This is my sorta-fake pet peeve this year. Has anyone else noticed how bad the gramin goal steps algo is? I’m talking in particular about my 920XT, but I assume it’s the same algorithm across similar devices. The algo basically has no long term memory and ‘penalizes’ you for missing your goal for a couple of days in a row.
Case in point: Last Thursday my steps goal was 13458 steps. In the ensuing seven days, I logged 110022 steps (average daily 15717), which is 16% more than I needed to hit 13458 seven times. And yet, in that time, my steps goal dropped from 13458 to 12894, a 4.2% decrease.
Okay, I know this doesn’t matter, and honestly I don’t really care. But is there anyone from Garmin on here who can comment? I assume the algo is simple linear regression, but that only one or two days of data are used to make the prediction for the next. Has anyone else noticed how ridiculous this is, or am I alone here?
Well, Garmin connect can’t even give me simple power/heart rate averages over time period, so I’m not so sure about that
Okay, maybe change my ‘can/should’ above to just ‘should’.
Honestly the daily steps goal should approach the year to date daily average steps. I’m having fun plotting those values for myself and watching the curves not converge at all.
I just enjoy when I’ve had like 3 weeks of hard, almost daily training, with some two a days, and I take a day off that I normally workout, and garmin pops up a message “oh you’re not moving as much today” lol I don’t really notice what the steps are telling me.
B) What were the totals for the individual days, not the average? If you did almost nothing the last day or two then it will adjust down. Mine changes daily, but I don’t really look at it. I like having the total number of steps per day but it’s more of a “oh look, cool” at the end of the day.
When I got my Garmin 235, I was most impressed with Garmin Insults (er… I mean Garmin Insights). It was happy to tell me that my least active day was Sunday and maybe I should stop being such a lazy turd and go out and do something.
Even though I received the watch on a Wednesday and hadn’t even owned it on Sunday yet. So much awesome.
I’ve got a Garmin Vivoactive HR and noticed that I got my move counter clear by not actually moving on more than a few occastions. Driving home it has happened a few times as it somehow is registering the back and forth of the steering wheel as walking…or when a couple of times washing or rinsing my face in the sink too. Then there is the time I was walking my infant son (holding his two hands) and walked back/forth for 10 minutes in my living room…move/step counter decided nope…I’m not gunna count that because F-your couch!! (for your Charlie Murphy fans). .
One of my cars has a very stiff suspension, it seems to be just the right amount of bounce going down the road that it counts as a step. Got 25k steps on a recent 8 hour road trip where I got out of the car twice to pump gas, and did nothing else.
One of my cars has a very stiff suspension, it seems to be just the right amount of bounce going down the road that it counts as a step. Got 25k steps on a recent 8 hour road trip where I got out of the car twice to pump gas, and did nothing else.
Probably got a few KOMs too
My 920 tells me to move during bike rides sometimes. Pretty sure it’s not supposed to do that, makes me laugh when it happens during a VO2 max interval on the trainer, I’m dying and dripping sweat and apparently it’s just not enough
My last half marathon was more than enough to hit my goal steps by 9am and I kept getting “Move!” Messages for the rest of the day.
Yeah, I also got those after an 11 hour MTB race last summer. Hilarious.
Yeah, well, their move algorithm isn’t that complex. It doesn’t look at what you did earlier that day, it just says move if you’ve been stationary for a certain time. It also only considers longer periods of motion to be movement (I think it’s around 3-5 minute of motion), which will result in it occasionally telling you to move when you’re already walking around.
