I’ve been on a Fitbit kick for a few months now, and have spent quite a bit of time getting to know the Fitbit One, which I really love. After I lost my second one, however, I started to think that maybe I should get a Fitbit Flex instead, since they’re harder to lose, being strapped to your wrist.
I don’t want to spend too much time discussing the merits and flaws in the various personal health trackers. I had a Jawbone Up and hated it. To be fair, I got one of the early ones, before they issued the refund for anyone who wasn’t happy. For those who are curious, the manual jacking-in to sync and the cap for the jack that always ended up lost were the main reasons the Up was a massive letdown for me, and why I like the Fitbit so much more.
I also have the Aria scale (love it), which lets me weigh myself and automatically logs it to my Fitbit dashboard, so I don’t have to face the shame of manually entering it. And I’m not a runner, so the Fuel band didn’t make much sense. I’m super-excited to learn more about the Amiigo, but I really hope it works with the Fitbit API so I don’t have to lose the single-dashboard awesomeness with the Aria and everything else.
The Fitbit’s developer API platform is a big reason I went with them. They integrate well with lots of other apps and services, and that’s pretty cool. This one in particular seems rad.
So first the basic difference:
Features both the Fitbit One and the Fitbit Flex have:
- Pedometer – although the One is meant to be worn on the torso, the Flex is a wristband
- Sleep tracker
- Silent alarm
- Access to the Fitbit dashboard which allows you to log food, water, other exercises, friend rankings, etc.
- Bluetooth syncing with iOS app and also to your computer with a USB wireless dongle
Features the One has over the Flex:
- LED display that shows step count, calories burned, etc
- Altimeter, to automatically track stairs climbed
Features the Flex has over the One:
- Lives on your wrist – put it on and forget about it
- Arguably does a better job calculating non-walking, arm-related activities like swimming
- Persistent connection link with your iPhone, so you get a real-time step counter display, versus periodic syncing with the One
The Flex measures steps by the swing of your arm, not the movement of your torso. This seemed like it could be a problem for me for a few reasons. First, I don’t always swing my arm while I’m walking (if I’m carrying something, walking the dog, etc.) — and second, I’m Italian and talk with my hands all the time. No, I mean All. The. Time.
I was afraid the hand-talking would over-inflate my numbers by counting non-step, hand-wavy business, and that I wouldn’t get credit for legitimate steps I did take if I was using a handrail or for whatever reason wasn’t swinging my arm. (Spoiler: you don’t.)
I expected the Flex to show inflated numbers overall, since I move my hands a lot.
I should mention, I’m already starting this with the notion that the One is probably more accurate, simply because of how it calculates steps, and because short of physically, manually calculating every actual step throughout my day, I have to assume one of the two of them is “more right”. My goal here was to determine how big of a difference there is and whether specific activities are more prone to erroneous step-counts than the other. Having used the One for a few months, I am very comfortable with its overall accuracy.
Naturally, I decided to test it out. I’m currently wearing both the Fitbit One and the Fitbit Flex at the same time, and have put together a spreadsheet of my findings, which I’ve been logging at various times during the day. I will continue this experiment for a week, updating this post (and the spreadsheet) with new data points.
A few points of interest: I am wearing the Flex on my right (dominant) hand, and have it listed in the dashboard as being on my dominant hand. I wear the Fitbit One clipped to my bra, between my boobs, which is one of the recommended placements for it. (Insert TitBit joke here.)
The Flex does not seem to erroneously register typing as a step, which I’m glad for. I type very aggressively, so that would have been an instant deal-breaker for me.
Also, the graphs below were captured using data from a Tuesday, which is my work-from-home day, which is why the step level is much lower than other days you’ll see in the spreadsheet over time. My step-count is always very low on Tuesdays.
Here’s the raw data so far from the spreadsheet:
And here’s a breakdown of the data represented in each column:[checklist]
- Date: date of entry where both the One and Flex were synced at the same time
- One: step count from the One
- Flex: step count from the Flex
- Delta: difference between the current step count of both (One – Flex)
- One Variance: difference between the current step count of the One, and the step count for the One from the last log entry
- Flex Variance: difference between the current step count of the Flex, and the step count for the Flex from the last log entry
- Variance Delta: difference between the One Variance and the Flex Variance (One Variance – Flex Variance)
- Notes: contextual notes on the activity performed
While the Delta is interesting by itself, the Variance Delta is potentially more interesting to me, since it will help me understand which activities tend to register with particular wide gaps in step counts. You’ll see the Variance Delta in the block immediately after “Dinner+drinks” is the highest so far, and dinner and drinks is a period of almost no walking, but lots of hand-wavy, enthusiastic talking.
As you can see, overall, the Fitbit dashboard graphs look pretty similar:
What’s more curious is that the Flex seems to think I had 34 Very Active Minutes, which I can sadly assure you I did not. The One thought I had 3 Very Active Minutes, which is closer but frankly, still might be pushing it. (I told you, I worked from home. Don’t judge me.)
Breaking it down, it looks like this so far (taken from the spreadsheet):
I will continue to log my data, and annotate any particularly unusual activities for context.
Keep checking the spreadsheet, and check back here in a week or so for my conclusions. Leave your questions or thoughts in the comments, or ping me on Twitter @snipeyhead.[box type=”note”]UPDATE: I have been logging from 7/15 to 8/5, and so far the difference in steps totals almost 50k. The Fitbit Flex has logged 47,784 more steps than the One, which I believe supports my theory that the Flex over-attributes steps. I will keep logging for the rest of the month, and we’ll see where it shakes out. You can view the raw data and updated charts here.[/box]