[FitzAusTex]

I'll leave others to comment on steps in activities like running and elliptical machines, but after 5 days these are my observations for walking:

1. S2 records about 80% of the steps that my Samsung Gear Fit records, and about 75% of the steps that my Android Wear Samsung Gear Live records.

2. I've always thought my Gear Live was "about right" in step counting, and Gear Fit was a bit conservative.

3. Although mostly correct for any extended walking in the same direction, the S2 seems to not count short walks (like from my living room to the bathroom, and back from the bathroom). In particular, it seems to not record steps where an abrupt about-face in direction occurs.

Obviously, this is less than optimal, and will definitely weigh in my decision to keep or return my S2. I know that this is something Samsung COULD fix, I just don't know if they will.

[winpeguy]

I have a Gear Fit as well. Guess I will wear both tomorrow and see how it counts

Sent from my SM-N920T using Tapatalk

[FitzAusTex]

I have a Gear Fit as well. Guess I will wear both tomorrow and see how it counts

Sent from my SM-N920T using Tapatalk

Hopefully, you won't have to do that in public, but luckily I don't give a crap that I looked silly wearing both at same time, and on same wrist. 😁

[winpeguy]

So I started today with my Gear S2 and Gear Fit keeping my step count. I have also had my phone on me most of the time keeping count as well.

6:14 AM Arrived at Gas Station
Gear Fit 847
Gear S2 641

6:17 AM Walked 140 steps
Gear Fit 986 (counted 139)
Gear S2 775 (counted 134)

6:58 Got to work
Gear Fit 1008
Gear S2 791

7:01 Walked in Office
Gear Fit 1131
Gear S2 906

8:00 AM - Not sure why my phone is that much higher, but I believe the Gear Fit is the most accurate.
Gear Fit 1741
Gear S2 1425
Note 5 1875

Yesterday when I got home from work, my phone showed a step count 1000 higher than my Gear S2. If the count is really this inaccurate, then it for sure is going back until Samsung can correct this with an update. Anyone else think their Gear S2 is inaccurate?

[FitzAusTex]

Winpeguy, your counts are exactly the type I'm seeing between my S2 and my Fit. Exactly. And I have a Nexus 5, so I'm using Google Fit and Android Wear rather than a Galaxy and S Health, but again similar results, in that my Fit has always been just a bit lower than Android Wear.

Due to a shattered heel, step counting is really important to me, and the S2 is just too much off for me.

[edubb256]

I'm seeing something similar. Here is what seems to be the different. The algorithms for the both the Fit and S2 require a threshold before step counting begins. Once it starts, all steps are counted, if it doesn't start none of the steps are counted.Here is quick test I did. with Fit and S2 on same arm:

- Took 10 steps (or anything below 10 steps): Both S2 and Fit record 0 steps
- Took 11 steps: S2 records 0, Fit records 11
- Took 12 steps: S2 records 0, Fit records 12
- Took 13 steps: S2 records 13, Fit records 13
- Took 14 steps: S2 records 14, Fit records 14

So the S2 seems to have a higher threshold (13 steps as opposed to 11) before it kicks in.

[FitzAusTex]

I totally concur that there must be a threshold, and that would explain why short amounts of walking are not counting. Hadn't actually tried to determine what threshold is, so thanks.

That being said, mine is failing to record way beyond that threshold, where my Fit does record. I can take 30 or 40 steps, and those OFTEN get missed on S2. I also think changes in direction can abort counting, but don't seem to abort on the Fit.

Just took dog for walk with S2 and AW Gear Live. Dog stops and sniffs ALOT, and S2 counted 555, and Gear Live counted 838, which has S2 only counting 66% of Gear Live.

My testing has generally found S2 to count around 75% of Gear Live and 80% of Gear Fit. And in my testing since June, Gear Live is about 95% accurate when I've actually counted steps in my head.

So, Gear Live is acceptable to me as a step counter, Fit is too conservative, and S2 is going in the wrong direction from an accuracy standpoint.

[winpeguy]

A quick 300 step walk and here are the results:
Note 5 (in my pocket) 2336 - 2646 = 310 Steps
Gear Fit 2517 - 2824 = 307 Steps
Gear S2 2006 - 2308 = 302 Steps

You may be right about the threshold on steps. So it is not counting my quick trips I guess.

[FitzAusTex]

I'd also be interested in seeing how off your S2 does or doesn't get without your Note 5 on you. The software has some jiggering to do when both devices are counting concurrently. One will overwrite the other, so not sure myself, which is taking precedence, the Note 5 or S2. And God only knows when Fit is also connected to your S Health (if it is), what that's doing to step counting.

[edubb256]

I'd also be interested in seeing how off your S2 does or doesn't get without your Note 5 on you. The software has some jiggering to do when both devices are counting concurrently. One will overwrite the other, so not sure myself, which is taking precedence, the Note 5 or S2. And God only knows when Fit is also connected to your S Health (if it is), what that's doing to step counting.

Each of the devices (Note 5, S2, Fit...) have separate pedometers. Within S Health you specify which one is the source of numbers reported in S Health.

[dnc0716]

I have been wearing my fitbit and s2 simultaneously and the difference is large. I really like the watch, but feel the inaccurate pedometer will make me return it. See differences below.

10/8
Fitbit 12,912
S2 6,254

10/7
Fitbit 12,763
S2 7,858

10/6
Fitbit 10,656
S2 6,535

10/5
Fitbit 4,957
S2 2,422

10/4
Fitbit 10,654
S2 6,958

I can walk over 100 steps and the pedometer will not move. Sometimes it does and sometimes it doesn't.

Posted via the Android Central App

[AndroHK]

I can walk over 100 steps and the pedometer will not move. Sometimes it does and sometimes it doesn't.

Posted via the Android Central App

That final comment is what frustrates me about the step counter. Sometimes it does and sometimes it doesn't. Mine will give up counting during a walking session. I keep walking, it stops counting. Then eventually, after a few minutes, it is back doing its job. Every day about 10% of my steps are not counted.

I am not doubting the threshold thing but think there must be something more.

[alex346]

I have been wearing my fitbit and s2 simultaneously and the difference is large. I really like the watch, but feel the inaccurate pedometer will make me return it. See differences below.

10/8
Fitbit 12,912
S2 6,254

10/7
Fitbit 12,763
S2 7,858

10/6
Fitbit 10,656
S2 6,535

10/5
Fitbit 4,957
S2 2,422

10/4
Fitbit 10,654
S2 6,958

I can walk over 100 steps and the pedometer will not move. Sometimes it does and sometimes it doesn't.

Posted via the Android Central App

That's really strange ... I use a fitbit everyday as well as my gear s2 and they are always pretty close... usually within a few hundred steps and I average anywhere from 7000-9000 steps daily

Even when I do a run on our treadmill the s2 auto logs the run into s health and it's numbers are always pretty similar to what the treadmill reads out for distance travelled and whatnot

Posted via the Android Central App

[e jen]

Regarding the treadmill run...when I walk on the tread, I have to keep my hands lightly on the rail for balance (due to a surgery), but I walk moderately fast, and when my arm is up, because of holding on, my Samsung gear s2 doesn't count at all the e
Whole time. The same occurs when I push the grocery cart. Real frustrating!!!

[icicle22]

Regarding the treadmill run...when I walk on the tread, I have to keep my hands lightly on the rail for balance (due to a surgery), but I walk moderately fast, and when my arm is up, because of holding on, my Samsung gear s2 doesn't count at all the e
Whole time. The same occurs when I push the grocery cart. Real frustrating!!!

Same experience when I run with the dog. My arm is out holding the lead instead of swinging freely and it misses a ton of steps. The sad thing is I don't feel that this was an issue on the Gear S that I used to have. Bummer.

[NuMotiv]

While I understand you want a several hundred dollar thing to work properly...do exact steps matter that much? The point isn't to tell you the bathroom was 10 steps away. Its to get you up and active. If its the weekend and I see 3000 steps when I average 13000 at work, maybe I should move.

[Hudsoner]

I compared my new S2 with the pedometer in my Galaxy S6 phone and found a lot of difference. I went to a large auto show, and at the end of the day the phone showed 17896 steps and the S2 showed only 14 560 steps! That is a big difference in counted steps. My previous watch the Se2 Neo was pretty close to the phone with it's step counter. This shows that Samsung can do it, but just seems not to be very interested in having fitness as a major feature in mind with their S2. I was sad that I had my Gear 2 Neo to let go, but I developed a bad skin rasch from the plastic case and band.

[paradox60660]

Regarding the treadmill run...when I walk on the tread, I have to keep my hands lightly on the rail for balance (due to a surgery), but I walk moderately fast, and when my arm is up, because of holding on, my Samsung gear s2 doesn't count at all the e
Whole time. The same occurs when I push the grocery cart. Real frustrating!!!

Put it on you ankle.

[tsparks1]

Put it on you ankle.

lol

[elron24]

Would like to ask for your opinion: which one do you prefer, Gear S2 with Fitness Apps (Skimble, Nike+, Pear) or Gear Fit 2 (which is solely made for fitness)? I'm a bit torn between the two. I do interval running, but do some workout in the gym too.

[munchiniwa]

For me, the pedometer function is the most important. My Gear S2 is also WAY off. I wear it with my FitBit Charge and the Gear is reading 5038 and the FitBit is 7022. When I take a step, I can see it increment on my FitBit instantly, one for one. When I check on my Gear S2 it seems to register about 3 out of 4 steps but not exactly -- more like 7 out of 10? This is OLD technology. I can't believe that Samsung cannot get this right. I'm also shocked that there is no settings for the pedometer -- length of stride, height of user, which hand (dominant or not). Thought this was supposed to tie into Samsung Health but instead it gives the impression that the whole program is buggy.

[Hudsoner]

I am pretty disappointed with my Gear S2. I come from a Gear 2 Neo, which, I agree, looked pretty ugly but had way more/better functions than the Gear S2. I feel it had a way better step counter and a very good and, with the phone integrated sleep monitor.
One should have thought that this more modern and more expensive Gear S2 would have at least similar functions or even more advanced versions of them.
I hope that Samsung will do some improvements with new programing.

[johnmcc0710]

I have always been a fitbit guy but got this Gear S2 free when I bought my wife a new phone. I was told I could sync it with my fitbit app but that doesn't work. My steps on this do seem low.

I ordered a fitbit one on Ebay so I will use that for fitness and the Gear S2 for everything else. I do like it and feel it is very useful. I will compare both. I have a charge HR I may wear for a day and compare.

Posted via the Android Central App

[cmb713]

I'm waiting to hear from someone that the Gear S2 4g works good as a standalone on a treadmill using Nike. I'll stick with my Apple Watch for now but hoping Samsung really fixes it based on feedback from multiple users who take the watch out alone on the treadmill.

[mrm_mrm]

I'm waiting to hear from someone that the Gear S2 4g works good as a standalone on a treadmill using Nike. I'll stick with my Apple Watch for now but hoping Samsung really fixes it based on feedback from multiple users who take the watch out alone on the treadmill.

I always use my Gear s2 3G version with GPS to go run and save my track. And I can say that is too weak, the GPS signal is weak, it takes some time to have satellites and always comparing the same route sometimes mark 4km other times 3km in same traject and not quite right with the phone track. So the Gear S2 3G no good for training.