Just How Far Off Is Virtual Power on Rollers

Great point and one I considered. I am pretty heavy so the usual 100psi might not play well for me. I though I was using 100psi constantly but perhaps I was a little higher before and the last workout was in actuality lower than previous.

Should probably write it down.

My bikes pretty heavy too at 11kg.

I think I will need to do the ftp test, and most likely the 20 minute one.

That being said, I’m building up an allez and it’s likely to be around the 9kg mark so would it be best to wait and test on that?


Really appreciate the help.

A power meter is deffo on my list now. I see 4iiii aren’t too expensive.

1 Like

To close the loop on this, the power curve for the Elite Arian rollers is direct from Elite so there is no data on what they used to build the curve.

I’ve seen 4iiii power meter is nit badly priced so I’ll look into that

1 Like

I dont think my last response fully closed the loop so i will pop a few more updates here to help in the future if anyone else has similar questions.

I now have a 4iiii left crank power meter.

I’ve just set it up and ensure i can calibrate and connect it up to Trainerroad and my Garmin.

I rode for a few mins just to make sure its all working and sending data to both devices (albeit in flip flop/slides).

Initial thoughts - one day i’ll get a dual sided but for now this will do lol.

I spun the legs up in a known gear up to a known cadence.

On Sunday 93-95 rpm had me at 140-145w using virtual power. Interestingly, with the power meter, 75rpm had me at about 140-145w. And 80 rpm has me at about 160w. I’ve been riding Z2 at about 85-90rpm for 90-120 mins for the past few months, so i was actually putting out 20 odd watts more than i thought i was.

I moved to another gear, again, fairly well known from previous sessions. A few days ago, 95rpm had me at about 200w with virtual power. I didn’t go up to 95rpm with the power meter (flip flops) but low 80s rpm had me at 225 (my currently set FTP).

Its not the quite a scientific test and tonight’s workout will give a better idea of the gap but overall it looks like for me (and my weight, bike weight, tyre choice and tyre pressure), virtual power has been under reading a bit.

I’m not sure just how close to the correct training zones i have been, given its all relative but its nice to have something that’s potentially a little more accurate.

(Or maybe the flip flops have massively skewed the results, or i calibrated wrong etc)


It seems virtual power was massively under representing my power on the bike, although now i am questioning the 4iiiis power meters accuracy.

Todays session was Seneca Rocks, 3x15 mins at 85%. Since i dont know what my actual FTP is (and i didnt feel like a 20 min test) i went off feel and decided to see where the power ended up.

All 3 intervals at 240w, and not overly difficult. HR at about 80-85% of max.
When i was doing 3x12 at 85% based on virtual power it was much harder. Hr was was up at lactate threshold (90% ish max HR) and it was a challenge to finish the interval.

With the power meter now reading 240w it felt like i could’ve gone for 20, 30 maybe even 40 mins if i really needed to.

Surely my FTP cant be that high. I’ve never had it over 240w (direto smart trainer).

It feels like i’ve just replaced one question with another.

Thursday is a Threshold session, i fear i may need to just suck it up and do the 20mins and see. Whether its over reading or not, it’ll give me a baseline.

By adding a power meter into the equation, you should really just do a new FTP test of your choice. We see similar issues when people swap from smart trainer data to power meter data and any combination of those mixed.

As much as it would be nice to have, power data is not created equal. There are countless threads here with examples of mismatches in data between any combination of two devices. Add in the fact that you are getting “single-leg doubled data” now as compared to “whole virtual power”, and you have more variables that can’t be reconciled.

Do an FTP test (as you mentioned) and move forward with your new power meter data as the “tape measure” and standard data. Also consider creating a new season as the transition between your prior virtual power and your new real power data set.

If you are questioning the 4iiii data, you should contact them for support to see what they can do to help you verify.

1 Like

Cheers @mcneese.chad

Do you mean restart SSB lv?

I’m just running the plan as I have no target race

  • No, don’t restart your plan.

Use the “Seasons” tools as covered here:

It’s not essential, but since you are swapping from power data A to power data B, your personal records from one may or may not relate to the other. If you aim to track your PR’s over time, using the new power data as the foundation is what you get by starting a new season.

1 Like


1 Like