Stepping out - A question of algorithms
I have two devices and three algorithms that count my daily steps...
(The fitbit died)...
An Apple Watch...
The iPhone health app...
The Fitbit app on the phone...
Never do the three agree.
I can understand the watch being different from the other two, I don't normally leave it on the desk when I get a coffee, I rarely wander off leaving it on the side.
The other two are using the same hardware - the phone, with its accelerometer and inbuilt gizmos...
How much of a difference? Well, I've had an interesting day at work, where I've been shackled to the desk, have been feeling rough and headachy - so a good wish test, no runs or gym sessions...
The watch - 8315
Health App - 7713
Fitbit - 7106
That's 10% difference between the two on the same hardware.
The next question is should I be bothered? The scientist in me is, I want to unpick what the difference is and why it's so stark on the same hardware. But there is a bigger question, how accurate are step trackers anyway... The answer is range of + or - ( at the extremes of) 10% Tracker accuracy
The even bigger question is does it matter - simply put something is better than nothing, but 10,000 steps isn't a magic number. Most people don't walk 10,000 steps, its closer to 5000, so walking 10,000 is a big increase and a large time commitment.
The 10,000 figure is an artefact of the Japanese olympics, in the mid-1960s... But as a rule of thumb for most people doubling your steps is a good thing Beeb - 10k steps
Why do I keep saying most people? Because not everyone can do 10,00 steps, not everyone can do 5,000. I'm not going to go into wheelchair exercise, because I'm not competent to do so, but I've seen more than a few campaigns for those if us with dicky tickers where our cardiologist calibrates the target, making sure that we slowly build up our steps towards a safe target. I also know a few who've been given an activity tracker to give a max figure, to ensure that the stay safe.
If in doubt, ask you doctor. They should be able to help, to point you towards someone who can.
So, 10% inaccurate, based on a dodgy genesis and with limited evidence for a threshold effect - why do I it... because its fun, and on days like today its that little nudge to get out in the cold and wet to get more steps in. Will I make my main target of 14,000 - maybe, maybe not.
One last point before this ramble concludes, there is a special ring of hell for those who cheat on their steps...
(The fitbit died)...
An Apple Watch...
The iPhone health app...
The Fitbit app on the phone...
Never do the three agree.
I can understand the watch being different from the other two, I don't normally leave it on the desk when I get a coffee, I rarely wander off leaving it on the side.
The other two are using the same hardware - the phone, with its accelerometer and inbuilt gizmos...
How much of a difference? Well, I've had an interesting day at work, where I've been shackled to the desk, have been feeling rough and headachy - so a good wish test, no runs or gym sessions...
The watch - 8315
Health App - 7713
Fitbit - 7106
That's 10% difference between the two on the same hardware.
The next question is should I be bothered? The scientist in me is, I want to unpick what the difference is and why it's so stark on the same hardware. But there is a bigger question, how accurate are step trackers anyway... The answer is range of + or - ( at the extremes of) 10% Tracker accuracy
The even bigger question is does it matter - simply put something is better than nothing, but 10,000 steps isn't a magic number. Most people don't walk 10,000 steps, its closer to 5000, so walking 10,000 is a big increase and a large time commitment.
The 10,000 figure is an artefact of the Japanese olympics, in the mid-1960s... But as a rule of thumb for most people doubling your steps is a good thing Beeb - 10k steps
Why do I keep saying most people? Because not everyone can do 10,00 steps, not everyone can do 5,000. I'm not going to go into wheelchair exercise, because I'm not competent to do so, but I've seen more than a few campaigns for those if us with dicky tickers where our cardiologist calibrates the target, making sure that we slowly build up our steps towards a safe target. I also know a few who've been given an activity tracker to give a max figure, to ensure that the stay safe.
If in doubt, ask you doctor. They should be able to help, to point you towards someone who can.
So, 10% inaccurate, based on a dodgy genesis and with limited evidence for a threshold effect - why do I it... because its fun, and on days like today its that little nudge to get out in the cold and wet to get more steps in. Will I make my main target of 14,000 - maybe, maybe not.
One last point before this ramble concludes, there is a special ring of hell for those who cheat on their steps...
You can do it a lot easier than this - attaching it to your dog seems to be common... as is just swinging it about... All you do is cheat yourself, and also demoralise people trying to walk their step legitimately. So, stop it!
On which point, I'm going to have another cup of tea, watch a documentary and think about going for a walk...
TTFN
Paul
Comments
Post a Comment