8 tweets · April 2020 · 11 likes · 0 retweets · read on Twitter
Idea 💡 Show logarithmic growth charts with the total population indicated as well. eg for USA - 10^8.5 people total - 10^5.5 confirmed cases now - 10^2.5 confirmed cases a month ago at that rate, would max out May3 at current rates, would max out May28 ofc, won't stay exp
but this shows *visually* that in the last month USA has gone halfway from 10^2.5 (~300 cases) to 10^8.5 (full population) sort of obvious with 300 300,000 300,000,000 but graphs still help fortunately, things have slowed somewhat, so actual current slope is about half that
If we include not just the max but also the min, we can see that it's *more* than half the way from 1 case to 327M (specifically, 5.5/8.5 = 65% ~= ⅔ of the way there)
I made this spreadsheet publically viewable if you want to play with it. I'm not an expert at any of this, but playing around has been helpful for me! (Make a Copy, then edit, or if we're friends/mutuals then request edit access & make your own sheet) docs.google.com/spreadsheets/d…
Oho! I figured out how to do logistic fns! I set them to assume it would cap at 70% of the population. This is revealing: the logistic is at most 3 days behind the exponential until half of the total (in this case 35%, ie 100k americans) has been reached, but then rapidly slows
It highlights how different these are: - "70% of population gets it at maximum growth rate" is (practically still exponential until huge) - successful containment: growth slows/halts *earlier* Say growth slowed from 30%/day to zero over a few weeks—total cases would stop at 500k
Hm. Above chart doesn't actually do it justice, because we fail to appreciate how far apart the purple & green dotted lines are. green is 200× purple. But if you try to do it on a linear scale... it looks weird the other way: current # of cases not visible at all on 50M scale!
Okay, this has been a satisfying experience of *learning* with the garage door up. I don't think it's exactly "work". It's been very playful! I'm also enjoying this experience of the interplay between tweeting & thinking/exploring. Mutually supportive.
