When I broke a finger and was forced to ride on the trainer last year I used an old wired Sigma computer to get the speed off of my rear wheel and (logically) as long as I was in the same gear and the resistance on the wheel was the same, it was a pretty good proxy of power.
The "data" it provided was very useful during a single training. As you progress through identical intervals, while maintaining the same "power", heart rate would go up ever so slightly on each interval.
However, I highly doubt that the actual readings were consistent from day to day, as the "power" reading depends on a lot of variables, including but not limited to:
- tyre pressure and temperature
- actual resistance of the trainer (I used a Tacx Satori, and because the resistance setting is cable-actuated, I doubt you could actually achieve the exact same resistance every day, unless you would completely fix the cable in a single position)
- temperature of the resistance unit (influences resistance on pretty much all trainers)
- room temperature, humidity, atmospheric pressure (which also influences your own performance, so you would need more or less actual watts to obtain the same "speed" on the rear wheel in a certain resistance)
Therefore, at least in my case, it was next to impossible to really track progress based on these "watts".
So although it seems to be a good concept, how reliable are these virtual power measures over the long term?