While SWAT should respond, shouldn't they cordon off the area and try to make contact? Actually going inside seems like it should be the last resort, after there is no response or there is observable probable cause or imminent threat.

Especially if the call is anonymous with few concrete details -- the police should consider why does this caller know the info they know? Why is the call anonymous or otherwise untraceable?

How many times have lives been saved by SWAT entering after a call like this vs an innocent person getting injured or killed? How many anonymous, untraceable calls to 911 have even turned out to be legitimate?

Not in the active shooter scenario. The standard approach now is to not even wait for SWAT, but to go in as soon as at least two officers are on the scene. The rationale is that active shooters usually stop killing people as soon as they get any kind of resistance, so timing is very important.

Of course "going in" does not mean shooting indiscriminately.

"Active shooter" scenarios usually only apply to areas with more than a few potential victims, e.g. not a single household.

It definitely seems like some simple heuristics could help a lot in giving police a sense of the probability that it's a false alarm. They will of course still need to assess the situation, but systemic police accountability issues aside, harming completely innocent people is still no picnic for the officers involved, so knowing when the fraud likelihood is particularly high should help them calm down and take a more measured approach.

Apart from call tracing, I'd expect simple things like whether the caller stays on the line and cooperates with the dispatcher (vs hanging up immediately) to be quite predictive.

  Actually going inside seems like it should be the last resort

and they did not go inside to shoot the victim.

> when it's really occurring, and isn't just some troll

I don't know how often hostage scenarios as extreme as you describe actually occur in the USA, but I read about instances of 'swatting' constantly. Given the false-positive rate of such reports, it's no surprise that sending military-style squads to 'shoot on sight' ends in tragedy again and again.

> I don't know how often hostage scenarios as extreme as you describe actually occur in the USA

Often. Mostly domestic violence barricading scenarios, which are usually handled by front-line officers with sidearms and negotiation. SWAT deployment is usually discretionary at such a scene, predicated on an active shooter, and anticipates forced entry. In an active shooter scenario, front-line officers will typically defer to SWAT to force entry, given the risk.

But yes, it does happen. A lot of domestic violence ends badly.

A lot of domestic violence ends badly because police actions ensured that it would. E.g. instead of leaving the bastard the woman develops a sort of codependent relation with local cops so that every busted lip she suffers means the bastard gets a tag-team beatdown from everyone on duty. Nightly entertainment for the whole neighborhood!

I'm no fan of wife-beaters, but somehow every other nation on earth has figured out less violent ways of dealing with them. Yes, in many cases "the way" is to simply accept that wives will get beaten. Admitting this is no more an acceptance of police violence than admitting that not all nations are governed with perfect justice would be an acceptance of all of our damned wars. This is a "won't somebody please think of the children?!" level of argument.

I think a big compounding problem is the overall prevalence of guns. In most places domestic abusers aren't untrained gun nuts living in a macho wave-your-gun-whenever-you-want culture.

People can obviously do plenty of damage beating each-other with objects, slashing with kitchen knives, or whatever, but guns seem to lead especially often to escalation from shoving to instant death.

When police show up, if someone has a gun everything is a lot more difficult/dangerous.

Agents, jet, Quantico? No.

Collect some evidence? Absolutely.