Power meter vs turbo trainer readings

Wheelspinner

I’ve a Tacx Neo 2T trainer, which seems excellent. I do pretty much only the filmed rides, not the fixed workouts at all. Allegedly the trainer needs no calibration. I’ve done a couple thousand Km on it now, and am fairly familiar with the perceived effort I’m putting in.

Recently picked up another road bike which has a Stages LH Power meter crank, which hooked up to the Garmin I use, and asked if I wanted to calibrate it - yes.

Both the Neo software and my Garmin unit have the same rider details - weight, age, etc defined in the user profile. I’ve done a couple rides with the Stages bike, which have been easy pootle about type.

The Stages reports much higher “average power” than the Tacx. Those easy rides show up as consistently around 175 W average on a 40 minute ride with a couple hundred meters elevation gain at only 18 km/h speed average. On the Neo, a similar route in terms of elevation gain per km, but done at a much higher average speed of about 26-28 km/h, the average power shows up as about 150 - 155 W. I certainly feel that the effort on the Neo is much higher - heart rate is up, and I feel a lot more knackered by the end. I have “easy” rides on the Neo which show average power numbers much lower again.

Curious why there’d be such a difference.

I’d expect some variation but within a few percent at most. This seems way greater.

itboffin

Not sure about the neo but my kickr I had to do three factory spin downs before it started to give good readings and feel more like my various crank and pedal power meters

daniel_b

Tacx themselves appear to say they expect approx a 12% difference, due to the fact it is going through the drivetrain, as opposed to being measured at the crank.

https://trainerroad.com/forum/t/tacx-support-12-loss-in-power-via-drivetrain-is-realistic/34178

pblakeney

Something else to consider. I wondered why the resistance on my 2T didn't feel right when the gradient went above 14% so I asked the question. Again and again.
I finally got the answer. The 2T is calibrated based on a 65 kgs rider going 10 kph up climbs. I'm guessing the bike weight was the UCI limit. Obviously some differences will be apparent if you are 80 kgs, and compounded if out of the saddle.
I imagine other manufacturers will do similar.

Wheelspinner

itboffin said:

Not sure about the neo but my kickr I had to do three factory spin downs before it started to give good readings and feel more like my various crank and pedal power meters

Pretty sure you can’t even calibrate the Neo - there’s no option to do that IIRC.

Wheelspinner

daniel_b said:

Tacx themselves appear to say they expect approx a 12% difference, due to the fact it is going through the drivetrain, as opposed to being measured at the crank.

https://trainerroad.com/forum/t/tacx-support-12-loss-in-power-via-drivetrain-is-realistic/34178

Ok, thanks that explains a fair bit of it - I’d guessed mechanical efficiency of chain drive was a lot better than that, typically in the mid 90s, so only a few percent loss at most.

Wheelspinner

pblakeney said:

Something else to consider. I wondered why the resistance on my 2T didn't feel right when the gradient went above 14% so I asked the question. Again and again.
I finally got the answer. The 2T is calibrated based on a 65 kgs rider going 10 kph up climbs. I'm guessing the bike weight was the UCI limit. Obviously some differences will be apparent if you are 80 kgs, and compounded if out of the saddle.
I imagine other manufacturers will do similar.

Ah, that sort of makes sense, except the software I use with the Neo is the Tacx Training App - their own one. That’s why I made sure the rider profiles were set up correct as me. Assumption (and implication from the blurb I thought) was that the trainer resistance is automatically calibrated based on rider profile details. I’ve only tried a few rides with gradients in the 14% and up range and they do seem odd, but I’m a lousy climber anyway.

pblakeney

My comments were about using Tacx software as I knew to eliminate that conflict. My test was their Koppenberg. I did a screenshot of the response.

"Slope, video, virtual reality, and GPS
All Tacx bike trainers are spec'd with a maximum incline value. This value is calculated with a body weight of 143 lbs (65 kg) in reference to the "strength" of the resistance unit when cycling 6 mph (10 km/h). It is important to understand that when your body weight is higher, the maximum incline the trainer is able to simulate will be lower."

Not sure it is related to your question but it does point to why some riders will feel discrepancies.

itboffin

Wheelspinner said:

itboffin said:

Not sure about the neo but my kickr I had to do three factory spin downs before it started to give good readings and feel more like my various crank and pedal power meters

Pretty sure you can’t even calibrate the Neo - there’s no option to do that IIRC.

the factory spin down is not an option on the normal app menu there's a sequence you have to do before it appears, worth googling just in case, made a huge difference for me on my Tacx before i sold it for a Kickr

whyamihere

Have you put the bike with the Stages PM on the Neo and run both at the same time on the same activity to see what the difference really is between them? Doing that may give some answers. Make sure it's a ride where you're constantly applying power in case there's something funny going on with the averages when 0 Watts are applied.

Unfortunately if they really don't match then you'd need a 3rd PM and recording device, with the idea being that hopefully 2 of the devices would agree and you'll know which one is inaccurate. Borrowing or hiring something like dual sided pedals would be ideal in case you have a significant L-R imbalance which is skewing the results.

ibr17xvii

Just to throw in my experience - since I've got my 4iiii (left side only) PM I've never been able to get it close to my Kickr V4 at all but outside it seems fine.

Today for example I did a Zwift race. Warmed up for 15 mins beforehand, did a spindown on the Kickr & a calibration with my Wahoo on the PM. Hooked the Kickr up to Zwift & had the PM hooked up to my Wahoo.

The Kickr gave me NP 221w & an average of 214w for just over 35 mins which I would say was about right.

The 4iiii gave me NP 305 & an average of 289w which is miles out, I don't think I'm even capable of that. I've never hooked up the 4iiii directly to Zwift though to see if that changed anything.

TBH it doesn't really bother me, the Kickr is fine indoors & the 4iiii is fine outdoors & effectively that's what I bought them for. I would just be interested to know why the 4iiii reads so high indoors.

Maybe interference with Bluetooth or something along those lines? Genuinely have no idea.

wongataa

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

Wheelspinner

wongataa said:

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

All well and good, but I am not especially interested in the (correct) absolute value of how much power I can produce. I am trying to understand why two devices supposedly calibrated to calculate power can produce such markedly different apparent values.

The Physics is relatively straightforward - power is the rate of doing work. Work in terms of a bike ride is relatively simple to define - if you're on a dead flat road, then essentially all your work effort is expended in overcoming aero and rolling resistance plus mechanical inefficiency in the drivetrain and you, in order to accelerate to a speed and then maintain it.

If you are climbing hills, all the above applies, plus the additional work required to move your weight uphill, calculated by determining for example the change to your potential energy from going up the slope.

Those equations are not that complex.

Power therefore (and especially average power) *should* be easy.

Examples from my recent ride history:

Neo 2T: Ride time 47:54, average speed 24.6 km/h; distance 19.58 km, total elevation gain 194 m, average power 148W

Stages out on road: Ride time 48:30; average speed 19.1km/h; distance 15.46 km; total elevation gain 172 m, average power 175 W.

On the Neo, I rode 20% faster, climbed about 12% more elevation all in the same time within about 1%, and yet the Neo reckons my average power is over 15% LOWER than the Stages.

Given that resistance (aero and rolling) increases with approximately the square of the speed, and I did 12% more work in gaining elevation alone, the Neo readings should be easily above the Stages if both are even only "reasonably" accurate.

I just find it odd that such expensive technology produces such apparently poor results.

ibr17xvii

Wheelspinner said:

wongataa said:

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

All well and good, but I am not especially interested in the (correct) absolute value of how much power I can produce. I am trying to understand why two devices supposedly calibrated to calculate power can produce such markedly different apparent values.

The Physics is relatively straightforward - power is the rate of doing work. Work in terms of a bike ride is relatively simple to define - if you're on a dead flat road, then essentially all your work effort is expended in overcoming aero and rolling resistance plus mechanical inefficiency in the drivetrain and you, in order to accelerate to a speed and then maintain it.

If you are climbing hills, all the above applies, plus the additional work required to move your weight uphill, calculated by determining for example the change to your potential energy from going up the slope.

Those equations are not that complex.

Power therefore (and especially average power) *should* be easy.

Examples from my recent ride history:

Neo 2T: Ride time 47:54, average speed 24.6 km/h; distance 19.58 km, total elevation gain 194 m, average power 148W

Stages out on road: Ride time 48:30; average speed 19.1km/h; distance 15.46 km; total elevation gain 172 m, average power 175 W.

On the Neo, I rode 20% faster, climbed about 12% more elevation all in the same time within about 1%, and yet the Neo reckons my average power is over 15% LOWER than the Stages.

Given that resistance (aero and rolling) increases with approximately the square of the speed, and I did 12% more work in gaining elevation alone, the Neo readings should be easily above the Stages if both are even only "reasonably" accurate.

I just find it odd that such expensive technology produces such apparently poor results.

I think the only way you can judge them is having both meters running in tandem as @whyamihere suggested but in my case that confuses the situation even more.

As I posted above I'm not losing any sleep over it as I "think" they are both right in their respective environment but like you I find it odd.

pblakeney

ibr17xvii said:

Wheelspinner said:

wongataa said:

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

All well and good, but I am not especially interested in the (correct) absolute value of how much power I can produce. I am trying to understand why two devices supposedly calibrated to calculate power can produce such markedly different apparent values.

The Physics is relatively straightforward - power is the rate of doing work. Work in terms of a bike ride is relatively simple to define - if you're on a dead flat road, then essentially all your work effort is expended in overcoming aero and rolling resistance plus mechanical inefficiency in the drivetrain and you, in order to accelerate to a speed and then maintain it.

If you are climbing hills, all the above applies, plus the additional work required to move your weight uphill, calculated by determining for example the change to your potential energy from going up the slope.

Those equations are not that complex.

Power therefore (and especially average power) *should* be easy.

Examples from my recent ride history:

Neo 2T: Ride time 47:54, average speed 24.6 km/h; distance 19.58 km, total elevation gain 194 m, average power 148W

Stages out on road: Ride time 48:30; average speed 19.1km/h; distance 15.46 km; total elevation gain 172 m, average power 175 W.

On the Neo, I rode 20% faster, climbed about 12% more elevation all in the same time within about 1%, and yet the Neo reckons my average power is over 15% LOWER than the Stages.

Given that resistance (aero and rolling) increases with approximately the square of the speed, and I did 12% more work in gaining elevation alone, the Neo readings should be easily above the Stages if both are even only "reasonably" accurate.

I just find it odd that such expensive technology produces such apparently poor results.

I think the only way you can judge them is having both meters running in tandem as @whyamihere suggested but in my case that confuses the situation even more.

As I posted above I'm not losing any sleep over it as I "think" they are both right in their respective environment but like you I find it odd.

Tacx power meter to app on screen and Stages to head unit simultaneously?

Wheelspinner

pblakeney said:

ibr17xvii said:

Wheelspinner said:

wongataa said:

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

All well and good, but I am not especially interested in the (correct) absolute value of how much power I can produce. I am trying to understand why two devices supposedly calibrated to calculate power can produce such markedly different apparent values.

The Physics is relatively straightforward - power is the rate of doing work. Work in terms of a bike ride is relatively simple to define - if you're on a dead flat road, then essentially all your work effort is expended in overcoming aero and rolling resistance plus mechanical inefficiency in the drivetrain and you, in order to accelerate to a speed and then maintain it.

If you are climbing hills, all the above applies, plus the additional work required to move your weight uphill, calculated by determining for example the change to your potential energy from going up the slope.

Those equations are not that complex.

Power therefore (and especially average power) *should* be easy.

Examples from my recent ride history:

Neo 2T: Ride time 47:54, average speed 24.6 km/h; distance 19.58 km, total elevation gain 194 m, average power 148W

Stages out on road: Ride time 48:30; average speed 19.1km/h; distance 15.46 km; total elevation gain 172 m, average power 175 W.

On the Neo, I rode 20% faster, climbed about 12% more elevation all in the same time within about 1%, and yet the Neo reckons my average power is over 15% LOWER than the Stages.

Given that resistance (aero and rolling) increases with approximately the square of the speed, and I did 12% more work in gaining elevation alone, the Neo readings should be easily above the Stages if both are even only "reasonably" accurate.

I just find it odd that such expensive technology produces such apparently poor results.

I think the only way you can judge them is having both meters running in tandem as @whyamihere suggested but in my case that confuses the situation even more.

As I posted above I'm not losing any sleep over it as I "think" they are both right in their respective environment but like you I find it odd.

Tacx power meter to app on screen and Stages to head unit simultaneously?

I might give that a crack, but I am not sure if the head unit will record anything if there's no measurable speed or distance - wheel off trainer so I can't use a simple speed sensor on the back wheel to tell the head unit how fast I'm pedalling. Average power theoretically will be infinity - all that effort to go exactly nowhere

pblakeney

Wheelspinner said:

pblakeney said:

ibr17xvii said:

Wheelspinner said:

wongataa said:

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

All well and good, but I am not especially interested in the (correct) absolute value of how much power I can produce. I am trying to understand why two devices supposedly calibrated to calculate power can produce such markedly different apparent values.

The Physics is relatively straightforward - power is the rate of doing work. Work in terms of a bike ride is relatively simple to define - if you're on a dead flat road, then essentially all your work effort is expended in overcoming aero and rolling resistance plus mechanical inefficiency in the drivetrain and you, in order to accelerate to a speed and then maintain it.

If you are climbing hills, all the above applies, plus the additional work required to move your weight uphill, calculated by determining for example the change to your potential energy from going up the slope.

Those equations are not that complex.

Power therefore (and especially average power) *should* be easy.

Examples from my recent ride history:

Neo 2T: Ride time 47:54, average speed 24.6 km/h; distance 19.58 km, total elevation gain 194 m, average power 148W

Stages out on road: Ride time 48:30; average speed 19.1km/h; distance 15.46 km; total elevation gain 172 m, average power 175 W.

On the Neo, I rode 20% faster, climbed about 12% more elevation all in the same time within about 1%, and yet the Neo reckons my average power is over 15% LOWER than the Stages.

Given that resistance (aero and rolling) increases with approximately the square of the speed, and I did 12% more work in gaining elevation alone, the Neo readings should be easily above the Stages if both are even only "reasonably" accurate.

I just find it odd that such expensive technology produces such apparently poor results.

I think the only way you can judge them is having both meters running in tandem as @whyamihere suggested but in my case that confuses the situation even more.

As I posted above I'm not losing any sleep over it as I "think" they are both right in their respective environment but like you I find it odd.

Tacx power meter to app on screen and Stages to head unit simultaneously?

I might give that a crack, but I am not sure if the head unit will record anything if there's no measurable speed or distance - wheel off trainer so I can't use a simple speed sensor on the back wheel to tell the head unit how fast I'm pedalling. Average power theoretically will be infinity - all that effort to go exactly nowhere

I was thinking the power readings would still be sent, even if live display and not recorded. Less to check average, more to check current. Apparently there are different algorithms used to calculate averages. There has already been posts on Strava having different averages to apps for the same ride.
Speed or distance is irrelevant for this test?

Wheelspinner

pblakeney said:

Wheelspinner said:

pblakeney said:

ibr17xvii said:

Wheelspinner said:

wongataa said:

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

All well and good, but I am not especially interested in the (correct) absolute value of how much power I can produce. I am trying to understand why two devices supposedly calibrated to calculate power can produce such markedly different apparent values.

The Physics is relatively straightforward - power is the rate of doing work. Work in terms of a bike ride is relatively simple to define - if you're on a dead flat road, then essentially all your work effort is expended in overcoming aero and rolling resistance plus mechanical inefficiency in the drivetrain and you, in order to accelerate to a speed and then maintain it.

If you are climbing hills, all the above applies, plus the additional work required to move your weight uphill, calculated by determining for example the change to your potential energy from going up the slope.

Those equations are not that complex.

Power therefore (and especially average power) *should* be easy.

Examples from my recent ride history:

Neo 2T: Ride time 47:54, average speed 24.6 km/h; distance 19.58 km, total elevation gain 194 m, average power 148W

Stages out on road: Ride time 48:30; average speed 19.1km/h; distance 15.46 km; total elevation gain 172 m, average power 175 W.

On the Neo, I rode 20% faster, climbed about 12% more elevation all in the same time within about 1%, and yet the Neo reckons my average power is over 15% LOWER than the Stages.

Given that resistance (aero and rolling) increases with approximately the square of the speed, and I did 12% more work in gaining elevation alone, the Neo readings should be easily above the Stages if both are even only "reasonably" accurate.

I just find it odd that such expensive technology produces such apparently poor results.

I think the only way you can judge them is having both meters running in tandem as @whyamihere suggested but in my case that confuses the situation even more.

As I posted above I'm not losing any sleep over it as I "think" they are both right in their respective environment but like you I find it odd.

Tacx power meter to app on screen and Stages to head unit simultaneously?

I might give that a crack, but I am not sure if the head unit will record anything if there's no measurable speed or distance - wheel off trainer so I can't use a simple speed sensor on the back wheel to tell the head unit how fast I'm pedalling. Average power theoretically will be infinity - all that effort to go exactly nowhere

I was thinking the power readings would still be sent, even if live display and not recorded. Less to check average, more to check current. Apparently there are different algorithms used to calculate averages.
Speed or distance is irrelevant for this test?

True, they are irrelevant.

The Garmin head unit does note that the various display options for power are instant, or some short term moving average, but the overall average should be a straight formula for the completed ride activity.

Will fiddle about when I have a spare hour to swap bike/cassette and use the Stages bike on the Neo. (One is 11 sp, one is 10)

pblakeney

An aside but I recommend having the same set up on all road bikes and turbo.
Cuts down on the required spares and tools. Being able to swap at short notice is convenient too.

N0bodyOfTheGoat

Zwiftpower has an analysis tool to upload and compare power data from multiple sources, which can be public or private.

Wheelspinner

Yeah I'd agree, but have you seen the prices people want for 11 speed components these days???

The 10 speed one is 7800 Dura Ace shifters/mechs stuff I've had for ages. Just bought another road bike basically because it was nearby, and a cheap deal. It has the 11 sp setup with the Stages crank and I can't just swap that onto the other bike because that one has an FSA crank and different BB.

pblakeney

Fair point on the cranks. Can't you just swap chainrings?
TBF, I last bought a groupset in 2015. 😱🤣

Anyway, way off topic.

ibr17xvii

Wheelspinner said:

pblakeney said:

ibr17xvii said:

Wheelspinner said:

wongataa said:

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

All well and good, but I am not especially interested in the (correct) absolute value of how much power I can produce. I am trying to understand why two devices supposedly calibrated to calculate power can produce such markedly different apparent values.

The Physics is relatively straightforward - power is the rate of doing work. Work in terms of a bike ride is relatively simple to define - if you're on a dead flat road, then essentially all your work effort is expended in overcoming aero and rolling resistance plus mechanical inefficiency in the drivetrain and you, in order to accelerate to a speed and then maintain it.

If you are climbing hills, all the above applies, plus the additional work required to move your weight uphill, calculated by determining for example the change to your potential energy from going up the slope.

Those equations are not that complex.

Power therefore (and especially average power) *should* be easy.

Examples from my recent ride history:

Neo 2T: Ride time 47:54, average speed 24.6 km/h; distance 19.58 km, total elevation gain 194 m, average power 148W

Stages out on road: Ride time 48:30; average speed 19.1km/h; distance 15.46 km; total elevation gain 172 m, average power 175 W.

On the Neo, I rode 20% faster, climbed about 12% more elevation all in the same time within about 1%, and yet the Neo reckons my average power is over 15% LOWER than the Stages.

Given that resistance (aero and rolling) increases with approximately the square of the speed, and I did 12% more work in gaining elevation alone, the Neo readings should be easily above the Stages if both are even only "reasonably" accurate.

I just find it odd that such expensive technology produces such apparently poor results.

I think the only way you can judge them is having both meters running in tandem as @whyamihere suggested but in my case that confuses the situation even more.

As I posted above I'm not losing any sleep over it as I "think" they are both right in their respective environment but like you I find it odd.

Tacx power meter to app on screen and Stages to head unit simultaneously?

I might give that a crack, but I am not sure if the head unit will record anything if there's no measurable speed or distance - wheel off trainer so I can't use a simple speed sensor on the back wheel to tell the head unit how fast I'm pedalling. Average power theoretically will be infinity - all that effort to go exactly nowhere

On my Wahoo you can set it to indoor & it will record Time, Mileage, HR & Power but obviously no distance.

Might be worth a shot.

ibr17xvii

pblakeney said:

ibr17xvii said:

Wheelspinner said:

wongataa said:

The only way you can actually see what any differences are is to do a ride on the turbo, record the turbo power numbers and the Stages numbers, and compare them.

Now any differences will be down do the accuracy of each device, the different points in the drivetrain where the power is being measured and your L/R balance. The turbo will be measuring total power. The stages is just measuring your left leg power and doubling it. The further away your are from having a 50:50 L/R power balance the greater the discrepancy from your true power output the number from the Stages will be. Note that your L/R power balance could vary due fatigue levels, power output, and cadence.

Trying to guess any variance due to how different rides feel is not the way to go.

All well and good, but I am not especially interested in the (correct) absolute value of how much power I can produce. I am trying to understand why two devices supposedly calibrated to calculate power can produce such markedly different apparent values.

The Physics is relatively straightforward - power is the rate of doing work. Work in terms of a bike ride is relatively simple to define - if you're on a dead flat road, then essentially all your work effort is expended in overcoming aero and rolling resistance plus mechanical inefficiency in the drivetrain and you, in order to accelerate to a speed and then maintain it.

If you are climbing hills, all the above applies, plus the additional work required to move your weight uphill, calculated by determining for example the change to your potential energy from going up the slope.

Those equations are not that complex.

Power therefore (and especially average power) *should* be easy.

Examples from my recent ride history:

Neo 2T: Ride time 47:54, average speed 24.6 km/h; distance 19.58 km, total elevation gain 194 m, average power 148W

Stages out on road: Ride time 48:30; average speed 19.1km/h; distance 15.46 km; total elevation gain 172 m, average power 175 W.

On the Neo, I rode 20% faster, climbed about 12% more elevation all in the same time within about 1%, and yet the Neo reckons my average power is over 15% LOWER than the Stages.

Given that resistance (aero and rolling) increases with approximately the square of the speed, and I did 12% more work in gaining elevation alone, the Neo readings should be easily above the Stages if both are even only "reasonably" accurate.

I just find it odd that such expensive technology produces such apparently poor results.

I think the only way you can judge them is having both meters running in tandem as @whyamihere suggested but in my case that confuses the situation even more.

As I posted above I'm not losing any sleep over it as I "think" they are both right in their respective environment but like you I find it odd.

Tacx power meter to app on screen and Stages to head unit simultaneously?

Yes that exactly.

ibr17xvii

N0bodyOfTheGoat said:

Zwiftpower has an analysis tool to upload and compare power data from multiple sources, which can be public or private.

Good shout, I didn't know about that.

I've just done it for the ride from this morning I mentioned above & Kickr v 4iiii is miles out but I knew that already.

There is literally 100w difference across the board.

itboffin

That is odd my three different devices are at most 2% out which is amazing considering one is a single sided 4iiii one kickr and the other dual sided pedals, my L/R balance is between 45-50 split so pretty even.

vincesummerskRoxcBTr

If there is 100w difference when you can run them at the same time then one must surely be faulty or missing calculated?

ibr17xvii

vincesummerskRoxcBTr said:

If there is 100w difference when you can run them at the same time then one must surely be faulty or missing calculated?

You would think so yet the 4iiii is fine outside.

Weird.