[BUG] Running nullifies steps


#1

That problem was brought up here many times, but I wanted to make a topic specifically for that.

Walking isn’t a problem - You walk and the Sweatcoin algorithm catches the steps 98% of the time, you just have to double-check that it actually does when you proceed.

But when it comes to running/jogging - Most likely due to the intense shake and the much higher speed, Sweatcoin loses those steps and ignores 1/3 if not 2/3 of the steps further on.

I can’t see how someone can possibly get or stay motivated to get the breaker/troublemaker coins cap when 1/3 of their steps are ignored while running or wasting about 4 hours of their day walking (Which is healthy, but time-consuming.)

Please, work on the algorithms so running will be possible with a much lower loss rate, because it’s a real issue.

Feedback is appreciated and encouraged to blow up the exposure of this problem.


#2

Same Crap, different day, even worse.


#3

Hi @ISh98
Sorry to read of the issues. You’re not the 1st to mention it. I have experienced the problems myself. To be honest mate I think you could be right, it seems most likely that’s the more erratic free movement of the mobile phone From the extra spring in the step when running causes steps to fail conversion. However, although I was having problems where I would lose up to 10,000 steps a day it’s not more, I have managed to near enough resolve the problems by carrying the phone in a different way. I have found that putting the phone in a less loose front pocket or wearing one of those “flip belts” often leads to very minimal losses no matter if running, jogging or walking. I use an iphone XS Max which is probably a little heavier than the newer android models and is quite bulky. The worst ways to carry the phone I found I’ve been in your hand (especially if using while moving) or strapped to your arm in one of those armbands (the phone just seems to swing back and forth around your arm every time you move it). It seems to closer and tighter to your body you can keep the phone, the more reliable results. Put it this way yesterday I completed over 107k steps and lost less than 1000 steps. That was with sections of walking running and jogging.
It’s worth a try, works for me and it has worked for a couple of others.
Please bear in mind that I don’t work for Sweatcoin, I can only share my experiences and my own findings and opinions. So, The suggestion might not be the ultimate fix, however it is one that has worked for me and I hope it can help you out too. :sunglasses::+1:t4:


#4

Thank you for the reply! I will keep that in mind.
Yet, a fix is indeed needed, because not everyone have a flip belt.

I have Ulefone’s Armor 3 which is bulky as it can be (About 400g weight) and still it is woobly in my pocket, and I also have a belt, but it’s been set aside because the last time I used it, it did pretty much the same, but then the steps back then were much more restricted.

I will give it another shot


#5

It’s worth a try mate, I know it isn’t always ideal. However, i’m sure with a little experimentation, you will find the most suitable way. I even went through testing of trying to strap the phone to my thigh in my pocket. Which, although is definitely not ideal and probably not the most sensible method if you want to avoid trapped nerves, it did resolve the problem.

This is something that did frustrate me for a while until I found a solution. Even when I start to get really tired or end up with a bit of a limp, steps to get lost. However, I can understand how the data from more erratic motions could easily be interpreted as a shake rather than a step motion, especially when the phone is bouncing around so much when running.

I have spotted that there are now available, double shorts. Running/gym shorts with compression shorts as a liner and a mobile pocket in the compression shorts. I considered trying them but don’t know if the pocket is big enough to carry the larger phones. Also, with the pocket being on the side of the leg, chances are it could have the same issue with excessive rotation as the arm bands.


#6

I experienced the same issue and also informed admin with no response however :frowning:

The problem however did not occur on older iPhones (6s) with the old algorithm.

With new iPhone SE 2020 and new algorithm, running does not work, only walking…


#7

@Hansi2019
Have you tried alternative ways of “carrying” the phone as I mentioned above? It’s only based on what I found useful myself but I had the same issue when the new algorithm was first introduced. Nowadays I don’t really suffer significant losses whether I’m walking or running even on a ridiculous number of steps. It is worth a try.


#8

Hi Phatt,
in fact I tried nearly all possible ways of storage during my run with no affect.
On my old machine I never had issues regardless of carrying position.
Really strange…


#9

Hmmm The old algorithm did maximise steps but to be fair there were probably a lot of steps counted that weren’t necessarily steps. As I say with the new algorithm, I did experience issues to start with once the pace went up. Although restricting the excess “bounce” did solve the issues for me personally. I won’t lie, it wasn’t an overnight fix and even now I can still have issues when running but 9 times out of 10 it isn’t really a regular problem. Only on the odd occasion. It’s also worth mentioning that if you are testing, it’s probably wiser to test in good 20 minute sessions as forcing steps every couple of hundred steps doesn’t seem to offer the greatest results on conversion even when walking normally.
The carrying of phones and the effect on conversion was actually something I was going to post about a couple of weeks back but managed to mess up my iphone7 when changing the battery and typing on the iphone 5s isn’t so easy when used to one of the plus size mobile. I generally leave the phone with Sweatcoin installed in a pocket or wherever and type on a second mobile.


#10

Thanks a lot for your helpful thoughts! I will give it a try and especially focus on the „excess bounce“ which sounds reasonable!


#11

I tried to run today with a phone strap and it converted at least 99% of my steps.
Tightening/fixing the phone to the body indeed helps with converting step at the best rate*.

*Google Fit steps matched Sweatcoin steps with 99.9% conversion rate on my run.
On my Mi Band 5 though, only 2/3 were converted, but unlike my phone, the Mi Band counts steps non-stop.


#12

Hey @ISh98
Awesome news! Great to read conversions are back up for you.

I think the new algorithm is based more upon physical motion data than the original was. In many ways I think it is a little more accurate once the motion thing is sort of taken onboard and you find your best method. It took me a while to figure out this little niggle, I still don’t get it perfect every day. Although, at least there is a logical explanation and the hurdle can be overcome :sunglasses::+1:t4:

To be honest I actually learned quite a lot of interesting stuff along the way. Reading studies and research papers on biomechanics and the accurate measurement and tracking of lower limb movement. I wasn’t too interested in maximising the number of steps, more the degree of accuracy.

I suppose that as with anything, there will be both advantages and disadvantages depending on how users actually use the app, or rather the focal purpose.

Happy Sweatcoinin’ :sunglasses::+1:t4:


#13

Quite a research you made for a few cents a day, haha

Sounds interesting though, might look into it myself


#14

Lol I haven’t really chased the coins since the first year. Once I’d seen the difference the exercise was really making to ealmost every aspect of my life, I knew I’d already claimed better health and not only restored but exponentially improved mobility as the reward… invaluable :sunglasses::+1:t4:

I’ve always been the guy that will be more interested in the solution than the actual answer… if that makes any sense. I guess a lot of it will be my age but even as a kid back in school, I could never see the sense in just reading and regurgitating facts. I always figured that there’s more to learn from methods and reason than there is to knowing answers. Although I was always pants with maths, which in truth makes no sense at all.


#15

Sry @ISh98 I only half replied. The problem is with the Mi Band that not only will it count ur your steps but also a number of other arm movements that have nothing at all to do with walking… picking up coffee, practising the piano, using an abacus :rofl:
To be fair, although mobile phone pedometers are amongst the least accurate devices, a huge part of the reason I appreciate the work that has gone into the Sweatcoin algorithm development. In most circumstances, Sweatcoin gives a more honest impression of the exercise you go through in a day. Although, the algorithm is probably more in place to protect the integrity of the reward scheme by limiting cheating and actually giving value to the steps you take.


#16

True, Mi band is indeed inaccurate when it comes to movement :joy:

Sweatcoin had and still has a great potential, but where they will go with it? we will see.

So far, it’s mixed for me but I hope it will get better.