Power comparison with Vector 3 Pedals

Luko79 might have confused accuracy with precision (see en.wikipedia.org/.../Accuracy_and_precision ). However, he was absolutely right about the following: Any device that measures power at the rear hub has no way of knowing how much power was lost in the drivetrain. They do not estimate this, they just measure force and revolutions where they are, and this, of course, with the accuracy and precision stated by the manufacturer.

In addition to the drivetrain, oval chainrings are known to cause a slight overestimation of power in pedal based power meters (at least in those that do assume a constant angular velocity internally, like the Vector 3s)

I know what is accuracy and precision :) (just some of the manufacturers are not)  these trainers (or pmeters ) have only claimed!!! accuracy but lots of them is far from accurate because the factory calibration is faulty (these trainers or powermeters have no any industrial accuracy cerftification like METAS or similiar, maybe /if the manufacturer is using/ dynamic torque calibration bench that has such a certificate, for example the Power2Max is using such a bench for every NG model and these powermeters are accurate and precise too not like the other toys on the market which glued a plastic onto a crank and already called powermeter...) , the Neo (if that is in perfect condition) and between a good precise and accurate powermeter (like Power2Max NG) there are always at least 3-4watts difference as drivetrain loss on small ring(34) with round rings, and 5-6watts on big ring(50) with straight chain line on bigger rear cogs, if You are using it on smaller rear cogs the difference is getting bigger because the loss is increasing on smaller rear cogs heavily, if the drivetrain condition not perfect (bearings in bottom bracket, the tightness between crank and frame on certain system, bearings in the pulley wheels, the chain, its lubing) these difference is bigger , could be even 15-20w easily, and depending on rider too, because if the rider is pedaling against one of his leg to the other /not pulling with left  when the right is already pushing, it happens often with lots of amateurs if the position is not good, sitting too high for example/ a small power (watt) is adding to the crank/pedal's measuring unnecessarily, but the rear wheel cant measure this pedaling imperfection and the difference will be higher than usual. (let me tell you an example, my girlfriend has 5-7w difference sometime on small ring with NG powermeter on her bike and on Neo, if i try her bike (we have totally same bike setup, so same chains, BBs, chainrings,sprocket, just the frame size is different) immediately after her,  i have the same 3-4w as with my own bikes with my NG powermeters, if she sit back and try again, she has the same 5-7w difference  as at 1st try, and we change only the "rider" (and she has good pedaling style just sitting a tiny bit higher than she should...), but in worst situation this difference could be bigger with other riders...)

The NEO wont add this drivetrain loss to its own measuring because it is a not constant value so it is measuring the raw power at the rear axle, so if somebody says his trainer (Neo or any other "accurate" trainer) and bikepowermeter is absolute matching there is some issue with the trainer or with the bike powermeter beacuse must be a difference ...  If you could put your pedals/cranks directly to NEO's axle (so without the normal bike drivetrain parts) and pedaling like this and comparing the 2 measurements, You should get exactly the same numbers but You cant do this and must accept that THERE IS DIFFERENCE BETWEEN THE 2 POWER MEASUREMENTS if you are using a direct trainer and a bike powermeter. (but of course, if this difference very high , or unusual, then must be investigated, need to validate your bike powermeters with known weight and/or using a 3rd accurate powermeters at same time which help to determine which one is right and which one is wrong.

Yes, the oval is modifier too, depending of which type of oval, the OSYM has the highest modifier, the Rotor's ovals a little lower and depending on the used powermeter too (some of them overmeasure with ovals, some of them not /its the rare/)

the other mistake what the users do that pairing the dual Assiomas via BT to the computer for Zwift for example but forgot to set the "unified L/R channel setting" and with the default setting the computer via BT receiveing only one of the leg (if i'm remember good only the left one) measurements doubled and if the rider has leg imbalance like 52-48 then the measured in Zwift will be 4% higher than the real power just due to a wrong setting and the difference between bike powermeter and trainer will be so huge...  
 

 This is a very comprehensive summary. I like that you could actually measure this imperfection, BECAUSE of the limitations all these power meters face.

drivetrain power loss is approximately 2-3%. Add that.

The engineers dont like the word "approximately"... :)  

But the 2-3% is just an average, sometimes lower, sometimes more, dependinging on the used bike (gear system) also,

here it is an example, with 53/39 chainrings on 2ways system, with 250watts at 95rpm,

You cant use "2-3%" because it is much more on certain gears with certain drivetrain type:

I suggest sell that (You shouldnt have bought it ... ) worthless Stages what You have  and buy an accurate and precise dual-leg powermeter, with a left side (or right side) only powermeter You will have just problems :) All my friends sold their Stages years ago and bought  (on my recommendation) Power2Maxes or Assioma Duos and after that they had no problems with differences or anomalies... 

Here is a few comparisons between ACCURATE trainers and bike powermeters, as You see , usually there are 6-8watts difference between them (these datas are coming from Zwift races so i suppose the riders were using big ring /52 or 53 probably/ front and middle at rear somewhere and the trainer difficulty was set to a low number (maybe at 0%) instead of 100% to avoid the frequent gear changing )

where the difference is just 3-4watts with Neo and P2M NG that is mine with ERG mode so i used small ring front (34) and 19T at rear all the time, i have a new bike with 53/39 with another NG too , i will test it on the winter that the 39 vs 34 (and 53 vs 50 ) is how many watts difference in drivetrain loss.

So the "pro" Zwifters could understand and accept that they have this difference always during indoor train/race , i dont know why so hard to understand for other people... :) 

the link to comparisons :   https://ibb.co/mC6cLp2

i just got my neo 2t (used to have a kickr) and feel that power reading was on the low side.  i hope this will be fixed soon

4-5% could be valid if You are pedaling at 100-120watt. if it is 20-25watts then it is bad news. Do You have comparison , can You share FIT files with the 2 power sources to check this?