Very true, but the same principle holds true for small gears, also.
Ok, thanks, I think I understand the original question.
IRL same gear, same cadence, speed will be same but power out put will be increased.
I’m not sure why it is different in Zwift.
I’m not sure it’s a Zwift thing vs a trainer thing.
Zwift may overestimate the power required to ride up hill.
Also, is your wt set right?
Hi all,
Thanks for all the replies, It’s an interesting one!
I have had a few mins to test out Nigel and Steve’s suggestion of changing the Trainer Resistance to 100% and this has solved my puzzle. At a cadence of 85rpm whether going on the flat or up an incline (I only had time to get to a couple of 1% ones but that was enough), if I don’t change gear and put in more power to keep the cadence at 85rpm, I do indeed stay at the same speed as I thought I should.
With the TR at 50%, as Steve and Nigel say, it “waters down” the representation of the inclines compared to real life, so the power I was putting in to stay at the same cadence in the game was only 50% of what I would have really needed to put in to stay at that cadence in reality, and hence to stay at my original speed.
@Tim_Camden_C yes you got the question right, the answer looks like it’s the Trainer Resistance setting in Settings having been set at the default 50%. Changing to 100% makes it behave as it should. And my weight is roughly right, but the power required doesn’t actually matter at all, see Nigel’s description a few replies up about distance, gear inches etc.
For reference, I noticed this on the R.G.V. ride in France. I didn’t have time to get very far through it to the 2-4% sections, but the 1% ones seemed to behave correctly with TR at 100%.
So thanks @Nigel_Tufnel and @Steve_Hammatt for your suggestion and resolution of this issue.
And thanks to everyone who’s contributed, I’m a happier rider, at least in part … but now all those inclines are twice as hard with TR at 100% 
Cheers all!
2 Likes
Looks like your question has been answered. You next question will be , why does the bike travel slower on different road surfaces in game with the same gear with the same cadence.
1 Like
Note that your expectations will only be met on non-negative grades. Even at 100% trainer difficulty, Zwift models downhill slopes at 50% of the actual grade, and once you do any coasting, all bets are off.
1 Like
@Lawrence_F one more thing I thought of today that I wanted to add to the discussion, before you come back here with a follow up WTF? question:
There is one more parameter related to Zwift speed that we didn’t really touch on here: crr (coefficient of rolling resistance). The discussion above holds true if you are always on the same road surface. However, each different road surface in Zwift has its very own crr, which also affects the speed of your avatar. Most places you won’t likely notice this, since the road surface stays the same for a fair amount of time (but you may grow frustrated with the ‘slowness’ of your avatar riding the Jungle Loop on a road bike). However, when the road surface changes regularly (I’m thinking the Watopia Marina on this one) you will notice that your avatar changes speeds (possibly up to 20%) as the road surface changes, even if you keep your gear and cadence the same, AND the grade stays at 0%.
Personally I find this annoying since it completely ignores the laws of physics. (Sure, Zwift isn’t perfect in this respect, but in most areas they at least make an attempt, but this is just blatant disregard.) So, if you find your avatar speeding up and slowing down around the Marina while you are riding the same gear and cadence, this is a ‘feature’ not a ‘bug’. 
If they didn’t ignore the laws of physics, there would be no reason to ride a gravel or mountain bike
To model different surfaces more like IRL, and to model draft effect too (e.g. you can feel draft effect IRL via the pedals), Zwift would need to alter the trainer resistance.
But I think that the protocols only work on a gradient basis, not purely resistance. I.e. the trainer relies on being told a gradient and it uses that to set the resistance. So Zwift would need to fudge the resistance to simulate this (e.g. putting you on a 1% gradient when you hit a flat section of gravel).
Maybe the protocols could be updated in time with a rolling resistance parameter as well as gradient?
That would be fine for most trainers, since the trainer ultimately has control over just one parameter: resistance. For the Kickr Climb and smart bikes with tilt, however, more advanced protocols would be necessary. But that’s already a problem today with the trainer difficulty slider, I would imagine. If you have a tilting trainer and you set your trainer difficulty to zero, do you lose all tilt as well?
1 Like