I don't think anyone is claiming it does. But what's your point, really? It sounds like a conversation that goes like this:

"Wow, Lee was likely killed by someone who knew him, another tech entrepreneur."

"Yeah, and San Francisco sure has a lot of problems with homeless people."

Huh? Discussions about SF's homeless problem are pretty irrelevant to the case at hand. And it's really showing of people's biases that, when we first heard of Lee's death, the evidence-free, go-to narrative was that he obviously was killed by a random mentally-ill homeless person.

This is not bias. You don’t know in advance what the cause of an event is for sure. People mentioned a probable cause (violent homeless) and it happened to be a less probable one (a tech co-worker), although we still don’t know for sure yet. In particular, the influence of drugs and mental illness among the homeless were thought to explain the event.

It’s like predicting the price of a stock that didn’t come true. Not every incorrect prediction is due to selection bias.

Being killed by someone you know is far more probable than being randomly stabbed by a homeless person.

> This is not bias

Now a textbook case of bias is not bias!! The world is crazy?

"Well known problems" - really? how well known? Unless you can provide good sources you are simply showing selection bias.

That also doesn't the change the facts about the well known problems with the cold in SF during spring time and the very frequent fogs.