.fit files sent from peloton have some formatting problems. Some apps/sites won’t import them at all. I’ve used the RunGap app on my iPhone and it seems to fix the formatting when it syncs files. Actually haven’t used it in a while cause I ride my bike on a kickr core most of the time. I think there are other work-around s but I can’t remember them as RunGap works really well and easily. It’s a paid app but I think only $10/year
I am Peloton user also. I can sync Peloton data to Strava, and then use RunGap to sync from Strava to TrainingPeaks and the data seems to be accurate. Second RunGap works well and easy.
I have been experiencing a similar issue. I don’t do Peloton rides often but my girlfriend has a Peloton bike and if the TR workout prescribed isn’t too complicated (endurance or something like sweet spot intervals) then I have no issues doing that on the Peloton Bike.
Questions about accuracy of the power reading aside if I do a ride and match the power numbers as close as possible my TSS is always ~2/3 of the anticipated number the workout gives. A recent example I have
I think this more or less the issue other people are reporting. Any resolution?
For starters, notice that the FTP is different (245w top pic / 250w bottom pic) on each of those screens, so I suspect that is a problem you need to rectify. TSS is based upon the load, and uses your FTP within the related calculations, so if your FTP doesn’t match, that will affect the results.
1 Like
The IF and TSS on the completed workout do not add up. The formula is:
TSS = IF x IF x duration x 100
And using IF from your May 13th screenshot:
TSS = 0.83 x 0.83 x 1.5 x 100
TSS = 103
But TR shows TSS = 57,
Like the first post stated, it appears TR calculation for IF is too high, and TR calculation for TSS is too low.
Send email to support@trainerroad.com
This sounded familiar - I had an ongoing, over a year exchange with TR to figure this out when I was traveling heavily and using a peloton bike in a hotel.
Yes, Peloton’s are notoriously out of calibration, but I think that is a red herring for the issue of TSS being lower in TR after being imported from Strava. I used HR and RPE to get the Peloton into the ballpark, regardless of the watts.
From 3 years ago, what I discerned from TR support was the issue of low TSS could be how (in)frequently Peloton measures power. I believe they said every other second. And when I clicked into the TR power curve that seemed to check out. Why does it look normal in Strava? TR said that Strava likely fills in missing data…
My apologies: this was a ride taken before I updated my FTP. I snapped the picture of the TR workout today just so I could share a reference. But good point that the expected tss will be slightly less when I was at 245. But not by that much.
2 Likes
Yeah, not likely to “fix” the whole issue. As others covered, there is more here but it’s useful to make sure FTP’s match whenever heading down this road.
No. And I can’t believe it’s the TrainerRoad approved answer.
The expected TSS of that workout is independent of FTP.
@IvyAudrain ^^^ The TSS of Iron Crag will be 76 if your FTP is 100 or 200 or 300. TSS is intensity factor squared multiplied by duration in hours multiplied by 100.
The total work in kJ will change based on FTP.
The TrainerRoad calculation of TSS for the completed workout does not add up. Something is wrong.
The three numbers above do not add up. The TR IF looks too high, using that to calculate TSS and you get TSS = 103. Using TR TSS = 57 gives an IF of 0.62. Something is wrong.
1 Like
This is a consequence of incomplete data being sent from Peloton.
To give you an idea, heres what the first 30 seconds of the ride file in CSV format. As you can see, we are missing ride data for every other tick (second):
That missing data is what’s causing TSS and IF values you’re seeing.
For example, in this particular situation there are 2940 seconds in the file that have ride data, which converts to 0.82 hours of total “ride time.”
If you plug that value into the TSS calculation, you can see how the formula arrived at 57 TSS.
TSS = 0.83(IF) ^2 x .82 (hours) x 100 = 57
Essentially, the missing ticks are misleading the calculations into thinking the ride was only 0.81 hours instead of the full 1.5 hours.
5 Likes
That makes sense and tracks what I see in intervals.icu
It seems Strava fills in the missing ticks with “mock”data while TR doesn’t?
If that’s the case can I make a feature request for TR to do that? I understand mocking data gets a little dicey but I think it can be done without making the ride data too crazy.
Edit: I also understand this might not be a high priority as I doubt many rides get synced from Peloton.
1 Like
You’re right that it doesn’t impact a significant number of athletes right now.
We’ve looked into this and the change is pretty involved, so if it applies to more athletes in the future, we can possibly revisit prioritizing it.
1 Like
What does Intervals give for TSS and IF?
@IvyAudrain I know you are the messenger, but take it from someone that’s worked with sampled data for over 35 years. There are 2940 SAMPLES of data (not seconds of data), and of course that impacts how the normalized power code needs to be written. 
I’m not saying that I might have mentioned this above, but I am not not saying that… 
Thanks for the validation @IvyAudrain !!
1 Like
100% fair. Thanks for the response!
1 Like
So I got COVID so I haven’t ridden in two weeks until today. Did a 30 minute structured class geared for recovery on my girlfriend’s Peloton. Check the data in intervals.icu and I see this:
or rather I no longer see the “This activity has been recorded with approximately 2 seconds per data point” message that I cited here:
Did Peloton change how they are reporting rides? Did intervals.icu do something different?
Side note: COVID sucked in terms of sucking fitness out of me. Hopefully it doesn’t take too long to get it back.