This is a snippit about pedometer accuracy from DCRainmaker.com - he does a ton of sports related technology reviews and is very in depth (probably to a fault).
Obviously, the goal of an activity tracker is to track your activity. But, that becomes a slightly more murky proposition than most realize. See, all of these devices depend on accelerometers for tracking motion – and thus, they have to filter out what is steps versus other random motion. For example, if I’m sitting at my desk and drink a 64oz Slurpee, is that mistakenly identified as steps? It’s motion, right?
If you look at any activity tracker on the market, you should take none as an absolute measure of exactly how many steps you did that day – down to the step. Instead, you should look at trending over time. In virtually all of the units I’ve found that if I go out and just walk 100 steps, it counts it just fine as 100 steps.
Rather, it’s my day to day wanderings that I get more variation between units. For example – how does one unit count a quick 5-10 minute ride on a Velib bike share bike to a local store? On the cobbles, are those triggering steps? And turbulence on a plane?
Thus, for fun, for the last 10 days I’ve tracked across 4-5 activity trackers to see how they’d chart. Obviously I don’t know exactly who is right, nor who is wrong. There are some cases where it’s clear one is wrong (for example, why the Basis appeared to take a vacation on the Wednedsay for a good chunk of it).
I thought it was interesting how much variation I saw in the first half, versus the second half. I have absolutely zero idea why there was that much variation initially, and the closest day overall was actually Tuesday the 11th – astoundingly close, especially given a run was involved.
this section is called counting steps and accuracy.