Regular readers of this blog will no doubt have come across at least one of my rants about how frustrating it can be that the power values I get from my Assioma Duo pedals do not match the power values from my Tacx Neo 2 Turbo Trainer.
Here are a few examples:
- https://cyclingindoors.co.uk/joining-the-sufferfest/
- https://cyclingindoors.co.uk/a-second-attempt-at-learning-the-sufferfest/
- https://cyclingindoors.co.uk/tour-for-all-2020-stage-2-longer-race-a-b/
- https://cyclingindoors.co.uk/better-connection/
I’ve definitely moaned about this more, I just can’t be bothered going back to find every such instance.
Zwift’s Peculiar Power Matching
What’s most peculiar about this problem is that, out of all the different cycling platforms I’ve used, only Zwift has ever managed to get the two to align.
I don’t know how it does it, and I don’t know why it specifically works on Zwift, but that has been a compelling reason for me to stick with Zwift for so long.
Along the way, I’ve tried a bunch of other cycling programs, some of which I really enjoyed. The Sufferfest, in particular, was one I would have liked to continue using, but it’s extremely frustrating when I’m putting in an effort and the numbers just don’t match up.
By this, I mean the numbers I see in The Sufferfest (or whatever app I’m using) compared to the numbers being reported by my Garmin head unit from my pedals.
Moving On from Zwift
The price hike around April this year was the last nail in the coffing for me, and I’ve moved on from Zwift. I was getting a bit bored with the app after being on it for at least three years without seeing much improvement or new content. Additionally I’ve been cycling outdoors more this year, making it feel like I wasn’t getting my money’s worth anymore.
However, since the accident, I’ve been exclusively back to indoor cycling, which is good in some ways because, at least now, the blog URL makes sense again.
I’m now predominantly, or even exclusively, using MyWhoosh as my training platform of choice.
Unfortunately, like all the other apps I’ve tried, MyWhoosh also suffers from this power discrepancy issue.
A Helpful Suggestion from Dr. Robert
As I’ve mentioned and ranted about in the past, this power discrepancy has been a long-standing frustration.
However, a blog comment by Dr. Robert Saunders, to whom I’m extremely grateful, pointed me to his really helpful and detailed post on comparing the power values from his Tacx Neo 1 and his Assioma pedals. I was immediately intrigued because this is practically the same setup I have. He was seeing a slight difference, which is to be expected, as the pedals are closer to the feet and there’s no chain involved, so naturally, a slight variation is always going to occur. I understand that.
Dr. Robert noted that cadence plays a role, with the higher the power, the closer the two values seemed to get.
However, in my own rudimentary testing, this difference is always there for me, regardless of the power level.
My Testing Process
I’ve never scientifically measured this power difference before, which in hindsight seems a bit silly.
I’ve searched for answers online many times but only ever found a really old forum post on the Tacx forum (not sure if it’s still there).
Here’s another good one with a similar issue – https://www.reddit.com/r/Velo/comments/kpjo1u/different_power_readings_smart_trainer_vs_assiomas/ – but it looks now like there are far results on Google about this exact issue than there were a few years ago.
The main gripe I have with Tacx turbo trainers is that they’re self-calibrating, so you’re pretty much stuck with whatever value they report.
Dr. Robert recommended DC Rainmaker’s Power Comparison Program, but since I’m a bit of a cheapskate, I tried to find something free. Reddit led me to https://compare-the-watts.com/, which appears to work exceptionally well, and that’s where I’ve sourced my data.
To compare, I used the power reported by the Tacx Neo 2 and my pedals.
As shown in the screenshot below, I had to offset the MyWhoosh data by 75 seconds since it took me that long to actually get going, but the pedals only started reporting once I set off.
Workout and Data Collection
For this ride, I followed the Rolling Hills No. 3 workout in MyWhoosh. I paused the workout just before the cooldown to do my own “scientific testing.” My plan was to do three different tiers of effort:
- Zone 2 Effort: Around 160 watts.
- FTP Effort: Between 200 to 230 watts, which I considered my FTP pre-accident.
- Max Effort: A one-minute, challenging effort above 300 watts at 95 rpm.
I didn’t manage to test at a low cadence as planned since the max effort at 95 rpm wiped me out, but I was happy to get the data.
The Results
When I looked at the output from Compare-My0Watts.com, the power difference between the Tacx Neo 2 and my pedals was consistently around 7.5% to 9%.
Click the image below to zoom in.
On average, the pedals reported about 8.5% higher than the Tacx Neo 2.
This difference doesn’t seem to scale with effort.
At my max effort of 310 watts in MyWhoosh, the pedals were showing 332 watts. A 20-watt difference might not sound like much, but you can really feel it, and I personally certainly cannot sustain that difference at higher outputs.
I had hoped the lap button on my Garmin would help with comparison, but it turned out it did nothing inside MyWhoosh. The only lap recorded was when I paused the workout.
Conclusion
So, what does all this mean?
It means the workouts I’m doing are still valuable.
I’ve adjusted my FTP down to 215w, compared to 235w pre-accident.
But if I’m training at a set power and still putting in the effort, even if the numbers on screen don’t make my virtual travels quite as far or register as many burned calories, back in the real world I’m still doing the work and getting the benefit of training at the higher wattage.
Dr. Robert once asked me, “Why don’t you just use the TacxNeo as the controllable trainer and as the power meter?”
Rationally, I know I’m still getting the benefits of the workout, and in Zwift, the numbers always matched up anyway, so I wasn’t losing out there.
But the frustration remains—seeing lower numbers on screen makes it feel like I’m doing less than I actually am.
Maybe it’s just vanity.
Final Thoughts
In conclusion, it’s good to have quantifiably proven this discrepancy exists. I don’t know how to fix it, but at least now I know it’s real and not just in my head. The difference is steady regardless of effort, and cadence matches perfectly between the pedals and the Neo. I’m not really losing out by using the Neo for my workouts.
Lastly, a big thank you to everyone who’s had to endure my rants on this issue, and a special thank you to Dr. Robert for suggesting this comparison method and writing such a helpful blog post.
Just a thought (and you may have considered this), but what does your LR balance look like? Perhaps Zwift takes power from both pedals, whereas other platforms take one pedal and double the power.
I did check this and it’s at zero – it’s not something I’ve ever touched. I trust the pedals 100%.
I reckon if I could put my own “man in the middle” between my kit, the pc, and the software, the “problem” would become very obvious. I know that’s possible – it’s how people cheat – but I’m not willing to invest the time in setting all that up. Right now, I know there is some issue. So long as I can get a decent workout based on whatever numbers, I’m happy enough.
I do wish I knew some local folk who did this sort of thing, then I could try my kit on their turbo. That would also answer a bunch of questions I’m sure. Such a pity the price of turbos etc is so flipping high.