Accuracy of Zwift Power Results

I was riding the Black Celebration Series today at 12:30PM. I am on an unlisted trainer, with heart rate, cadence, and speed sensors. I noticed that my results # position on Zwift Power do not correctly reflect the ride that I did. I noticed that there were riders that were slower than me with a higher rating from the ride. Can anyone explain this phenomenon?

Hi @Q111, those events are group rides, so the results don’t matter. I’m pretty sure the position is random.


Hi Mike,

Thank you for your response. If you are correct, then they should eliminate the position # and give out results of group rides in an alphabetical order. However, I put in a really good effort today and I am monitoring my progress as I am recovering form an injury. i was really disappointed to put in that kind of effort for an hour only to see that the accuracy of the result is off.

For timed events the results is all random. Group rides are made to ride with the leader.

The number of people who crossed the line in front or behind you has no effect on your progress/training. Why does that matter? You won’t be riding against the exact same people the next time you enter an event, so comparing your results doesn’t matter.


Hi Mike,

Thank you for your response. The category was rated as E 1 - 5 w/kg. Even though it is a group ride, I find that on Zwift, there are groups that form along the way. I was working well with a group that was doing approximately 3.5 - 4.5 w/kg however, I was in the red (HR) along the way but, I was able to stay with the group. I think that it would be nice mental boost to see the performance level rather than some randomized result. Again, I think that Zwift should either list results alphabetically, rather than # position if they are randomizing the results for group ride. I use these group ride as a way to judge to what category I should select when I start to race.

You need to be careful how you interpret your performance.
3.5 - 4.5 on an unsupported trainer indicates most conclusions may be highly suspect.


Hi Tim,

This sounds quite plausible. I am 70kg so 4.5 w/kg is about 315w. I am on am Minoura Magturbo.

That is the biggest complain from Ride leaders, they take time out of their day to lead a ride and then people treat it as a race?

Zwift Power should do that. Good idea for improvement.

Racing against others that don’t race, that wont tell a lot.
You should race as a B rider according to Zwiftpower.

1 Like

Yes, I am trying to figure out which category I should race. Realistically, I do not think that I am a B category rider. However, I was not racing against other who were not racing, the event was listed as a group ride. I Like I said earlier, a group formed and we rode together, I was trying not to get get dropped that is all.

Looking at your performance you are putting out A level numbers. Last week you did A performance in the TDZ and today you did 3.976w/kg that is top B numbers.

I would think you would be upgraded to A in a week or two.


1 Like

This comes back to a point made by @Tim_Camden_C. An unsupported trainer may lead to spurious results.

Even if it is unsupported the power numbers will still make your avatar go as fast as a A rider therefore you need to ride with them.

It is not about how fast you are on a bike outside it is how you perform in Zwift so if your trainer reads high you need to ride in the higher category.

1 Like

Hi @Gerrie_Delport_ODZ

Hmmm, I feel like I am mechanically doping with this unsupported trainer. This is how I set my trainer up. My resistance level on the trainer was set to match my HR at a particular speed on the road. I matched the trainer resistance to the HR of a speed of 30kph on a flat road. However, I am able to turn 53 x 11 at over 90rpm on Zwift, which is something that I could never do on the road.

You probably are, but with that said you should still race in the correct category. Once you get more accurate power meter your Zwift racing performance will drop and you will race in the lower category.

1 Like

Thanks. Maybe I should not race till I get a supported trainer. I do not think it would be fair otherwise.

1 Like

I think you have the right idea.
For the first 2 years on Zwift, I tried to make my first trainer accurate and when I realized I could not do so, I tried to make it consistent but I found I could not do that either.

This experiment has been performed by many people in Zwift.
The first Zwifters have been Zwifting 6 years now.

The wheel on, zwift power curve trainers , supported or not, were never designed to be used to race against other people.
They are fine for work outs, group rides etc.
Zwift, initially, was not designed to have people racing against each other.
Like Gerrie said, if the results are consistent, then you can race in the appropriate category but freq the errors are either inconsistent or are nonlinear and grow in magnitude with increased power.

Please don’t be discouraged and keep at it.

Hi @Tim_Camden_C, @Gerrie_Delport_ODZ, @Mike_Rowe1

Thank you all for you insight regarding the aforementioned topic. You have helped me understand the Zwift Power better.

Best regards.