Oh about the stupid idea how war-related deaths have been decreasing for a handful of decades. I’m really thinking about publishing an alternative analysis every year: an event costing how many lives would have to occur this year to totally invert those silly linear regressions. I bet it wouldn’t be that big.

That is, I’d guess if 25k deaths happened in an official war in 2018, none of those reports would be able to say “war has declined since 1950s!” without serious data cherry picking 🍒.