Race Cheating

Sorry, I don’t do facebook so I don’t know if this is done.

I did a zwift race and noticed people could instantly lay down 10 watts/kg and chase me down.

Looking at the strava data, it is obvious they’re just booting their pedal to make their power meter spike, and then Zwift smooths it out over a sustained period of time. You do this several times and you can keep it going for 30 or 40 seconds easy.

So basically you can just toss down 10+ watts/kg for very little effort.

That kind of sucks and is highly demotivating. I totally get there’s not much to be done about weight doping, etc., but this just seems like a bug.

Otherwise “everybody” needs to learn to do this, which is a bit silly.

If you suspect people of cheating submit a ticket: https://support.zwift.com/hc/en-us/requests/new


Community Guidelines: https://support.zwift.com/hc/en-us/community/posts/201078465-Community-Guidelines

I’m more wondering about the bug aspect. Reporting people for doing it seems a bit of a losing battle, since the game seems to be set up to encourage it.

Is it even cheating? I can’t even say that  I suppose.

Without having more info to go off of I couldn’t begin to say if somebody is gaming the system or not, but I can say that the standard says a power packet is valid for up to 3 seconds, and that Zwift does not smooth power data at all.   The only smoothing that is done is simply on the number that appears at the top left of the screen while you pedal. It does nothing to the avatars performance.

M: I’m not familiar with the phrase you used, “booting their pedal.” What do you mean?

Ah, the 3 second thing is the issue. Here’s an example, the power spikes from one hard stroke, then rest, then another.  The in-game experience is a sustained 600+ watt effort over the time period, while the individual is not actually sustaining 600 watts.

The 3 second average is only a screen display, having it set to 3sec or one sec has no effect on your performance.

If I look at your plot above it does not seem that this rider got any benefit from what he was “trying” to achieve. Look at his speed it is dropping from the moment he rested. 

I agree with Gerrie. I see nothing in the image that constitutes cheating. If they are putting out the power and then backing off then repeating a couple of times then good on 'em. They will burn all of their matches pretty quickly.

Now what concerns me about cheating or unintentional cheating is all of these low end wheel on smart trainers with dodgy power accuracy.

This is very easy to replicate and once done, you can see the remarkable advantage this pedaling style will give you. Pedal a few revs and end with a hard stomp. 5-10w/kg. Stop pedaling and the last power number sent from the power meter will continue for around 3 seconds, maybe more. Do this over and over and it will be like you’re holding a steady high power with just a hard stomp every few seconds.

Mark. Try doing this. In the settings set “Power display” to “Instant”. I think you’re assuming that the 3 second averaged power is being applied to the rider. It isn’t and this has been stated by Zwift’s Jon Mayfield. All it does is smooth out the onscreen power display numbers. 

Also be aware that there is a lag between what you do on your bike and what happens in Zwift.

At the end of the day if someone wants to ride in this fashion then good on 'em. I doubt that they can sustain it for long.

Nigel: No. I have the 3 second smoothing off. This has nothing to do with that.


Extreme example. Same ride, same power meter, one is Bluetooth data and the other is ANT+. Compare the average speeds. The Bluetooth spikes are spikes, ANT+ are plateaus.


If you’re convinced there is a problem I suggest filing a support ticket with Zwift.