[Jeff Allen1]

The Gear S2 step tracker is horribly inaccurate. I went for a walk with my wife and her FitBit on Sunday and she ended up with 6,000 steps while I only had 3,000.

For my 2 week Gear S trial period, I was averaging 10,000 steps a day but, with this Gear S2, I'm getting half that.

[alex346]

I thought the step counter on my s2 was off as well so today I tested it out for myself. While I was out on my run I stopped at a local school and walked around the track there. I literally took and counted out 1000 steps and compared it to what my gear s2 recorded. To my surprise when I looked at the pedometer on the s2 it had registered 998 steps. I was a little surprised because I thought on my day to day it seemed off. But coming in at 998 when I counted out and took 1000 steps is more than acceptable. I'm wondering if during day to day use its having trouble picking up steps after I get up from being seated? Maybee it's taking to long to kick in and start recording steps?

Posted via the Android Central App

[FitzAusTex]

The problem with the S2, for me, is that the threshold for counting steps is apparently 11. So if you take 10 steps and pause, the S2 will likely not count those. And if you take 10 steps 100 times a day, that is 1000 steps that are uncounted. Taking a walk where there are no pauses, the S2 seems to do pretty well.

I found that my S2 only counted about 75% of what my pretty accurate android wear watch counted, but I pause very often.

[supersoulfly]

I feel like it's been alright. It's in line with the expectations I established with my last watch (Pebble Steel). Do around 12,000 steps a day with my usual routine and that hasn't changed.

[alex346]

The problem with the S2, for me, is that the threshold for counting steps is apparently 11. So if you take 10 steps and pause, the S2 will likely not count those. And if you take 10 steps 100 times a day, that is 1000 steps that are uncounted. Taking a walk where there are no pauses, the S2 seems to do pretty well.

I found that my S2 only counted about 75% of what my pretty accurate android wear watch counted, but I pause very often.

Ahh that makes sense now... I was wondering why on earth when I got up out of my chair at work just to go to the next office over or the lunch room or whatnot many times it seemed like the watch didn't kick in counting steps. If it only starts tracking after 11 steps thats unfortunate because It seems very accurate once it starts counting. But like you mentioned you can easily accumulated over 1000 missed steps throughout the day at that rate.

Posted via the Android Central App

[FitzAusTex]

Ahh that makes sense now... I was wondering why on earth when I got up out of my chair at work just to go to the next office over or the lunch room or whatnot many times it seemed like the watch didn't kick in counting steps. If it only starts tracking after 11 steps thats unfortunate because It seems very accurate once it starts counting. But like you mentioned you can easily accumulated over 1000 missed steps throughout the day at that rate.

Posted via the Android Central App

Yeah, the Samsung engineers must be happy with the threshold, but I hope we can convince them to lower it. They could just change the algorithm, or perhaps create a setting for Agressive, Moderate, or something like that.

[FitzAusTex]

Still very disappointed in the step counter of the S2. For the past week, have been wearing both an Android Wear watch and S2 on same arm (I'm at my home, so only dog could comment on silliness) and S2 is generally about 25% lower. S2 does fine for, and only for, walks that are continuous. Stopping and/or changing direction seem to be the cause of the undercounting. Android Wear seems to count short walks much better. I often walk 10 to 20 steps many, many times per day that don't end up being recorded. Lots of that this morning, and S2 is at 1341, and Gear Live is at 2025 which is 66% of Gear Live.

[rickl43065]

Works for me just fine. Gear S2 Classic

[FitzAusTex]

I agree it counts steps, but not accurately. I have two S2s, and they both undercount. Rather disconcerting to accurately go from 10k steps per day to 7.5k.

[cindylou1028]

Thanks for this thread. I've found that on runs, walks, and the like the S2 does a great job with steps. However, with everyday 5-steps-at-a-time walking it really stinks. I teach in an anatomy lab and spend 4 straight hours on my feet walking from station to station (6-8 steps per walk), and after an hour I'm receiving push notifications that I need to get moving. :0 On average, short walks are 20- 25% of my steps. I've had to adjust my steps goal down to 8,000 from the 10,000 goal I had on Fitbit.
Is there some thought that Samsung will tweak this algorhythm?

[Delirious D]

I was curious about that too. Does it register my gps location or accelerometer? I woke up this AM watch was on zero steps. I walked to my living room and it still said zero. I told myself i will just google it later. I started doing some chores i looked at the watch said 70 steps.

So the walk and pause conclusion makes sense. How does it work if i have my watch and my phone. Which steps are recorded?

[Delirious D]

Just saw this too. Its been 24 hrs since last update.
O_o
How do i sync them? I cant figure it out.

[Blknight96]

The Gear S2 step tracker is horribly inaccurate. I went for a walk with my wife and her FitBit on Sunday and she ended up with 6,000 steps while I only had 3,000.

For my 2 week Gear S trial period, I was averaging 10,000 steps a day but, with this Gear S2, I'm getting half that.

Why your fitness band might be lying to you video - CNET

[KahneFan]

I'm not too worried about 10 or less steps. If I get up, take 10 or less steps, then sit back down - I probably didn't burn that many calories anyway. What I am concerned about though is; does it accurately count steps when pushing a shopping cart or pushing a stroller? Also, if I am holding my little one with the arm my watch is on, will it count those steps since I'm not really moving that arm in a "normal" walking motion? You could replace baby with a clipboard or tablet or many other items you may be holding stationary as you walk through a warehouse or around a job site or whatnot.

[Mznik60620]

It appears that the update has made the step count more accurate.

[B3cca1973]

I've been noticing this and came to look it up because I looked at the step count - 617, got up from my desk and walked around the library putting away books (children's librarian, so I walk a lot, but lots of stops and starts as I pull books, talk to people, help with computers, etc.) and some other tasks. Because of what I was doing, I literally walked around every shelf, but had to stop frequently. I came back to the desk and looked at the step counter - 617! I am really frustrated with this thing!

[Nikkiekoala]

With the recent update I have found that the step count has greatly improved. With my gear fit I easily got to my 10,000 step goal. Since I've had my s2 I have been hard pressed to achieve 8,000 steps. Since the update I am getting my usual 10,000 steps per day.

[ShaggyKids]

Where are you guys getting this Step Counter app? I tried a search in Sammy store and nothing popped up.

[TreoRock]

It's built-in the watch OS. Search for it in the watch.

Posted via the Android Central App

[AndroHK]

My Samsung S2 step counter undercounts by about 10% in any kind of walking.
It tends to freeze for short periods. Any steps taken in that time are lost.
It must be doing something in the background when the counter freezes but I don't know what.
The update that is mentioned here, does that download automatically or do I have to go looking for it?

[AndroHK]

I went for a walk with my wife and her FitBit on Sunday and she ended up with 6,000 steps while I only had 3,000.

The difference between the measurements for you and your wife may not be only about the difference between the S2 and the Fitbit although I'm sure that is part of it. If she happens to be shorter than you she will take more steps to cover the same distance.

[AznCajn]

And guess what, the new Fit Gear 2 activity tracker also has a 10 step threshold. Nobody could tell me anything about it, I even went and got another FG2 and it was the same, thinking my unit was defective. Guess those engineers aren't changing anything but it does bother me that there is usually a 2000 to 3000 step difference between my Note 4 and the FG2 (although I believe that my Note 4 overcounts steps often) and I've been tracking steps on my Note 4 for a while now.

[tsparks1]

I've worn numerous wrist based step counters, or pedometers if you will, and the one thing I've observed with ALL that I've used (FitBit Surge, Garmin Vivofit, Basis Peak, JawBone etc.) is that you can also get a lot of false steps recorded depending on your arm movements, or lack of arm movement as well. As an example I work part time stocking shelves for a retailer, this morning my S2 showed me at 6000 steps in slightly under an hour while there is no way I walked that many steps. Compare that to a different type pedometer that is worn at waist level (FitBit One, Google Fit and probably others) Right now my S2 is showing 9984 steps while Google Fit, which is based more on actual body movement rather than arm movement is showing 6,405 which I'm sure is much closer to my actual step count. I just don't put too much stock for any wrist based pedometer unless it is only used while actually walking, if that makes sense.

[John-117]

I just got the Gear S2 a couple of weeks ago, and the step counter is horribly bad, not just in terms of the 10 step buffer, but in the fact that it counts when it's not supposed to. When I get on the bus or in my car, I look down at my wrist and see the step count going up at an almost comic rate. By the time I get I'm going (usually 2-3 miles away), my step count will be close to 1000 steps higher than before, even though I hadn't taken any.