The color scheme itself is kludged using arbitrary thresholds to highlight where things are happening. A linear scale doesn’t work because NYC is such an outlier currently (it washes everything else to white). I tried variations of log scales which didn’t quite work.
I also tried quantile and sequential quantile scales, and while those did allow for differentiated display across all the data, I don’t like how it put counties with ~40 cases in the same bucket as counties with >30k cases.
If anyone has any ideas about that, I’d love to hear from you.
The interactive Observable notebook can be found here: