This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

What is the population base for Insights?

At 72, I find I am in the top 1% for average weekly running distance. However, if I select any age group other than 70-74, the percentage is the same. Does the comparison with Garmin users of my gender include people who are doing some other sport, like swimming, or golfing? Is the data global (including those who might not be running as much during winter in the Southern Hemisphere? I'd like to see much more insightful data for comparison with my activities, such as:

  • The population size
  • Number of days per week running, by age group
  • Average pace for tempo runs, by age group
  • Average distance per week, by age group
  • Same data as above, broken out for the top 10 countries
  • My guesses:

    - the population is all Garmin users (which matches what the text on the page says).

    I doubt Garmin tries to figure out which of its users are actually "runners" (or cyclists, golfers, etc), but I could be wrong. What would be the threshold for being a runner? Ran at least once with their device? At what point would a threshold that's too high mess up the data itself? For example, you could define a runner as "someone who runs more than 10k per week", but that would clearly be a problem if you're looking to compare your average weekly running distance with others.

    - the fact that it says you're in the "top 1%" for running distance for all age groups is simply because of rounding / simplification. (i.e. I'd imagine that 0.01% 0.1%, 0.5% and 1.2% are all represented as "1%". This is a funny edge case because 0.1%, for example, would be more accurately rounded to "0%", but "top 0% of users" doesn't sound quite right.)

    Same argument goes for the actual phrase they use, which would be "You ran farther than 99% of other users" (at least on the Connect website). If you actually ran farther than 99.982% of users, they're not going to round that to 100% for obvious reasons.