After a few months absence from my blog….here is a short story about what I learned after weighing myself for a year using a Withings WS-50 smart body analyzer. I say “a” Withings and not “my” Withings since it was not really mine… it was my Wife’s Wandering Withings (WWW). I bought it for her birthday, in 2014, partly since we didn’t have a scale (and she had been asking for one) and partly since I thought it would be a way to get her more interested in quantified-self (but that didn’t really work). She downloaded the app, connected her phone and used it once or twice, end of story, at least for her.

So, the scale stayed on the bathroom floor for a while, wandering around and finally last year (2016), two years after buying it for my wife, I connected it to my phone and started weighing myself (a belated self-present, as it were). It was very easy to connect, the data went straight to the Withings web page, which was also quite intuitive, and it could all be downloaded easily as a “.csv” file. I weighed myself daily, shortly after getting up, with basically the same clothes on, or rather PJ’s. During the half minute when the scale talks with your phone and sends your weight to the whole world, including those Russian hackers, I would hum to myself, “just standing here weighing myself with my wife’s wandering withings”.

Nokia knocks

A few weeks ago (June, 2017), Withings was acquired by Nokia. The new owners wrote me some nice emails, assuring me that things would not change, but ended up changing things shortly after, which is a shame, since it was all working so nicely. So I had to get used to a new way of doing things (yawn), or at least new fonts and rather barren open white spaces (more Finnish I guess).


So the Withings is a great example of what I was just reading about in Alberto Frigo’s recent book/thesis “Life-stowing from a Digital Media Perspective: Past, Present and Future: i.e., one of the problems with massively produced quantified-self gadgets is that users are forced to use online templates that keep changing, all the time, to prevent users from migrating to other products/apps. The Red Queen anyone? The end result is a disruption in usability, or worse, disruption in data sets. Luckily, I was still able to download my data and the change went well, but others have not been so lucky. (Aside: I had no idea that Withings was originally French, vive la France!, and that they had made this cool video, with naked people…from a respectful distance).

The data

After importing the csv data into Excel, I cleaned it up a bit (eliminating wonky data, some bad readings). Then I made this graph for all the world to see, including my mother, (Figure 1), in Excel (yes, yet another one), which shows my weight variation over time. Modesty aside, my average weight was 71.3 kg (standard deviation 0.701, coefficient of variation 1.0% and range 68.8 – 73.1 kg). Some may say too skinny, some may say just right. In any case, my “heaviest months” were April and May, lightest September and October. That was most probably since I had kidney stones in August (2016), and ate less after that, at least for a while.



Figure 1. Variation in body weight over one year, in someone not trying to lose or gain weight, or do anything really.

But really, a 1% coefficient of variation in weight is not high. And I’m not worried about losing weight. So, you ask, Why in heavens did you weigh yourself every day for a year then? Good question. Well, as I explained, the Withings was just wandering around…. it could happen to anyone… and I was also interested in how much my weight really varied, but for no good reason. Maybe to compete with someone? Anyone else out there have a lower coefficient of variation in weight? I may have set a record. Please let me know.

So, as opposed to most other stories about weighing, I was not thinking about gaining or losing weight. If one looks for videos related to Withings on the web site, there are nine entries. Most people talk about losing weight, and most also suggest concentrating on weekly or 10 day averages. Four of the most interesting to watch are: Akhsar Kharebov (A Smart Scale for Healthy Weight Loss April 20, 2016) stresses that the Withings shifts the emotional stress of losing weight to a quantified one; Julie Price (Family Visits, Races, and Games December 17, 2014) who found the family was mostly to blame (gained weight after family visits); Nan Shellabarger (26 Years of Weight Tracking. November 26, 2014) appears to be the longest “weight-quantifier person” and mostly concludes that she gained weight when she didn’t focus on planning, tracking and cooking meals; and Amelia Greenhall (Weigh Everyday = Understanding weight, January 2, 2012) is one of the first videos about the topic, and was also nice to watch (I like the use of a “physical” notebook).

Time stamps

So, compared to those nine entries, I have little more to add, except for one idea. The most original thing I found playing around with my data had little to do with weight. I figured out that I could use the time stamp of every weighing for other things. I usually weigh myself first thing in the morning, around 6h00, so it’s also a measure of when I sleep in. Aha.

I found that I tend to wake up later weekends (no surprise), but not too much later, like 7h30 instead of 6h00. In addition, during the year I woke up earlier in February and September, both “beginning months” for me, starting new courses etc. So the time stamp provided by the scale also could be used to indicate work load, or an indirect indicator of work stress. Funnily enough, the months when I got up later, I also tended to weigh a bit more. Food for thought, or rather, probably better said as “if you eat more food, you’ll probably wake up later”…Tah da!