I am using the DarkSky API to get the local weather forecast.

Here are some different statistics for the probability of rain:

  1. They say the chance of rain today is 85%.
  2. The average probability of precipitation over the next 24 hours is 1%
  3. The sum of the probabilities of precipitation over the next 24 hours is 21%.
  4. The maximum probability of precipitation over the next 24 hours is 4%.

Here is a bar graph showing the probabilities (rounded decimals, not percents). Showing this to the user is not an option:

Bar graph

I can only show one statistic, so which of these do you think would make the most sense to users? Would you use a different statistic altogether?

  • I'm finding it hard to square "the chance of rain today is 85%" with the individual per-hour probabilities or any of the other three "summary" figures. I would also question at least the phrasing, if not the usefulness, of those "summary" figures. Are they your summaries, or the web-site's? Averaging and summing probabilities is rarely the right thing to do. – TripeHound Nov 12 '18 at 14:17
  • Using the 24 per-hour P(rain) figures, I estimate the overall chance of rain "today" (in those 24 hours) as 17.5%... first calculate the chance of not raining in every hour (P(dry) = 1 - P(rain)), then multiply all 24 together to give the chance of not raining all day (~82%). Finally subtract from 1 (or 100%) to get the chance of rain at some point during the day being about 17.5%. – TripeHound Nov 12 '18 at 14:20
  • @TripeHound, so would you actually suggest just getting rid of that statistic? Now that you mention it, I realize that people don't really know how to interpret that. On top of that, the API also gives me a human readable summary that basically does what the chance of rain statistic was supposed to. Thanks for helping me realize this! – Daniel Nov 12 '18 at 19:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.