At a normal walking pace, a tracker lands within 3% of your true steps about half the time, and worn on the wrist it tends to run low. Push a stroller or lean on a walker and the count can come in a fifth to a third under. So your daily total is a good estimate, not a measurement, and you can find out how good yours is with a 500-step walk and about five minutes.
I’m not a doctor, a nurse or a pharmacist, and this isn’t medical advice. A step count isn’t a diagnosis. If your doctor gave you a walking goal, their instructions come first.
How a Tracker Decides You Took a Step
Your tracker has a small motion sensor inside, and it’s looking for a repeating rhythm. Each time it sees the swing and bump that looks like a step, it adds one to the display. It never actually sees your feet.
That’s the whole reason the errors happen. A wrist tracker is watching your arm and making a guess about your legs. If your arm swings and your feet don’t move, it may count steps you never took. If your feet move and your arm stays put, it can miss steps you really did take.
Bumps, arm motion and shaking can all look like walking to a sensor. With big, exaggerated arm movements, wrist trackers read more than 3% high in 2 of 3 comparisons. Folding laundry, washing dishes and riding down a rough road are the usual suspects. You can’t stop every phantom step (I don’t think anybody can), and a handful a day won’t change anything. If your count looks oddly high on a day you barely left the couch, there’s a setting worth checking.
Makers know this, and some give you a setting for it. On a Fitbit, the wrist setting changes how sensitive the step count is: the dominant setting turns the sensitivity down, and the non-dominant one, which is the default, turns it up. Check the manual for your own model, because it’s worth two minutes to set correctly.
Which wrist you actually wear it on matters too. Over one full day with 12 people, the dominant wrist averaged 1,253 more steps than the other wrist, though the Fitbit in that test showed a smaller gap of 613. Either way, pick a wrist and stay with it.
The Error Bands, in Plain Numbers
Here’s the fine print nobody puts on the box. Fitbits came within 3% of a hand count in only 46% of 191 lab walking tests, which is closer to a coin flip than a sure thing. When they missed, they nearly always counted too few: more than 3% low in 51% of the tests, and more than 3% high in just 3%.
Where you wear it changes your odds. On your wrist, expect a count more than 3% low about 70% of the time. Clip it at your torso and it does better, landing within 3% in 65 of 114 comparisons.
Out in regular daily life, the spread gets wider. Across 20 comparisons against research-grade step counters, 55% were within 10%, 30% read more than 10% low and 15% read more than 10% high.
Against people counting by hand, Apple Watches averaged anywhere from 47 steps under to 39 over, depending on the test. And across 10 different trackers on a treadmill, the average miss ran from 0% all the way up to 26.4%, so the brand and model really do matter.
The honest way to read your display, then, is with a band around it. About 10% either way is a fair place to start, so if it says 7,000 steps you likely took somewhere between 6,300 and 7,700. Even that band held in only a little over half of those daily-life comparisons, and if you spent the afternoon pushing a cart, I’d widen it on the low side. A display that’s running 25% low and says 7,000 means you really took about 9,300, and that’s a lot of walking to leave off the books.
Slow Walkers Get Shortchanged
This part matters most if you’ve slowed down a bit. Trackers do their best at a normal or brisk pace, and they get worse as you slow down. On that treadmill, people walked at about 2, 3 and 4 miles an hour, and the trackers did better at the two faster speeds than at the slowest one.
The Fitbit numbers show the same slide. At slow speeds the devices read more than 3% low in 12 of 23 comparisons, and at very slow speeds it was 19 of 26.
So if you walk with a cane, you’re recovering from something or you just take your time, your tracker is more likely to be shortchanging you. That doesn’t mean you aren’t doing enough. It may just mean the display is missing some of your effort, and that’s worth knowing before you get discouraged.
Carts, Strollers and Walkers
Push a walker or a stroller and your wrist stops swinging. The tracker sees a lot less motion even though you’re covering the whole hallway, and a shopping cart locks your arm the same way. Movement like that read more than 3% low in 19 of 24 tests, and restricted or stop-and-start movement averaged 21% to 35% under the true count.
Moving the tracker off your wrist is worth a try, but it isn’t a sure fix. At a normal pace, a tracker clipped at the waist stayed within 3% in 24 of 30 comparisons. At a very slow pace it can still read low, and in 2 of 3 daily-life comparisons with older adults who had limited mobility, a waist tracker ran about 25% under.
I’d test it both ways on your own walk, with the cart or walker in your hands, and keep whichever spot reads closer. A phone in your pocket is one more spot to try, if it counts steps.
If you use a walker every day, I’d treat the wrist number as a rough floor. Your real total is probably higher, and your doctor or physical therapist can help you pick a target that fits.
Read more: What is The Simplest Fitness Tracker for Seniors?
Rings Can Run High
A smart ring has an even tougher job, since your finger isn’t your foot and it isn’t your hip either. Over 14 days with 32 people wearing an Oura ring alongside a pedometer, the ring averaged 13,446 steps a day against the pedometer’s 11,322. That’s about 2,100 extra steps, or roughly 19% high, and the gap on any single day averaged about 50%.
A ring can still be a nice way to see whether you moved more this week than last. A pedometer isn’t perfect either, so the ring may not be quite that far off. I just wouldn’t use it to settle a bet about whether you hit 10,000.
Test Yours on a 500-Step Walk
Here’s the test, and you don’t need anything fancy to do it. Pick a flat stretch where you can walk without stopping. If your balance isn’t great, bring someone along or stay close to a rail.
Write down the number on the display, then walk while counting every step out loud or in your head, and stop at exactly 500. It’s easy to lose your place around 300, and if you do, start over instead of guessing. Check the display again and subtract. Now do the math: take the difference from 500, divide it by 500 and multiply by 100, and that’s your percent error.
Say your tracker added 470. You’re 30 steps short, and 30 divided by 500 is 0.06, so it’s reading about 6% low on that walk. If it added 510, it’s about 2% high. I’d call anything within 5% a good result for daily use.
Do it at your normal pace first, then again at a slower pace, or while pushing a cart if that’s a regular part of your day. Two or three short tests will tell you more about your own device than any review can, because they’re measuring your gait and your wrist.
What to Do With the Number
Compare your tracker to itself, not to your neighbor’s. A step count from the same device, worn the same way, is good at showing whether this month is better than last month, even if the exact total is off by a few percent. Switching brands mid-goal will make your numbers jump for reasons that have nothing to do with you. The same goes for changing wrists or changing that sensitivity setting, so once you’ve got it set, leave it alone. I’d also write the number down once a week and look at the weekly average, because a single low day doesn’t mean much.
There’s no single normal range for daily steps that fits everyone, and I wouldn’t get hung up on hitting a round number. If your doctor gave you a target, ask whether they want it counted on your tracker or some other way, since a slow walker’s display may run low.
Walking should never come with chest pain, fainting or trouble catching your breath. If it does, stop and call your doctor, and call 911 if it’s severe or doesn’t ease up. If you live alone, call someone. No step goal is worth pushing through that.
Your tracker is a helpful nudge with an error band attached. Use it for the trend and don’t sweat the last few hundred steps. If you’ve tried the 500-step test, I’d love to hear how yours did in the comments.
Read more: Calories Burned on Your Fitness Tracker: How Accurate Is the Number?
