Zwift power figures query

[Paul Drawmer]

I'm not a Zwift user, but my mate is, and he's in the same age group as me.
On road segments in Strava he's a bit more powerful than I am. So his Strava road figures look OK to me.

But the Strava figures for Zwift segments look as he's going much too fast for the power being recorded. Will the Strava wattage figure be from his Wahoo Kicker or some derived figure output by Strava? He says he's freshly calibrated the Kicker.

Also looking at Strava leader boards for Zwift segments shows huge variations in power output for similar speeds. Is Strava doing something silly with the rider's weight rather than just taking the power figure from the trainer?

Basically he always looks fast on Zwift!

[Mastodon2]

If your mate looks unusually fast on Zwift, he may well have given flattered himself a bit when inputting his weight in the game. It isn't uncommon from what I gather.

[millen]

Similar question from me. Several in the club I'm in are doing a lot on Zwift now and seem to average 15-20% faster speeds than their typical outdoor rides. They're using different turbos and I doubt they're all misreporting weight!

I'd assumed this is partly that turbo riding is far more intense than outdoor, with much less freewheeling, no braking, no wind, and rides generally shorter.

Doesn't bother me as I don't have a turbo; just curious that the speed discrepancy always favours the turbo.

[anonymous-user]

You will find out next time you ride with him :-)

[frisbee]

Zwift does give the appearance of being a lot faster than outside. However there are a number of factors that bump up the speed so as a simulation it isn't that far out:
- you are pedalling and putting pressure on the pedals the whole time. Outside I spend about a quarter of the time every ride below 50W.
- On a lot of Zwift rides you are getting sucked along in huge groups
- The gravel roads in Zwift are more similar to UK roads
- A lot of Zwift rides are very flat
- On Zwift you are in a perfect aero position on a very aero road or TT bike

Or he's fiddling his weight and height...

[IroningMan]

Strava power figures are pretty meaningless: they're just a basic calculation, not a measurement, so they might have some value in that you can probably say that you worked harder on a ride where Strava says you averaged 250w than you did one one where it says you averaged 175w, but that's about it. Trying to use it to compare two different riders is not really valid.

Zwift, in contrast, is measured power - either with a power meter if the trainer has one or with a speed/power conversion algorithm if it doesn't: you can't really compare Strava power with Zwift power with any expectation of consistency because they're not the same thing.

[ian in lancs]

Mastodon2 said:

If your mate looks unusually fast on Zwift, he may well have given flattered himself a bit when inputting his weight in the game. It isn't uncommon from what I gather.

This and / or 'difficulty' is dialled down.

[frisbee]

ian in lancs said:

Mastodon2 said:

If your mate looks unusually fast on Zwift, he may well have given flattered himself a bit when inputting his weight in the game. It isn't uncommon from what I gather.

This and / or 'difficulty' is dialled down.

Difficulty doesn't work like that, it doesn't affect power.

Trainer Difficulty scales the gradient simulated by the trainer, so at 50% difficulty a 10% gradient climb in world would be reduced to 5% at the trainer. The power the trainer measures is unaffected, it just means you don't run out of gears and don't have to change gears so much when the gradient changes.

[Juffled]

Wheel on trainers generally over-read compared to direct wheel flywheel trainers (can be up to 15-20%), so they may just be using a wheel on one and getting a boost.

Strava power data is meaningless, its disregards drafting/wind direction and estimates a number of variables such as your CDA, rolling resistance

Weight is the primary variable in zwift, if they are using their 'racing' weight and not their actual weight then they will go faster on segments. Also just ignore zwift segments, it doesn't translate to the real world very well. We all know many people who can suddenly magically hold 22mph for an hour on zwift but cannot do anywhere near that outside.

Its a good training tool for engagement in what is a very boring process....turbo training, that is all, you cannot really compare different people on it unless they all use the same equipment, including the outside where the turbo becomes a road.

[Paul Drawmer]

Thanks for the replies..

The bit I don't understand is if a Zwift segment is showing up on Strava and a smart trainer is being used; then will the speed and wattage in the Strava table be the figures of speed and power as output by the trainer?

[ian in lancs]

frisbee said:

ian in lancs said:

Mastodon2 said:

If your mate looks unusually fast on Zwift, he may well have given flattered himself a bit when inputting his weight in the game. It isn't uncommon from what I gather.

This and / or 'difficulty' is dialled down.

Difficulty doesn't work like that, it doesn't affect power.

Trainer Difficulty scales the gradient simulated by the trainer, so at 50% difficulty a 10% gradient climb in world would be reduced to 5% at the trainer. The power the trainer measures is unaffected, it just means you don't run out of gears and don't have to change gears so much when the gradient changes.

and because the gradient simulation is dialled down you go faster for a given power input

[frisbee]

ian in lancs said:

and because the gradient simulation is dialled down you go faster for a given power input

No. Your in world speed is calculated solely from your measured power. Your cadence, gearing and trainer speed is irrelevant.

[wobert]

Interesting stuff.

I’ve a friend who has recent acquired a smart turbo and is posting his Zwift exploits all over FB.

His last ride was 60 odd miles at an average speed of 18.9 mph.

On his road rides he’s averaging 15.0 mph, so Zwift appears to ‘flatter to deceive’.

Having tried Zwift, FulGaz and TrainerRoad, I prefer the latter, as ‘it does what it says on the tin’ from a personal perspective.

[lufbramatt]

I agree, from looking at people I occasionally ride with, their speed (and therefore distance covered) in zwift is about 20-30% higher than real life.

When I do Trainerroad sessions it reports a "distance" as I have a speed and cadence sensor on the bike as well as the power meter. This gives the speed of the back wheel rather than a calculated game speed. My Road Machine I find is slightly harder than real life so a pretty hard session may be 270watts average power but only about 18.5mph "speed". but I keep the setup consistent so I record it and realise that it's only relevant to me. I expect all the Zwifters that follow me on Strava think I'm just taking it really easy on my indoor training!

If you have a power meter on the bike then Strava reports that data rather than it's estimated power on segments, look for the little lightning bolt symbol next to the figures in the leaderboard.

Edited by lufbramatt on Monday 18th January 09:20

[okgo]

Speed is irrelevant.

Is he doing the same power numbers inside that he is out, and if he uses heart rate, do they broadly match up?

Strava power as far as I know has no effect on Zwift, as in to use Zwift, you are using either powermeter, or turbo trainer inferred watts, Zwift is not running on strava est power at any point, is how I think the chain of command is.

Or just post his profile, I'll tell you in 5 minutes whether any of it is legit

Edited by okgo on Monday 18th January 09:35

[wobert]

okgo said:

Speed is irrelevant.

Is he doing the same power numbers inside that he is out, and if he uses heart rate, do they broadly match up?

Strava power as far as I know has no effect on Zwift, as in to use Zwift, you are using either powermeter, or turbo trainer inferred watts, Zwift is not running on strava est power at any point, is how I think the chain of command is.

Edited by okgo on Monday 18th January 09:29

I can’t answer that question, but given he’s considerably heavier than I am by more 20 kgs and on the Zwiftride previous he averaged 170w.

This sixty miler was 156w for 18.9 mph average, which seems erroneous to me...

[okgo]

wobert said:

I can’t answer that question, but given he’s considerably heavier than I am by more 20 kgs and on the Zwiftride previous he averaged 170w.

This sixty miler was 156w for 18.9 mph average, which seems erroneous to me...

Watts are the only measure that matter really. As said above. 200w from a reliable powermeter will see you go much quicker on Zwift than in the real world, its as simple as that.

I ride around on zwift on the flat at 40kph, not the case on the roads sadly

[emicen]

okgo said:

I ride around on zwift on the flat at 40kph, not the case on the roads sadly

Amen

I don't have a power meter for real life, but my average on the flat in Zwift is around 40km/h, in real life its more like 32-34.

[z4RRSchris]

zwift speed / kms are pointless. all that matters is power

I did 300km at 140w @ 33.4kph.

a) I could never find a road that flat IRL
b) for 140w i'm not going to be going 33.4kph IRL
c) i would never do 300km outside, its pointless

https://www.strava.com/activities/4541253998

[okgo]

z4RRSchris said:

zwift speed / kms are pointless. all that matters is power

I did 300km at 140w @ 33.4kph.

a) I could never find a road that flat IRL
b) for 140w i'm not going to be going 33.4kph IRL
c) i would never do 300km outside, its pointless

https://www.strava.com/activities/4541253998

Real world comparison, and that was two up, so probably would have been more like 240W solo.