@Maz_Fletch so here you go. Your problem is HR. I could show examples of rubbish HR on Sense 2 every day. A few hours ago my bike ride:
In general, I don't trust wrist heart rate monitors (not necessarily optical because PolarOH1 and the latest Polar Verity Sense are on-par with chest straps but they are not meant to be worn on the wrist). This time I went for a small experiment. I paired my H10 chest strap only with Edge 1030+ and let the Fenix 7 record HR using a wrist sensor. I wasn't expecting impossible and it's well visible I didn't get anything even remotely impressive. Yet still, it follows the chest strap a lot better than Sense 2 (although, for me, this is not enough). I'm a bit lazy to compare only moving times and unfortunately, I forgot to disable auto-pause on the Edge device. Hence, the Edge AvgHR is higher (shorter period of moving time and nearly 100% accuracy). Fenix 7 seems to perform worst right after pauses (or, in the case of both watches, stop and rest periods as they weren't paused like Edge was). However, the majority of the time it catches up with the chest strap (probably "good enough" for calories and AZM but not for any serious training). Sense 2 on the other hand does whatever it wants. There are visible holes which means that it couldn't read my HR at all. I don't expect 100% accuracy when cycling using a wrist sensor. For casual cycling, I guess Fenix 7 did quite a decent...ish job. I would need to either test them together with auto-pause... oh wait, that feature has been taken away on Sense 2 🤣... well, so the only way is not to pause (or never stop). I'll give it a go tomorrow and maybe will add Galaxy Watch 5, although GW5 is a pain when it comes to extracting activity data.
So this is about Sense 2 and HR accuracy - probably the worst sensor in the market.
Another thing is weight training and calories. Despite your Galaxy Watch doing a better job reading HR it very likely overestimated the energy expenditure (unless you were not taking any rests). Here's why. Weight training is quite specific and unless it's a CrossFit or very fast HIIT (but then it's less weight training and more cardio-ish activity) most of the time you'll spend resting. The actual energy expenditure takes place when you are performing work (work in physics terms, joules, watts etc.). To lift N kg from the floor you need to generate a certain amount of energy and this is what you're going to burn. Once you dropped the weight and went resting the power output drops. Your HR will still be elevated but you're not burning more because you're resting (so you're being granted "false calories" :)). This may even out when you start a new set because with the first rep, your power output is the same as with the last rep but your HR is increasing over time. It means, the watch will give you fewer calories for the first rep and more for the last one despite all the reps needing approx. same energy from you to be performed. So it's a lottery. Can't cheat physics 🙂 The most accurate result when it comes to calories is steady-state activity when your HR stays at a certain level and it kind of "attaches" to your steady power output for a longer time. Sense 2 in this case is wrong when it comes to HR but do this: add calories from both devices, divide by 2 and you're on the "safe side" 🙂 I know, sounds stupid but there is a chance you'll be closer to your real energy expenditure 🙂
Best Answer