and Strava seem to ignore it (basically power applied without a full pedal stroke). It’s not a big deal with steady state efforts but in Cyclocross and MTB this causes big differences.
Equipment:
Garmin 840 Solar - set to include zeros
Stages Gen 3 R8100 Dual-side - gyro setting on
Edit:
Garmin and Strava show (857KJ)
Did some math 857 kJ / (1:02:38 = 1.0439 hours) / 3.6 = 228W
Strava shows 190W and 857KJ, LOL
For the record, Trainingpeaks shows 229W and 857KJ
Are you sure you don’t have Garmin set to exclude zero power? That’s typically the case when you see a discrepancy like that, especially between Garmin and Strava/TR.
Edit just reread and saw you said you are including zeroes. Not sure what else, that’s way off
To clarify, you’re saying you record a ride on your Garmin, and when it uploads to TrainerRoad it says 190W average power, but on Garmin Connect it says 229W average? Should be the same file so I don’t know how it differs that much. Try downloading the FIT file from Garmin Connect and upload it to another service like intervals or TrainingPeaks. If the uploaded file is saying 190W average, then you have a setting somewhere on Garmin that’s removing the zeroes.
Another idea is use a smart trainer and do an erg workout at like 150W. If they give you the same data from that, then it’s definitely removing the zero power readings because erg you wouldn’t have any zeroes.
A random guess, there’s a dodgy peak or two (max say 3,000w) that’s putting the average up in GC but TR is automatically correcting that stupid peak so the average is lower.
That is correct. I’ve tried downloading the FIT file and manually uploading it and the same thing happens. I am 100% sure that zeros are being included on the Garmin.
To your point, for steady state efforts like ERG mode it does give virtually identical numbers. However, it’s not absurdly low like this recent file. The one I referenced was a CX race where 229W is the expected level. I was certainly not at 190W.
Just curious, are the power zones similar with similar time in zone between the platforms?
Also, I assume you’re using the watts reported by each platform, not converting from kjs into watts. Every platform calculates kjs with a different multiplier.
This is incorrect. Different platforms calculate kcal differently because they assume a different efficiency. Watts to kJ is a defined conversion factor because the units of a Watt is joules per second, ie, 1 W = 1 J/s.
The underlying data for an activity is the same irrespective of summarizations performed by an app, like calculating averages, NP, Weight Avg Power (STRAVA), TSS, IF, etc… I expect (hope) that TR utilizes the underlying data and does not rely on another app’s calculations and summarizations.
