Earlier quoted context omitted.
It's still missing a crucial ingredient: you have to demonstrate that your list of events is an unbiased sampling of all relevant events, as opposed to a cherry-picked list of useful events. For example, it would not be an "empirical argument" to convince everybody that a coin was unfair by publishing an article whenever it came up heads. An example isn't evidence, and a list of examples aren't evidence. Evidence is…
> It's still missing a crucial ingredient: you have to demonstrate that your list of events is an unbiased sampling of all relevant events, as opposed to a cherry-picked list of useful events. That's if you're writing a paper. If you're making an empirical observation in the real world, the demands are much less rigorous, and if one can't tell what's going on without first demonstrating "that your list of events is a…
The demands are only less rigorous when you agree with the conclusion. Link an article from a racist website dedicated to black-on-white crime (that ignores other crime), and watch as the demands suddenly become much more rigorous.