Is Quantified Self Worth It? (Post #1)

I’m inclined to be a huge fan of Quantified Self. Taking entrepreneurial initiative, using numbers, even Meetup groups – it’s right up my alley! I’ve been doing it for years, including many months of daily 30 second recording, a few months of tracking every meal, and some self-designed Randomized Double Blind Controlled Trials (No, that’s not a typo. Yes, double blinding by yourself is tough). I’ve also seen several people try personal things at various stages of complexity, talked to people at the meetup groups, and watched many of the highlighted QS videos.

What I’ve taken away from it is that the human body is complex. Moreover, a human life is complex. So you can try to say “I ate carrots from days 1 to 10 and felt terrible, then avoided them from days 11 to 20 and felt great”, and infer that you’re allergic to carrots. But think about all of the things that could go wrong in this minimal example:

  • Maybe some nutrient that’s relatively high in carrots was accumulating in those first 10 days, and was actually the only reason you felt so good in the last 10
  • Maybe you ate celery instead of carrots in the last 10 days, and that’s actually the only factor that matters
  • 10 day cycles means that one of your samples probably had two weekends in it, while the other had just one. Maybe the difference in averages is simply because you feel different on weekends
  • Maybe there was a party at work, and you ate something you don’t usually, which contained the actual allergen
  • Maybe it has nothing to do with allergies, and responds to the weather
  • Maybe it has nothing to do with anything external, and relates to hormonal cycles (which also happen in men)
  • Maybe your definition of “felt terrible” shifted as you acclimated, or relative to other events in your life during the experiment

There’s also the issue that we often can’t directly measure the thing we’re interested in. Energy levels, focus, even “pain” are hard to put a consistent number on. If the true effect is small, it can easily be swamped in the difference between your definition of a 6 and 7 on whatever scale over time.

Finally, there are a near infinite number of interventions you can try, and just randomly guess-and-checking them is going to make for slow progress. There are some where your priors can be high enough to make them worthwhile (lactose intolerance is a good example), but once you get down into cutting out trace ingredients, or unconventional sleep routines, or anything like that, you’re looking for a needle in a haystack. Also: https://www.xkcd.com/882/

Also, very few doctors know what to do with your data. It makes sense, since few of their patients are arriving with anything like it, and those who are probably bring totally different formats to address different problems.

Still, you’re going to live your life today anyway, you might as well take some quick notes about how it went. At the very least, it might form the baseline if you decide to try a more significant intervention down the road.

Leave a Reply

Your email address will not be published.