How close does TR virtual power correlate with actual power?

ftp

#1

My only experience of using power on the bike has been TR virtual power (with a basic Elite Crono turbo). I’m curious - how accurate is my TR virtual FTP likely to be in the real world when using an actual power meter? For those of you who have used both virtual power and a power meter, have you found your FTP numbers are pretty close?


#2

The short answer is that it varies. Slight changes in roller pressure, tire pressure, even the actual tire used will all impact “accuracy”.

Sometimes it can be dead on accurate, other times way off. The take away is to not assume it is more than a measurement device for you. Don’t compare FTP to others.

Make sure you are very consistent everytime you set up the trainer to do a workout. Get everything the same, and you will be able to do workouts consistently as well as measure your own progress.


#3

To give you a real world example, I have a Kurt Kinetic Rock and Roll ‘dumb’ turbo. One of my bikes has a Power2Max power meter and I have used this to compare the ‘true’ power recorded on the P2Max, against the virtual power indicted by TR, across a variety of different power outputs from 150W to 350W.
Within this power range, the virtual TR power reading was consistently 30W higher than the true power reading on the P2Max. The important thing here is that the 30W offset was consistent across the power range that I use and that it was reproducible from one ride to the next, as long as I used the same tyre pressure and applied the same number of turns to the tyre clamping mechanism on the turbo. This means that I can get a pretty good indication of my true power when I use other bikes on the turbo that don’t have a power meter fitted, simply by subtracting 30W from the virtual TR power.
Obviously, this 30W offset is specific to my turbo, power meter and bike, but in principle the same method could be applied to any other set up. However, I suspect that the Rock and Roll has a top quality fluid resistance unit and therefor has a very predictable resistance curve that the TR virtual power algorithms can use. It may well be the case that cheaper trainers do not give such consistent resistance curves and are impacted significantly by external factors, notably the turbo resistance mechanism heating up during a ride.