I feel all manufacturer’s claims of accuracy should be confirmed by the end user by dual recording.
If enough people do this and make results public, then we will see which trainers are accurate and which are not
I feel we will find that most trainers are accurate under some but not all conditions and the cheaters will try to take advantage of those situations.
We may not be able to prevent them from doing so but we can expose them, if Zwift would permit us to do so.
More important, are the folks who are inadvertently in one of those situations.
They are people who can be educated and taught how to correct issues and become good racers.
I feel along the lines of the opposite. It’s up to Zwift to do this; it shouldn’t be the consumer’s fault. Zwift shouldn’t publish a (shopping) list of compatible trainers without at least a disclaimer of some sort. For that matter, it should also be up to Zwift to publish what constitutes cheating – aside from the pro e-sports stuff, AFAIK this doesn’t even exist.
So starting with the fundamental 2 things listed above that do NOT exist, it seems arbitrary, especially to the uninitiated, to label them cheats, nor to assert that they need to go out and invest in something better, when Zwift put their equipment choice on a list that indicates compatibility.
This is spot-on the core of the problem. For people coming from outside the sport, things we accept as pretty simple are going to be complex. Measured Power vs zPower vs A Bike That Effectively Calculates zPower Onboard And Transmits As Measured Power. Sheesh.
The software can’t tell that a reported signal of ‘watts’ is not real watts–that’s how bikes like this Horizon work. Or the Decathlon D100, or the old Kinetic pseudo-power remedial-smart trainer. Onboard software sends a watts signal, Zwift don’t know. The only way to go would be a whitelist/blacklist. But Zwift wants to be cycling for everyone, non-competitive included. And in fact that’s the majority of the user-base. If someone wasn’t going to race at all, why should they spend for real power measurement?
Aaron’s suggestion above might still be a good way to go. Because Zwift can’t stop janky manufacturers from tricking people into thinking they are getting quality. Let these people ride non-races, but do simple things to indicate that they’re not getting accurate power numbers. Make their data numbers orange or something, with a popup on pairing that says “Just so you know, you will be getting estimated power with this device.”
And then of course some clear indicator on the List of Devices of what it means if you own/are looking to buy one of these fake smart machines.
That, and do the community race organizers even care? Filter Race events on Zwift’s own events page and filtering that there be a Power Meter (aka “Hardware Requirement”), it seems striking that virtually the ONLY events where this Hardware Enforcement is set, are ZHQ events (Zwift Games and Crit Racing club). There were 2 other outlier events where this was the case.
FYI @Jesper_ZwiftHacks the Hardware Requirement or No Hardware Requirement rules filters on ZH are removing every race event for some reason.
Need to stop comparing things as racing and non racing as it just means it doesn’t get the focus it deserves.
If someone pays their membership doesn’t race but values the leader boards they have ever right to expect a fair playing field. So things like ADZ leaderboard should be clean data, sprints should be clean data.
If people are using spin bikes/QZ app/zPower then perhaps they should be exempt from the leader boards.
Trying to say only racing should get a sanitised environment is unfair on those who don’t race but value their own efforts in game.
That exists, it just doesn’t seem to have been adequately developed. Blacklisting some number of spin bikes is playing Whac-a-Mole since there are so many random products out there, more every day. There are some spin bikes excluded from hardware-enforced events. It would be better done as a whitelist IMO. Publish a list of trainers approved for racing.
Sounds good – a white list. If a manufacturer/model wants to be certified, they can ship a sample of their product to Zwift to verify if Zwift doesn’t have one already.
I was suggesting combining a whitelist/blacklist with non-racing features–namely something that signals the rider that their data should not be taken as being accurate.
I think that’s fair. There is a clear racing/nonracing distinction that can’t be ignored, and there should be limits (and are, but maybe need to be better) about what machines can even be allowed to race. There shouldn’t imo be limits on what machines are allowed to free-ride/group ride outside of racing, as that would price things way out of some people’s reach. But you’ve got a good point about the leaderboards–excluding these machines from leaderboards also shouldn’t affect them all that much either, if they are made aware of the shortcomings of their hardware.
Fair–what I meant and should have written was “why should they be forced to spend for real power measurement?” My entire point there was that plenty of people using Zwift have no desire, and no need given how they use it, for accuracy.
That’s a matter of precision, not accuracy. Although of course it’s an open question as to whether these machines have that either.
(Edit: if someone isn’t comparing themselves to others, then if their machine is inaccurate, it might not matter for them as long as it’s inaccurate the same way all the time–precision. Progress can still be measured. It’s when they want to say “I have such-and-such FTP” to someone else that accuracy becomes a factor. But a lot of people don’t care about that.)
I still don’t see how we get away from dual recording by the end user.
Didn’t we just have a race series cancelled or put on hold because of power inaccuracies caused by virtual shifting under high torque/low cadence situations?
These were high end accurate trainers that turned out not to be so accurate.
My low end kinetic trainer doesn’t have that problem but is only accurate when calibrated against a second device.
A list based on manufacturer’s testing misses conditions not tested for and ignores users who have taken steps to ensure accuracy.
And.
You can make a user use a specific piece of hardware but then we have no clue how much they weigh or their gender or if they use testosterone replacement.
Black list yes.
White list maybe.
Grey list - let the user demonstrate accuracy.
Monitor and ban unrealistic performance.
It’s why I have been arguing to allow e-bikes in racing, but not elsewhere. It’s the reverse of the “as long as it doesn’t affect racing I don’t care” argument.
ADZ leaderboards are so useless because it’s always cheaters, and I don’t mean subtle, but outrageous cheating. That must be cleaned up!
It shouldn’t be up to me to flag and hope someone gets banned after their fourth 30min lap of ADZ in a row at 39kg. That should be automatic. I’m sick of trying to type explanations with sweaty hands about why these folks should be excluded. When it’s so screamingly obvious that should be done automatically.
It’s even worse when someone with equally dubious weight and power figures is loudly shouting about another person cheating on ADZ. That was observed last week. I felt like calling it out but kept quiet. Do they not realise people can still find out their weight…
On the topic of dual recording, I do that because I need to know my power balance (because of injury) and my kickr bike doesn’t give me that.
Also some people may not race in Zwift, but they do race IRL and want to track things same as I’m doing.
How do you dual record WITHOUT getting duplicate rides which require a lot of work to deduplicate ? (I also have a huge imbalance but I have given up recording on zwift because of the hassle)
Edit: zwift really should record balance btw, Rouvy does…
I just deal with it. The saved data from the Garmin has a different name to the Zwift recorded ride so it’s not terribly hard to sort them.
I don’t save every ride from the Garmin, but I do keep a serious watch on it, but it’s probably a lost cause trying to recover such a big power difference.
Some of these bikes are quite accurate, but those are the exception.
Most have an estimate simply based on resistance lvl
Some have based on resistance lvl but combined with a calibration procedure (than often can be manipulated)
