I have a conundrum. my stages sb20 reads a consistently lower power number as my outdoor bike’s power meter. Since it is consistent, the indoor training went fine.
however, no that i am transitioning to outside, i am getting red days after workouts when i really wasn’t stressed terribly hard
Not sure what to do and how to resolve this. todays workout which was somewhat based on steamboat was a solid sweet spot effort but due to the meter issues it registered much harder than that
pretty significant. a good 20 watts. Enough that TR AI thinks that i spent most of my time @threshold
If the offset is consistent I wonder if you could game the smart bike by manually adjusting the crank length until the power is closer to what you get outside?
Are you 100% sure it’s the power meter?
It’s a pretty well known phenomenon that outdoor and indoor efforts are different for most cyclists.
Intervals.ICU let’s you set an indoor FTP.
“As we touched on earlier, most riders will produce more power and record a higher FTP on the road than the trainer.”
What is FTP and why does it matter for cyclists? | BikeRadar.
I’m not sure how true that is. There are lots of factors, but I’d definitely start with the power meters - the Hunter Allen article you linked touches on this under Calibration as a precursors to all the other factors.
I have a similar problem when I go out on MTB (left crank powermeter): the power is slightly higher than my Zwift Hub and even my road bike Favero dual. In my case I think it is because my left leg usually produces the 52-55% of power.
I usually avoid to do important workouts on MTB, but is a bit annoying, because I train for XCM and I need to spend time on my mountain bike!!
But in the other side, I don’t want inaccurate power activities on my Trainnerroad…
Not easy solutions… Well, change powermeter to a more accurate one?! 
You’re really going to try to refute the guy who founded Trainingpeaks? Let’s hear your argument.
Wouldn’t the simple answer be to set your FTP 20 watts higher?
You already have, and it’s in his article - I’m pretty sure my post was very clear on this.
Some PM software have an “offset” function. I know Sram/Quarq has.
If that’s not the case. Another - annoying - alternative would be:
a. identity your PM standard and the offset % you need.
b. use a tool to edit your activity. Intervals.icu does this. It will be annoying editing your activity and re-upload it.
Since you’re using a dedicated indoor bike, I’m not sure of a simple solution here. 
Can I assume that you’re using the Stages app to check firmware and calibrate your SB20 as often as recommended?
What about for your pedals?
It’s tough to find a way where this works well with the software if you’re significantly overshooting the power targets when riding outside, but the intervals are appropriate inside.
An accurate smart trainer would be a good, easy option. 
I’ll dig into this a bit more and see if I can find a solution of some kind, but if the discrepancy is as large as it seems to be and things are all calibrated as best as possible, there might not be a cheap, easy solution. 
thanks for looking into it and asking at home base
for now i am just ignoring the adaptions and hoping AI will adjust to the higher power of the outside workouts
we shall see
Stay in touch and let me know how things go.
I’m here if you need any help! 
Can I bump this thread with a very similar problem.
I have been testing my SB20 trainer power against my two pedal power meters the Garmin V3 and the Favero AD MX Pro.
I found that the Garmin vs SB20 were inconsistent. While the Favero’s were pretty consistent and within a very close tolerance. However I had to change the batteries in the SB20 PM’s and it started to go south.
I swapped Panasonic CR2032 for Energiser CR2032 and was getting a difference from -6 to +23. I have put in new Panasonic’s after a little bit of Googling and 2 rides in and the difference is +11 to +13.
So it means the Sweet Spot ride was a Moderate as opposed to a hard which is what I and AI was expecting.
My feeling is to just keep at it, mark the rides accurately and let AI do it’s thing. I’m getting easier rides and a higher that expected FTP, I guess that’s a win-win.
I have same problem. Since my power meter is showing around 4% higher I have set my scaling factor to be 0,96 so it fits with my home trainer. So perhaps similar could be a solution if you only use a left side lower meter?
FWIW, I had tons of trouble with my Stages initially. It took me a long time to figure out that cheap 10 or 20 packs of batteries off Amazon didn’t work well. Even oem packs of Energizers didn’t work. (Maybe they are fake?)
What did work was retail packs of batteries - Duracell, Panasonic, Energizer. I’ve been buying the Panasonics lately off Amazon but only the retail pack and not the bulk industrial packaging.
I don’t think this is the cure to your power reading “mismatch” between meters. And I’m not saying this is “the problem” or even “a problem” in this case, but SRAM has this to say about CR2032 batteries - in this case when used in SRAM shifters. But I wonder if this may be contributing to your inconsistent results with various batteries.
Please note that batteries with a bitter child-safe coating can cause interference between the battery and terminal, resulting in potential performance issues.
I wasn’t implying that it was a cure for mis-matched power meters. I was just confirming that using crappy batteries in my Stages led to inconsistent results.
I’ve used Duracell bitter batteries in my Stages without a problem.
I just think that Stages in general are voltage sensitive and while a 20 pack of inexpensive amazon batteries might work in all sorts of things, they don’t work well in Stages.