I started using TR 2 weeks ago and really like it, but one thing confuses me. I didn’t track my power before using TR and I just used to do GCN YouTube workouts, which seemed focused mainly on cadence and perceived exertion. However, my average speed was higher when I did the GCN workouts, even though I feel like I’m working my ass off way more now with TR.
Does anyone have any thoughts about this? I hope I’m doing my TR workouts correctly, maybe it’s because I’m still in the base phase?
Are you using a power meter? If you are then just ignore the speed data. Speed data is essentially meaningless - it just tells you how fast your flywheel was turning, but that will depend on which gear you’re in - by which I mean that 200W in your highest gear will be much quicker than 200W in your lowest gear even though the power is the same.
I’m using virtual power with speed and cadence sensors. That makes sense, thanks.
Agreed. Speed is meaningless on the trainer. For a non-smart trainer when you’re using virtual power it takes a mix of the speed at a particular friction setting though. So just set a friction level which will mean that you can ride at the required intensity on a gearing which means that you don’t end up with your back wheel doing 50,000rpm as that achieves nothing except noise and tyre wear.
When training indoors you can safely ignore speed and distance metrics.
Focus on virtual power and maintain consistency with your set up as much as possible.
Zwift & Strava didn’t get the memo. 
Zwift is understandable given they are simulating mileage but strava I think are probably just lazy in showing you pointless metrics.
Strava is merely reporting data as received. Any fault lies in the reporting app.
It’s really the grade that matters and Zwift has a grade and speed is governed by gradient. 300 watts at 0% will have a higher speed than 300 watts at 3%. That’s what trainerroad cannot replicate visually, which is why some of us try to choose a gearing that might simulate a flat-ish rolling grade