zPower: Don't be a "Flier"

(Jason K) #1

Hey everyone!

Thank you for helping to test the Zwift beta! We’ve been getting some great data and feedback that’s helping us to figure out what to develop and prioritize. You’ve all been a huge help!

We know there’s been discussion about zPower, “fliers” (cyclists who - whether intentionally or not - use zPower incorrectly), and the effect this situation is having on the Zwift community - and we wanted to talk about where we’re at on this.

Our game industry veterans are very familiar with this kind of situation, and we’re working on a number of potential software-based solutions for making sure that everyone can enjoy a ride in Zwift. However, we are still in beta, and developing these solutions will take time. This is why we’re counting on you, the community, to step up now and beyond launch.

There are some Zwifters who are approaching the situation with a level-headed perspective, and this is what we want to encourage: give your fellow cyclists the benefit of the doubt and realize that most people doing this just don’t know their trainer isn’t set up right (or our virtual zPower curves aren’t 100% dialed in yet). If everyone works together to educate them, we can have a stronger, healthier, more inclusive community.

We’ve created a couple different articles explaining both how to set up your classic trainer and how to confirm that your zPower setup is accurate. Please share these articles and this message so that we can all work together and make Zwift the best it can be!

All of us here at Zwift want both a product and community where everyone can enjoy a ride and each other’s company at the same time. Thank you for helping make that vision a reality!

Ride on.

-Team Zwift

(Ronald Termaat) #2

Thanks for the explanations.

Using a Kickr myself I see different behaviour of other riders compared to what I am doing. It would be nice to have in the ranking list a small indication showing on what kind of trainer the rider is riding. This will probably increase the quality of the feedback the Beta testers can provide.


(Christopher Pallotta) #3

Just a counterpoint with what I hope is also a “level-headed perspective…”

I don’t know what the solution is, or if there ever could be a solution. Or even if a solution is necessary. But… Even with riders using legitimate power devices, working correctly, properly set up and calibrated, there will be a range in how power is reported for the exact same effort.

I use an SRM power meter, a Stages power meter and a Kickr with its own power meter. They all are working according to the manufacturers specs*. All have been calibrated and zeroed-out or in the Kickr’s case, properly “spun-down.”

If I forgo the automatic resistance-change of the Kickr and use the SRM or Stages power meters to control Zwift, I’ll get wattages/performance that closely match each other. If I use the Kickr for power, I get wattages that average 25-30 watts higher than the SRM or Stages, and different (much better) performance on Zwift. 

Again, all three methods are considered legitimate to Zwift and they’re all being used on the island by conscientious people with good faith.

So, should I ride Zwift with my SRM/Stages? The wattage results are probably more “accurate.” My experience tells me, though, that this puts me at an unfair disadvantage as far as getting jerseys or keeping up with fast group rides.

Or, do I ride Zwift with my Kickr power? There are many on the island riding with a similar configuration and this (I believe) levels the playing field for me allowing a reasonable shot at jerseys and hanging with fast riders. This is less accurate power-wise, but more representative of my real-world performance where I actually can ride A-level groups and have earned many KOMs competing against very strong riders. 

As I always say, I love Zwift and don’t want to stir the pot. But these are questions worth asking. Ultimately, though, there may not be an answer and maybe we don’t really need one. 

*I’m confident that my Kickr is functioning according to manufacturer’s specs and that it is representative of the way many, if not most Kickrs report power. My first Kickr was sent back for recalibration and was returned with a big thumbs-up from Wahoo. It was the same as when I sent it back. Eventually, they agreed to take that Kickr back and replaced it with another. The new one had the same power profile. I then purchased the $100 calibration kit. Same results. Spindowns after a long warmup? Same results. Many, many others with Kickrs reporting their Kickr power in user forums? Same results.

In turn, I’m also confident my power meters are working according to manufacturer’s specs. For one, they both measure power very similarly. Second, the SRM really is considered the gold standard for power. Third, the SRM was recently sent in to have its calibration certified by the factory. And finally, the ftp I get when using the Stages or SRM matches the performance I can achieve in the real world based on calculations. The ftp I get when using the Kickr would appear to put me into an entire class above where I really am. 

(Simon Oxenham [VC10]) #4

I think it’s pretty obvious when you’re lapping the whole field every five minutes @ 10+ watts/kg!  Don’t underestimate how annoying this is.  The jerseys are an important part of the experience and they’re now effectively redundant.  

(Kickr Japiooh) #5

In Virtualtraining this is partially ‘solved’ by the possibility to set a filter so you can select the people / trainers you will see during the ride.


(Craig Howard WBR B) #6

I understand the difference between PM’s is variable but it’s also very small in comparison to Z-Power when not set-up correctly. This is enough to offer everyone with calibrated power a fair experience. However someone winning all the jerseys on ZPower pushing 10w/kg isn’t fair no matter their lack of understanding. It only takes one person to not know what’s going wrong with their set-up and the whole experience for everyone on Zwift is marred.

Why not have two grades of rider jerseys? Those using PM’s and those on Z-Power? I’d bet 99% of PM users are so data savvy, that they wouldn’t want to cheat and that is really the only way you’re going to cruise around the island at similar incorrect power readings from those using Z-Power setup incorrectly.



(Frank Goehner) #7

I have a feeling that once this becomes a pay service, the people that are cheating will stop using the product and move on to something else.

(Andrew Williams) #8

I did a little experiment with zpower last week, trying all the listed curves and comparing them to reality.  Reality in this case is represented by a Quarq which was warmed-up, properly zeroed and recently re-sloped to be as accurate as possible.


Some were way off, like 200% or something.  Some were about 40 watts off.  Some were lower than reality.  Most were higher than reality.  My trainer is not on the list (It’s a 1-up original model) so I wanted to find a curve that was as close as possible.  I eventually found one that was only 10 watts high, and settled on that one (In case I ride someday on Zwift without a power meter).

The final choice in the list, “unsupported trainer,” was the one that gave the lowest power readings.


It is therefore unlikely that people who choose “unsupported trainer” are going to end up with Superman results. 

I think people might either be knowingly choosing a trainer curve that makes them faster, or if they are ignorant, they are just clicking on the first one in the list.


Therefore, I suggest the following change:


Instead of calling it “unsupported trainer”, call it “generic trainer” and put it at the TOP of the list.


Worth a try?


(John Fighting Brain Cancer) #9

Perhaps the solution is to have a zPower leaderboard and one for all other devices that actually calculate power.  I know that there will always be some difference from one power meter to another but I think that it is close enough for leaderboards and sending to Strava.

The best thing about Zwift is that it makes riding on the trainer fun.  Sure I care about leaderboards et cetera, but I think that the principle goal should be expanding the island with new functionality/maps.  Whether you are number 1 or in last place the goal should be to do your best.

(Eric C. (Zwift HQ)) #10

For the record, if you choose a trainer you do not have, it will *always* give you wildly inaccurate numbers because the algorithm thinks you have a specific resistance unit.

Changing ‘Unsupported trainer’ to ‘Generic trainer’ might be worthwhile.

(anon46748293) #11

Here is my take on it. I do not care about winning any jerseys, scores or anything else for that matter in environment like Zwift. Even with powermeters there are ways to cheat. With trainers like KICKR and zPower we are measuring weather in California rather then rider’s true ability. So If I see some guy flying past me at warp speed I do not care What is good about Zwift is that since there are always enough riders of various capacity I can always find someone to engage in short competition with. It is like random interval workout with great motivation since you wanna beat that sucker right in front of you :wink:

(Kevin Morphet) #12


(Jeff Hirsch) #13

I have to agree.  When it’s obvious someone is doing an inhuman effort, I really don’t care.  I pretty much just ignore them for they might as well be on a motorcycle.


What I do care about though, is being able to ride with my friends or others of comparable ability and actually be able to ride with them.  The last few rides I had done on my CT and my buddy on his CT, both calibrated properly etc. etc.   I kept getting dropped, mostly on the downhills, but sometimes even on the flats.  It was like if you were being bounced backwords and forwards while drafting, eventually someone would end up shooting down the road.  In real life, he can’t stick with me even on a moderate effort on the flats. 

So what exactly causes this, be it the fact that the Racermate library isn’t used, or some other physics elements are out of whack, but it certainly doesn’t represent reality even with 2 of the same trainer types that are reporting proper power.


Does this really matter?  Well I suppose as long as I can get in a few 4-6 hour rides in February - March in Zwift, then I don’t really care.  Would I want to attempt racing or a real group ride?  Not a chance.

(Stewart G teamWBR) #14

Jeff, have you set the correct weight? I only ask as when the option to change weight came in I changed mine from the set 75kg to my actual weight, 95kg. The onlt problem was that with my setting at imperial I ended up with a rider weight of 95lbs! I didn’t notice this at first and couldn’t understand why I was being dropped on the downhills. I thought it had just been the way the CT interprets the data.

Another thing is that lately (maybe after recent updates) I find it difficult to pull away from others on a downhill even though I am heavier than most and even if I put in a sustained burst around 400-500W the other riders just pull up alongside me. This wasn’t the case a few weeks ago.

(Jeff Hirsch) #15

Yes, weight is correct.  I also tried ±5 pounds and it didn’t make any difference.

(Graham Easton WBR D) #16

I’m fairly new to Zwift and using zpower. For my (Elite) trainer the set-up seemed pretty spot-on. I have been doing interval sessions using the FTP levels I set on a wattbike, and if anything I think the power may be slightly under-reported on my zwift, but only very marginally. I only highlight this that probably for every one ‘flier’ there are probably 50 guys / girls using Zwift as it should be used.

I wonder if part of the answer is new users get introduced on a 5-lap ‘trial’ and zwift could alert them if something seems odd about their power. This would at least eliminate the unintentional ‘newbie’ errors. 

I don’t have a good solution for intentional cheating…

(Ride ide Safely) #17

Personally, I couldn’t care less what w/kg other people are pulling out. For me , Zwift is all about personal accomplishments. There will always be better people than you and even if someone out there is putting half the effort as I am for the same result, in the end, I do not think i’m the loser, they are.

You can cheat all the numbers you want but you can’t cheat your fitness. If someone is “flying” by me during my Zwift rides, I just ignore them. They could be a pro or they could be cheating themselves. The real beauty of zwift is when you compete against someone as strong as you. It’s great motivation just trying to keep up so let the cheaters be. 

(Timour Chayipov) #18

why not to make it possible to compete against yourself? just made couple of fantoms of your previous rides and ride with them

(Marcus Durant) #19

The only time it bothered me what w/kg other riders were putting out was when I (it) uploaded my first ride to Strava. There are a whole bunch of riders who seemed to be able to average 25-30mph up the KOM whereas I can only muster @15mph

I’ve now used Zwift a total of 3 times now and every time there has been some Muppet (or top level pro) lapping at 7-8w/kg average. Using a power-meter and a set of rollers there is no way I can “compete” unless I suddenly “weighed” 35-40kg’s.

I quickly came to the conclusion that what happens on the island should stay on the island. The whole thing is so open to abuse / misuse / error that I really see no place for it on Strava other than a record of a workout if that’s what you are using Strava for.

I now upload my data only from my Garmin & not from Zwift and treat it like the (more fun) interval session that it really is.

(Dave Clark | Michigan) #20

Potential solution - require a fitness verification every six months by requesting an upload of an out-and-back ride from their GPS/cadence(/heartrate) bike computer.