I’m curious what other people think about the calculated kj expenditure, and therefore calorie expenditure, when riding in erg mode. I’m curious if the natural smoothing of the resistance when sitting at a fixed wattage actually has a little room for over estimating power, I can easily take a half pedal stroke easy without registering a dip in power, for example.
Cruising at 250 watts for 90 minutes being 1400kjs expenditure, even if I’m consuming 120g of carbs an hour I’m still generating nearly a deficit of nearly 500 calories in an hour, I’m not sure I trust it
If you’re on a Wahoo trainer, turn off ERG Mode Power Smoothing. Once you do, you get more real looking power instead of the insane smoothing they apply. I would guess it’s more accurate overall (including kj) but I can’t say for sure.
My Wattbike atom doesn’t smooth in the same way. I see frequent fluctuations in power in real time, and even trying hard to keep what (I think) is a consistent smooth technique reveals many small changes in power, in particular if cadence fluctuates too much.
I actually find ERG mode significantly harder in terms of fatigue and stress than riding in non-erg mode or outdoors. It feels like outdoors I get numerous micro breaks, even when going uphill on long climbs.
Whereas in ERG there is nowhere to hide. Coupled with my inability to hold the same power indoors as I can outdoors (cooling I think makes up the majority of the reason for this) and I feel like my indoor workouts are quite a bit tougher than an outdoor equivalent.
To my mind this is a right-side failure and how I’d prefer it, given the flip side would be easier indoor efforts that I then can’t replicate outdoors when I’m attempting to pace.
This is all to say that my guess is, a lot depends on the setup you are using, as to whether this is a material issue for you or not.
I might be wrong. But doesn’t the TR app still use the raw data at the end to calculate the power and is the smoothing just for show?
Even if it’s not. Without deliberate sabotaging the workout, i think the estimates are pretty close. That half pedal stroke easy, works both ways. You could easily do a bit more power without even realizing. After 90 minutes i think it will balance itself quite nicely .
So experiment in erg mode with smoothing turned off on the kickr, you can see the much higher variance in the power output but the kjs ends up being spot on to the plan
Power smoothing is really only used to clean up the visuals of power data. It’s kind of like 3s avg vs. instantaneous power in a way.
Your average power numbers and the data are still all the same, so your kJs aren’t going to change with this setting on or off.
Kilojoules are simply calculated from power and time (Energy (kJ) = Power (W) × Time (s) / 1000) so the work done based on your power numbers is going to equate to the same amount of kJs no matter how you look at it. Power smoothing just makes the data look less erratic which most people prefer.
Taking the thread on a slight tangent, I actually really like seeing the raw data because it really shows how my technique evolves over an interval session.
As for the expenditures, they’re the same, as everyone says, so more carbs for me!
This is true for TrainerRoads smoothing; but I believe the OP originally had Wahoo’s (terrible) ERG smoothing on which produces perfectly flat lines.
Wahoos smoothing actually changes the data sent to TrainerRoad and in that case the accuracy of the various calculations could be affected - but I would have thought the impact would be pretty small unless you actively try and trip it up by doing micro coasting for excample.
Like others have said and as I’ve always understood from reading countless articles and forums, power smoothing (and wahoos in particular as I own a CORE and was always curious myself) does little more than make the power graph flat….
@GPLama once again doing the legwork so we dont have to
1min in he says something like it “reports what you are setting the trainer to not what you are really doing but it will be close”.
I’m not saying it’s a big issue - but I thought I’d just call out the difference between TR smoothing which leaves the raw data intact and wahoos smoothing that alters the data sent.
