I’ve been counting steps for several years now. I started out with a Fitbit One, which eventually died when I dropped it once too often. I replaced it with another Fitbit, but then decided to buy an iWatch.
I do a weekly shop on Friday’s at a local market (South Melbourne Market for any other Melbourne residents out there. Melbourne, Australia…) The market is two stops away on the light rail, and walking around the market usually racks up a fair few steps. Up until recently Fridays were the one day where I could reliably get to 10,000 steps every week. Since I switched from Fitbit to Apple Watch this has not been the case.
I push a shopping trolley around the market, and after some weeks of being disappointed by the step count I was getting after switching to the Watch, I discovered that pushing the trolley two handed caused the Watch to seriously undercount steps. If I pushed the trolley with only one hand (the one whose wrist doesn’t host the Watch) step counting worked fine. I can only assume that the algorithm that interacts with the gyroscope in the Watch expects the user’s hand to be swinging freely.
When I used the Fitbit it was in my pocket, but while at the Market I use the watch to pay for groceries, so putting it in my pocket wouldn’t work for me.
I haven’t been to a gym for years, but I would imagine that the same thing would occur if you went to a gym and used one of those walking machines with handlebars.
While swinging the arm is part of it, my understanding is the main factor the Apple Watch looks for in counting steps is the heel strike (the momentary shock when the heel hits the ground). The effect of grabbing the shopping cart would be much the same though. Because your hand is fixed to the cart, the watch can’t sense the input it needs to register a step.
I recall getting Pedometer++ because I was led to believe that it somehow gives more accurate (whatever that means) step counts than the built-in counts. It’s been a couple of years since then, so I can’t remember the particulars, but I can see in its app store entry (https://itunes.apple.com/us/app/pedometer++/id712286167?mt=8) that it specifies that you should use it with your phone in your pocket, so maybe that’s related. Anyway, I do know that Pedometer++ registered higher step counts for me than I was getting without it.
In Apple Health, set the phone data as a higher priority than the watch for step counting. The phone will pick up steps in all situations because it’s not cradled by the shopping cart. When you’re not carrying the phone, the steps from your watch will fill in the blanks.
Really though just don’t worry about step counts. You did the steps so who cares if your phone thinks you did or not? One of the benefits of the Apple Watch is how it accurately tracks calories, exercise minutes and stand hours. These, in my opinion, are much more closely aligned with your activity levels than a step count.
My understanding is that the step count is one of the inputs into those calculations. If the step count isn’t accurate, the Move and Exercise rings are going to be off as well (see #2 at this Apple Support page, for instance).
I’m not sure it’s that relevant to be honest. It’s more related to your heart rate in my experience. If you go for a long bike ride, you don’t count any steps but you count loads of calories and exercise minutes, even if you don’t tell the watch you’re cycling.
The Health app records seperate step counts for both devices, but as far as I can see it passes on to the Activity app either one or the other, depending on the priority that’s been set, not a mixed count depending. How would it know which one to pick?
I’ve never been able to work out how the watch decides what’s an exercise minute and what isn’t. When I’m walking the dogs I tend to average around 20+ minutes a kilometre, when on my own I can do around 15, but the exercise points I get don’t seem to me to be particularly consistent. I like the step count because steps get counted regardless (except when pushing a trolley…)
I tend to find exercise minutes are only counted when my heart rate is over 100 bpm while walking. At 95 they don’t go up. My resting rate is around 55-60 so it might be a percentage increase they’re looking for.
I don’t know what to tell you with the data sources for steps. All I can say is that I have mine set to phone first, and watch second and I don’t see the problem you are having. I don’t push shopping carts much either, though.
When I walk a mile at a pace that takes about 17 or 18 minutes to complete while wearing my Watch with my iPhone in my pocket, I get credit for burning about 70 calories. When I stand in place and juggle three balls while wearing the Watch and with the phone in the next room, I get credit for burning about 10 calories every minute.
I’m not saying juggling doesn’t require some effort, but there’s definitely something askew.
It’s true, there’s only so much the watch can do with frankly very limited data about the movements of one arm. It doesn’t know you’re only moving your arms wildly and it has to either give you very little credit or assume you’re moving your whole body like that. The latter seems more likely so I guess they go with that.
If you want to monitor calories more accurately you’ll need a tracker on every limb. For most people the calories tracked on one wrist is good enough to compare one day and the next.