Power Meter Discrepancy

My indoor rides are done on my road bike, which doesn’t have a power meter, but the trainer (Elite Direto X) does, so I use that for TR and setting my FTP. According to TR, my FTP is 244.

I recently added a single sided SRAM (Rival) PM to my gravel bike. I have done two rides outside with the SRAM PM, and both have been 2 hour rides with normalized power of 215 and 225 respectivel, which would be 92IF (approximately). My HR was not sky high, nor did I feel dead at the end.

I am curious what’s going on? Are the power meters THAT far off of one another? I can’t imagine I can ride that much harder inside than out?

What do you think?

I’d attach your gravel bike to the Elite Directo X, do a workout in ERG mode with the trainer connected to TR and you bikes PM connected to a head unit. Do a couple of longer intervals and you can see the difference in real time. You can also save both files and see the difference in each interval.


To save a LOT of time and guesswork, grab a set of Assioma (UNO, or DUO and convert it to UNO mode), or a set of Vector/Rally and convert them to single sided. Then you’re comparing apples to apples.

Next would be to ensure your two recording devices are similar. Two Garmins, two Elemnts, something. Mixing a bike computer and a PC/Mac/etc app that records data can result in different figures overall. Averages for intervals should be very close regardless of how they record data… and speaking of intervals… perform a set of intervals at different ranges/zones. Steady state, over/unders, ramps, peak power, etc.

Compare the intervals. Comparing overall averages isn’t really giving you the true picture a lot of the time.

All power meters should be close*.

*Give or take a few watts… or a few % depending.


Thank you Shane! I’ll see what I can put together to test more equally.

That said, if all power meters should be pretty close… what would you suppose is causing the discrepancy? Just being excited to ride with the boys outside?

Could be a number of things. The Rival unit I had was 6% low (irrc?). A quick upscale and it was bang-on. That was an early unit though and I wouldn’t say with any confidence they’d all need similar adjustment. Collecting more data will make things clearer.

1 Like

Another thing to throw into the mix is your indoor v outdoor environments. The natural cooling of outdoors may mean you can operate at a higher power without overheating, whereas indoors you are overheating which is a limiter. There’s ton’s of other variables too which could be causing a discrepancy.

1 Like

What’s that?

You can go into SRAM’s app and scale up the power readings if you like.

It is essentially the option to apply a multiplier, that will alter the power meter data output. You can shift the data (by percentage?) up or down by a chosen value entered into the approrpiate app.

The most likely reason is to align the power meter with another data source that is known to be good, or otherwise desired to match a source than may not have scaling options.

1 Like

Thanks. It’s a pity you can’t do this with turbo trainers as well so you can match the turbo to the pedals rather than the other way round.

1 Like

Yeah, would be nice in some ways, to have the option from any power source. But there are potential issues with that as well. I could see it misused innocently by changing the “good” power data source to match the “bad” power data since there is not necessarily an easy way to know which is which when you just have two devices in use.

That can be less of a problem for the few devices that allow users to check with static weight calibration and such, but even that is open to misuse if people don’t have very accurate weights and practices when doing the process.

The other “bad” way to misuse it is intentional adjustment by people for reasons like Zwift racing, but that’s a separate discussion.

Maybe we will see more options over time, as the whole Power Data Accuracy side of this world is far more tricky and messy than people and manufacturers might want to admit. :stuck_out_tongue:

Completely agree.

My suspicion is that that is the main reason they don’t offer it on turbos.

Also the manufacturers maintain that their equipment is accurate so “why should we offer away to alter it”

1 Like

This. As soon as you open this to trainers ergos etc it will be wild west.
Point of setting the slope is doing a user calibration. With a known weight for example. Which is impossible on most such devices since most of them do not have strain gauges and/or do not report an instant torque value.
Setting the slope to match two power meters is a last resort. I do it, not proud of it but it is the only way to dynamically calibrate a stubborn pm (mtb quarqs, I am looking at you).
I have the rule of at least 2 good pms approving a third one. Never take a sole pms data for its face value.