What annoys me the most: They only include one year of data before. So, assuming they would have shown US and California data for, say, 5 years before legaliza

What annoys me the most: They only include one year of data before. So, assuming they would have shown US and California data for, say, 5 years before legalization, and during that time, there would have been a strong correlation between the two, in the numbers shown. And then a sudden split (decrease in correlation) in the two years following, it would have been much more convincing.