> Maybe police should default to assume swatting and develop techniques to evaluate actual risk prior to storming in based on an anonymous phone call?
You read about a few false positives and decide the police need to start assuming by default that every call is a false positive. You have no data on the number of calls where sending a SWAT team is the correct response (true positives), the number of calls where the responder correctly decides not to send SWAT (true negatives) or where the responder incorrectly decides not to send SWAT (false negatives).
No, you have no data, just a handful of anecdotes and on that basis you want to completely change how first responders work. You talk about “Bayesian” but you couldn’t possibly calculate how many lives could be saved or lost by this change.
Please, let’s stop acting like we’re experts at everything because we know a bit about software. Maybe other people are competent too. Maybe they know shit we don’t.
Oh and by the way, like someone else pointed out, the article doesn’t even say SWAT was deployed or that any law enforcement entered his home.
Your comment confused me, the article says the kid in the U.K. called the local police department so until I read this I assumed they googled the local number and dialled the police directly internationally, not going through 911. But it sounds like that would have got them 911? Here it would just get you the guy at the front desk in that station.
I am surprised if it’s possible for someone in the U.K. to call 911 and get “911 stateside” and not redirected to 999 in the U.K.?
I don't assume competence of 911 responders because they regularly get the wrong house. Your aggressive screed really added nothing to the conversation. "It's probably correct so we should ignore the problem" isn't a sane perspective on police violence based on a prank call.
My point is that they have more data and experience on deciding whether a call is legitimate or not. More than the GP here, who read two articles about SWAT-ing before deciding to upend the entire system.
> regularly get the wrong house
Citation needed, with data? Or do you, like GP, also want us to assume that your anecdotes are data? And getting the exact address isn’t the same thing as deciding the legitimacy of a call.
> “it’s probably correct so we should ignore the problem” isn’t a sane perspective
You’ve mischaracterised my comment and called me insane. That was uncalled for. I did not say this wasn’t a problem. I merely questioned whether a person with 0 data and 0 experience was qualified to call for upending the current system. More data is needed before we can hold such strong opinions.
I would request that in future you don’t call people’s opinions insane without even understanding what they’re saying. You’re currently in violation of HN guidelines (https://news.ycombinator.com/newsguidelines.html)
Depends if you count a minimum of 10% "regularly"?
From 2003: "After the New York City raid that killed Alberta Spruill, Police Chief Raymond Kelly estimated that at least 10 percent of the city’s 450+ monthly no-knock drug raids were served on the wrong address, under bad information, or otherwise didn’t produce enough evidence for an arrest. Kelly conceded, however, that NYPD didn’t keep careful track of botched raids, leading one city council member to speculate the problem could be even worse."
We’re not talking about the same thing are we? This conversation is specifically about where dispatchers are sent after a person makes a call requesting law enforcement, giving their address.
Although I think it speaks to the general competence of the US police that they can't get the right address even when they've spent days/weeks organising a raid - I wouldn't necessarily trust them to get it right on a fast rollout on a SWAT call.
You read about a few false positives and decide the police need to start assuming by default that every call is a false positive. You have no data on the number of calls where sending a SWAT team is the correct response (true positives), the number of calls where the responder correctly decides not to send SWAT (true negatives) or where the responder incorrectly decides not to send SWAT (false negatives).
No, you have no data, just a handful of anecdotes and on that basis you want to completely change how first responders work. You talk about “Bayesian” but you couldn’t possibly calculate how many lives could be saved or lost by this change.
Please, let’s stop acting like we’re experts at everything because we know a bit about software. Maybe other people are competent too. Maybe they know shit we don’t.
Oh and by the way, like someone else pointed out, the article doesn’t even say SWAT was deployed or that any law enforcement entered his home.