A user posted on this issue back in April. There was recently a YouTuber accused of cheating (edit: they actively use the bug, even knowing about it), and I am almost certain it is related to this issue.
I recently purchased the Zwift Ride and installed the Cog v2 on my Kickr v6. Prior to this, I had been using the Wahoo Kickr v6 for about 18 months, and I dual recorded about half of all my rides with Favero Assiomo Duals. The power metrics between the Kickr v6 and power pedals was almost identical. I have a significant power imbalance between legs, due to nerve damage from back surgery, so I appreciate having pedaling dynamics from the power pedals.
I recently was promoted from Cat C to Cat B, and I was having PRs all over the place - from 1 minute all the way to 2 hour power output. Ignorantly, I thought I was really hitting a new peak - until I came across the video of the YouTuber accused of cheating. It was during this video that I became paranoid that something was going on with my data - so I put my power pedals on the Zwift Ride and did a quick test. The results were awful. The average power on the Trainer with virtual shifting was between 10-15% higher than the power pedals. This is not even remotely close, or acceptable. I then just started to pay attention to what happens when I stop pedaling altogether, after building up to a decent power and cadence. Upon stopping the power will initially drop to zero, but then the watts go up - in addition to the in-game cadence. Not pedaling at all, the watts then stay at 25-40 watts, RPMs up to 110. Overall, my dual recorded test ride resulted in an average power variance of about 35 watts - this resulted in a significant difference in distances, and skewed everything. I came across a Reddit thread that highlighted that this issue is related to virtual shifting, and recommended a factory spindown - after doing this, everything was completely normal (and accurate). However, since doing the factory spindown, the problem appears to be creeping back - but not nearly to the extent it was initially. In addition, the Reddit thread mentions that the problem occurs only between gears 22-24 in virtual shifting, however I believe that it is a problem no matter what gear - as I haven’t even hit 22-24, and the problem is coming back.
The post from April didn’t have any resolution. It seems very unlikely that I am the only person experiencing this problem as well.
Question 1: Are your Assiomas paired to a GPS head unit or to Zwift? They should be paired to the head unit, same as you’d do outdoors. Your KICKR should feed power data to Zwift.
Question 2. Is your GPS head unit adjusting the Assioma’s readings for the crank length? If no - that can skew your power numbers considerably. The pedals have no idea what length cranks they’re installed on - the conversion of torque to power happens in the head unit or Assioma app, and the crank compensation happens during that conversion.
My power pedals are paired to my Garmin Bike Computer. I’m not sure about GPS head unit adjusting for the crank length. The primary issue isn’t really the Power Pedal accuracy/Bike Computer - it’s that it appears I’m getting a bump in watts from the Kickr v6 and virtual shifting. This is currently better after doing the factory spindown, but it also appears it is coming back. It is very noticeable when I stop pedaling - in-game the RPMs stay up, and the watts are not zero.
Check the settings for your pedals within your garmin, there will be a crank length element, default being 172.5mm. If you use your garmin to calibrate your pedals, the crank length on your garmin will be fed back to your pedals as part of the calibration, if that crank length is wrong your pedals will then be out. The free watts thing though is concerning…
I’ve been using virtual gearing on Kickr Move (which is a v6 on rails) for a couple of months, but not noticed any power drift from setting new PBs etc - I’d need a power meter to verify of course! Watching the thread with interest however.
If you run the trainer without virtual gears (e.g. no Play or Click controllers) so they aren’t active and the gear index-calibration doesn’t happen (I assume it doesn’t?), does that help with the auto-calibration (showing it’s using the virtual gears is the problem), or do you think that it’s because the trainer has virtual gear capability whether you use it or not causes the problem? Does it still have an auto-calibration problem if you do a workout in the Wahoo app?
I’ll also look out for the power bump when stopping pedalling on my next ride.
Once the issue was present, it didn’t matter if I connected to the Play controllers or not - watts wouldn’t go to zero, nor would RPMs. The only thing that helped was doing the factory spindown. I did not try doing a workout in the Wahoo app, but I’m willing to try (if the issue comes back - everything is still looking pretty good from the factory spindown).
Hey all - we’ve discussed this with our colleagues at Wahoo, and this is a known issue that will be addressed in the next firmware release for KICKR v6 and KICKR Core.
The short-term workaround is to perform a factory spindown test + power cycle to put the KICKR back to normal.
The really bad thing is, and something that I probably didn’t make clear - but everyone should understand, is that when the issue is present, even when pedaling, the “phantom free watts” is active. This is what really made me sick - I very seldom coast, it’s just easier for me to keep the legs moving. When I tested, my dual recorded rides indicated that my watts were being artificially boosted by the same watts - as what I was seeing when I wasn’t pedaling at all. So basically, anyone who has this happening is receiving a bump in overall power. I do understand your point though - yes, reverting back to the original state, after doing the factory spindown, not using virtual shifting - would definitely be a good way to prevent the occurrence, you’re right. I just wanted to make sure that it is understood that if a rider is currently experiencing what is outlined, that they are receiving a bump in watts - pedaling or not. It simply is most obvious when not pedaling - watching the rpms jump up to 110 or so, and watts go up to 25-40.
That would explain some of the ZRL and club ladder races I watch on youtube. The odd one you get the rider has stopped pedalling and their watts are still active,… worse part is, some of them stream their rides.
That makes sense… a power offset that’s there nearly all the time. I’ve checked a recent ride where I was using virtual gears. The .fit file shows all zeros for a section where I know I had stopped pedalling down the Epic KOM and there was no evidence of free watts I’m pleased to say, so I don’t think it’s affecting everyone. I’ll do a factory spindown later though to try and make sure things are accurate.
Yes - that is good. I tried to take a screenshot of a short 4 minute effort - I moved my cursor at as close to the same point as possible to highlight the effect - granted, this is when I wasn’t pedaling.
From the above - the screenshot on the right is from my Favero Assiomo Duals - you can see both cadence and watts are zero. On the left is Kickr v6 and virtual shifting. In the top portion you can make out the overall effect - and it is more than impactful, we’re talking 2.3 vs. 2.6 w/kg - and this is just a short, zero effort spin. Imagine the impact on larger efforts. I have suspicion that the watt boost is also variable, that’s why I almost tend to characterize it as a % increase to power. While not purely statistically significant in the pure sense - I really believe that I was getting a 10-15% boost overall.
Yes, that clearly shows what’s happening, and it’s not insignificant, as you say. I wonder why some are/aren’t affected. I often run ERG workouts in a different training app, with no virtual gear pairing, so it’s possible the auto-calibration is working outside of Zwift.
I’m genuinely embarrassed. I feel like crap about it. If I could save rudimentary aspects of my rides to somehow get credit for being active (for Strava and general fitness tracking), I’d try and adjust everything since around September 11th or so - that’s my best estimate for when the issue really started in my case. I received the Zwift ride a couple of days before that, installed the cog v2, and it was likely in one of the Zraces, in which I “tried” to sprint (can’t really sprint too well, due to bad nerve damage in my calves and hamstrings). That’s when I probably triggered the auto calibration in the upper virtual gears 22-24 - something mentioned frequently in the Reddit forum.
