Garmin HRM PRO Power Readings

I think it's the same for any watch if you compare power readings derived from the wrist movement (watch) and torso movement (HRM strap). I noticed the same on my Forerunner 965.

I personally consider the HRM strap power more reliable, as trying to estimate running power from a moving wrist is incredibly difficult. Probably each person will have a little different wrist-based power readings, as each persons arms and running style is different. It's the same kind of thing as comparing Stryd power with Garmin power, both give you somewhat different results.

But I'm not saying that wrist-based power is useless (if you accept it's probably less accurate, but HRM strap based power is not totally "accurate" either, because there are no real power meters for running). It's just that if you want to use running power (power zones etc.), it's probably better to stick to either wrist-based power only or strap-based power only, then the results can be compared with each other.

I've noticed that also other running dynamics figures are different between wrist and strap (like vertical oscillation, for the same reasons as power), and seem more reliable using the HRM strap.

So, if I want to pay attention to power (as I usually do), I run with the HRM strap. If power is not important (like in trail running, where there's no way to get reliable running power), I might leave the strap at home (but usually don't as it also gives more reliable HR data, especially HRV data).

The last time I looked at this, I found that the difference in pace measurement explains the difference in power readings.

forums.garmin.com/.../a-quick-test-of-wrist-power-vs-hrm-pro-power-for-running

I am running with stryd and AW so the pace is the same on both instances. 

Edit...

I will try setting the pace to HRM tomorrow to see if what you saw is present in my case. But what I cannot understand is although in the case of a treadmill there could be differences between wrist and HRM, outdoors the GPS supposed to be more accurate and the HRM itself is calibrated with said GPS data, so...

I've been running outside, and my pace has always come from GPS (I've set HRM Pro pace to "indoor only"). So there should be no pace differences in my wrist-based power vs. strap-based power.

If I understood Etupes25 correctly, He thinks that the power generated on HRM is based on the internal pace calculations.

Etupes25 said:

forums.garmin.com/.../a-quick-test-of-wrist-power-vs-hrm-pro-power-for-running

Somehow this post did not pop up when I searched for related posts before creating this thread. Thank you for a very detailed post.

The pace and power measurements stand on their own and the pace data is not an input to the power model. 
When you let the watch measure pace by itself you get however some insight in the sensor data that is fed to the power model to evaluate the horizontal power.

Of course for a valid comparison, you need to make sure that the real pace is actually the same, which is why running outdoors is problematic for this question, versus using a treadmill (notwithstanding running form variations that I  cannot eliminate).

I'm pretty sure the actual power calculations are always done on the watch. So the watch uses whatever pace it is selected to use.

Even Garmin's own running power page (https://www.garmin.com/en-US/garmin-technology/running-science/running-dynamics/running-power/) says "Speed from the watch" and lists HRM strap as a potential source in other places.

Bitti said:

So the watch uses whatever pace it is selected to use

I don’t know for sure.

It seems like the models are set up in a specific way, independently of what pace data is written to the FIT file. 
For power, it seems that either the watch is calculating the power using its own sensors, or is sourcing power from the strap, which uses its own sensors, but not the pace data that the watch is storing concurrently.

i have the same assumption for the VO2 model: it seems to be using the GPS pace data regardless of what sensor is feeding pace to the user/FIT.

I don't of course know for sure either, but I would be surprised if the HRM strap calculated the power and used it's own step-based pace estimate. Because remember that HRM Run, HRM Tri, and the Running Dynamics Pod don't provide any step-based pace, and they can be used to provide running power. So for those, the calculation has to be performed on the watch using GPS pace, anyway. So it wouldn't make much sense for Garmin to make HRM Pro(+) to use another power calculation method based on step-based pace which they otherwise only use for indoors running by default.

I agree with you on that VO2max probably always uses GPS pace for outdoors running (which makes sense, since VO2max estimation is not interested in rapid changes in pace, where step-based pace is better than GPS pace).