PowerMatch - +/-30W variability with Stages Gen III and Wahoo KICKR

Hi, I wonder if I could get some advice on a matter that has been driving me nuts?!

I have been using PowerMatch for several years without any particular issues using a Wahoo KICKR and Stages LH PM. However, recently I have switched bikes from my road bike to my MTB. The MTB has a Stages LH PM the same as the road bike but I am experiencing much greater variability in power data in ERG mode during workouts.

I have learned to ride in a small gear and go really steady with the cadence - something that I have perfected over years of interval training. With the road bike, I was experiencing only small amounts of power variation, typically no greater than +/-10W against target power. However, with the MTB Gen III PM, I am experiencing much greater variability from target power, typically bouncing +/-30W from target power, despite an absolutely rock solid cadence. This is driving me crazy as in a recent ramp test, it was jumping all over the place and I found it really difficult to hold target power, the consequence being that the result of the test was a suggested drop of 11W on my FTP, despite the fact that I had been nailing all of the preceding sessions in the build plan at the higher FTP. On a VO2 max session tonight, the power was typically low versus target, such that I was constantly not hitting my target power for each interval, to the extent that I just upped the FTP by a couple of percent to ensure that overall my average power for each interval was where it should be.

This is extremely frustrating as I’ve had the best season of consistent training ever due to lockdown, not missing a single session on the HV plans, yet I feel that its limiting my progression as the trainer is just not giving me the resistance that I should be working to.

Is there some way of improving the data from the Stages, such as modifying the power smoothing or something? Any thoughts most welcome. Last resort is going back too the road bike but I’m trying to do all my workouts on the MTB since that will be my race bike this season.



How exactly are you judging the power data swings?

  • Looking at the 3/4-digit live power info vs target power or looking at the final TR workout info in the yellow power data graph vs the blue power target blocks?

Is it possible to see a similar workout done on the roadie vs mtb to see these variations?

What is the actual smoothing setting you have in the TR app?

  • Of note, keep in mind that this setting only controls the 3/4-digit live power info in the window, and does NOT alter the yellow power data graphed in the final workout info.

Do you know the specific gearing used with each bike setup on the trainer?

Are you watching power data while in ERG?

  • Sometimes that can lead people to try to influence the power by altering cadence. You say you are steady, but I just wanted to verify this because watching power in ERG seems more problematic than useful in my experience.

Without any more info, I wonder of the data reporting between your two power meters is the real issue here, but I have no way of knowing or reason to suspect other than it’s one of the biggest variables from your comments.

Hi Chad, thanks for following up on my query. I respond to your points in turn…

I am looking at the post-workout data for each interval block, looking at the yellow power line vs the blue target. I largely ignore the actual power value in the dashboard during the workout for the very reason you highlight, that in my experience this only results in a cat-and-mouse game of chasing the target vs cadence. Hence I focus almost solely on cadence when in ERG mode, really trying to hold it steady right through the interval, making sure I am holding target cadence at least 20 seconds before the start of the interval.

I checked the app and it looks like 4s power smoothing but I have literally just updated to the latest beta version and I’ve not done a workout using it yet. I was however using the default power smoothing in the previous version of the app.

Gearing is an interesting one since I have a 1-by chainring on the MTB (32T) and granny ring on the roadie (34T). I am using roughly the same sprocket on the cassette for both bikes - 16T.

It my well be an issue with the PM itself. I have decided that I will do all ramp tests using the road bike and training using the MTB as the road bike gives a more accurate test result it would seem.

See attached two images of similar intervals on the two devices (upper one is MTB) although I’m not sure these show the variation adequately at this resolution.

Thanks again,

MortStage MTB Power data Stage Road Power data

Try the DCR Analyzer Tool Overview & Manual (dcrainmaker.com) if you’re concerned about PM variability. It will really help with the comparison and analysis.


Will check that out, thanks @Henry_Lee!


Great info and all sounds pretty good, with one exception.

I advise against testing with the roadie and swapped to the MTB for training. Without and way to verify, there is no guarantee that the two are close enough to each other to be “even”.

Maybe you vahe done this comparison with your trainer and I just missed it? But if not I’d suggest doing a comparison via the DCR tool to really see if they are close enough to each other to call them the same.

If they differ by enough, I’d stick with testing and training on the same one.


Will do, thanks Chad. I felt that the two PMs were close enough based purely on feel but will do a proper test to benchmark them. Cheers!

1 Like