Wyze Band Step Count Issues

Agreed about their is variability, but using 4 different devices, the only one that is significantly off is the Wyze band. Since I know the other 3 track similarly, and it is easier to wear 2 bands on 1 arm, I chose to compare it with the Garmin Vivofit. I walk every morning for a few miles.

Wed: Wyze - Vivofit steps
7:45 am 7244 - 8602 Walked
9:45 am 7619 - 9159 (UPDATED Firmware on WYZE band)
10:30 am 8222 - 10134
6:00 pm 9871 - 12573 Walked

Thurs: (so far)
10:00 am 8620 - 9542
I walked around the house in short bursts then sat down, I manually counted the steps:
Actual Steps - Wyze - Vivofit ( arms swinging and some cases not)
80 - 35 - 78 steps
124 - 59 - 120 steps
201 - 168 - 204 steps
Some of the observed variations originally are due to the counter not catching up but I always take a few extra steps to insure that it counts.

Assuming this is not a lower quality 3-axis accelerometer utilized in the Wyze band, the algorithm can be tweaked over time and would be repairable. Otherwise, it won’t be a keeper :frowning:

2 Likes

Yesterday just before I jumped into the shower my step count was in the 500 step range. After I dried off and headed upstairs it was over 1000 steps. Since then I have been watching it. I have noticed that sometimes it will record 30+ steps when I am standing still and working with my hands in just a few minutes.

I have been using Samsung health on my phone for years and i average about 6000 steps per weekday. The Wyze band is registering over 10000 steps per day.

We donz’t realky want to know how you get the high counts in your shower.

As of yesterday, the Wyze band undercounts over the course of a day around 20% which is too much especially for those that require data for insurance premiums. This morning, I was tracking the vivofit against the Wyze band. In order to accurately determine not counted steps, watch your band switch and as soon as it shows an updated count, count your steps for 200 counts and wait for it to catch up/update. Consistently, I observe 40-50 counts just not registering. Vivofit with 2. The solution is not using you cell phone to update the incorrect numbers as suggested rather how the 3 axis accelerometer is interpreted in defining a step. I will give it another week, and hope that Wyze updates the firmware to address this broad variability. I was really hoping this would be a disruptive design as their other products have been for the most part. :frowning:

I’m getting about 10% lower than my Fitbit. I can’t try my phone because it is rarely on my body these days. I agree - that is too much of a difference when the steps go to the insurance company.

I am working with Tech support concerning the unit being off so much 12-23% off. They gave me three things to do, which I did, but I have already done that previously several times. I notice that I walk around several hundred steps and no steps are shown on the band. The sensitivity, or something in the hardware is off. Right now this is only a cool watch or toy with changeable clock face. Steps are way off. Alexa is spotty at best, Getting messages works good. Battery life on phone dropped hours in total time before charging. This is my 1st time participating in an early release with Wyze. I have done many others. What can be expected with a product that is not up to par? Do they continue working it? IF hardware needs to be changed (i.e. the accelerometer used) do they replace it? Refunds? Any help understanding as I am probably going to take it off now, put it in my dresser and just have the vivofit on so I have reliable step counting for insurance. purposes.

I had a sneaking suspicion my Wyze Band step count was inaccurate, (low) since I’ve been wearing fitness trackers for some time. I know about how many steps I take on an average day. So, today I wore both the wyze band and a Fitbit Inspire HR.

Fitbit read 9206. Wyze read 8691.

By my math, that’s 6% low!

Should I just accept that the Wyze number is consistently wrong? I mean… not cool.

Getting my activity right is kinda like the main thing the band should be good at, right???

Before you all ask, both bands were worn on my non dominant hand in the location and tightness recommended by their manufacturers.

Thoughts?

Your wyze band seems closer to your Fitbit than mine! My wyze band reports about 2,000 steps lower than my Fitbit Charge 2. I don’t know if my Fitbit is accurate though…it gives me steps when I’m not moving sometimes as well.

The last test I did I walked 250 steps my Fitbit said 271 and the WYZE band said 241, neither were correct but the WYZE band was technically closer, I am not sure how close they are expected to be. People should also test versus an actual count since in my test the band was only 9 off of actual but 30 less than the Fitbit.

1 Like

I did notice a slight improvement on the latest beta firmware v1.0.6.43. When I mean slight, its now only 500-700 steps off compared to more than 1000 on the previous versions. Still horrible accuracy though :joy::joy:.

One thing that is important to note is, Early Access is more or less a beta program. This is the time when our feedback counts the most. AFAIK, the previous testing phases were mainly hardware, this phase is more software testing.

I definitely think that the accuracy can be improved if Wyze offers a mechanism to calibrate the band to the user.

People keep saying EA is basically beta testing, I disagree. Early access is just that, you’re getting early access to the product. I would love for someone to post a link to where Wyze has posted saying EA is us helping work out the bugs, I’ve been on these forums since the beginning, I’ve never seen such a post, although I may have missed it. That’s the purpose of the testers prior to release.

Btw, I don’t expect any fitness tracker to be 100% accurate, sure some are more than others but there is a lot of variables to consider.

I normally do 10-16k steps a day. it is consistently off by nearly 20%. Always ran the latest firmware in these tests. One interesting observation that I need to validate further is that the Wyze unit counts much less steps swinging my arms like I normally do on my walks, The vivofit & fitbit excel in counts. When I place the Wyze band in my pocket 1 whole day, the counts were within 5% with all devices!. It sounds like perhaps the algorithm utilized on Wyze, negates counts more aggressively to not count arm swings. I then ran a test of 500 steps, with all devices in my pocket and they were all within ~5% of the actual steps. Perhaps a clue to the inaccuracy of the Wyze band?

FWIW, wore Wyze, Fitbit AND an Apple Watch today. Fitbit and apple were within 100 steps over a total of anout 6000. Wyze was nearly 1000 steps off from Apple and Fitbit.

I guess my experience is in the minority here. The issue I am seeing is that my WYZE band is counting many many MORE steps than my iPhone Health App!

Now, I know that my iPhone undercounts my steps. It has always been low compared to my brother’s iPhone when we’ve taken the exact same hike at the same time, But my Band is counting way too high, and this must partially be due to the fact that it is wrist-worn and counts a lot of non-walking activity as steps.

Here are my observations, so far;
Steps are averaging 93% higher on the WYZE (yes, almost 2X)
Miles are averaging 132% higher on the WYZE (more than 2x)
Calculated stride length for the WYZE (Miles in feet/steps) is running 17% higher on the WYZE, but is extremely consistent at 2.57~2.58 feet/step. The iPhone varies more per day, but averages 2.19

The number of steps varies dramatically, from as little as 11% high to as much as 4X as high, while the miles varies from as little as 31% to as high as 3X.

Not sure what to do with this data other than share it, and see what happens over time. And I have not seen any trends or measurable changes with firmware versions (currently on V1.0.6.43), or App versions (currently on IOS V2.10.70)

You should also try testing by manually counting say 500 steps and seeing what the others log in comparison

1 Like

I have to agree with everybody in that the step counter seems wildly inaccurate, I’ll get no steps actually walking around but get steps standing at a counter or doing laundry. I usually get at least 50 steps putting it on. So I’m not even sure what it’s even measuring.

Mine is doing the same. It’s counting arm movement not steps. Can it be fixed. Maybe, Lessing sensitivy to arm movement would help. Otherwise it just a watch. But, I like the get up and stretch notice.

I think it is important to keep a few things in perspective. Step counters vary wildly among people. What works for me, will almost certainly not work for you. It’s something to try as many choices as you can, find one you are happy with and stick with it.

A good background article to better understand why we see differences is here:

1 Like

I mean, to me personally the step counter doesnt really matter since I got it for the notifications and alexa. But thought I would chime in. Good article though.

Understood, and no worries. I know a tri-athlete that uses a FitBit of some kind. He swears by it. His wife who is also a tri-athlete uses a Garmond device because the Fitbit is wildly inaccurate for her. That’s why I jumped in. For some people the Wyze step counter is going to be fine. For others not so much. It would be ridiculous to expect all step counters to be accurate for all people. We all have different strides, wearing styles, use cases and so forth. If the feature is inaccurate for you then move on. If it works for you, well Bobs your Uncle!

2 Likes