Ahh. I would rarely consider, let alone object, to things outside of context. X of course means very different things given varying Ys. The reason I am talking

Ahh. I would rarely consider, let alone object, to things outside of context. X of course means very different things given varying Ys. The reason I am talking about 10 million is because it's the magical number that the study the NYTimes chose to run an article on, decided to pick. It has nothing to do with reducing variance, though of course that's used as the justification. In reality even in populations of tens of thousands there is surprisingly low variance in homicide rates. I mean if you look at what they're doing it's really scummy. They're juking their numbers to try to create a correlation when one does not exist. That's already really bad. But then they take it a step further by just jumping to correlation = causation. Try to imagine that this was a topic that did not confirm your biases. I think you'd be out of your mind with how really trashy this whole article is. It's like they took a list of things you should not do when reporting on statistics, and decided to do them all.