The chart shows that it's more or less the same across middle skin tones, with the darkest being almost non-existent, and the lightest being quite lower than th

The chart shows that it's more or less the same across middle skin tones, with the darkest being almost non-existent, and the lightest being quite lower than the median. However, in the text, it says that US twitter accounts are a 4:1 ratio of white vs black, but black people are more active on twitter overall, which skews the results somewhat. It all feels inconclusive to me, and also rather pointless (see my earlier comment).