If you don’t think that dog breed is a good predictor of behavior, you have not spent enough time around dogs.

For thousands of years dogs have been bred for specific purposes. These behaviors are innate. They do not need to be taught. Sure, you can train them to be better, but the behaviors are written all over their genes

My grandparents had shepherds. The dogs had never seen sheep or been taught anything about herding, but they would attempt to herd all my cousins when they were children, then get agitated when the children wouldn’t herd. Here’s some puppies doing it

Here’s some pointers pointing. They have not been taught this (and frankly I can’t imagine even training most dog breeds to do that)

Here’s a boxer dog boxing. Here’s one spinning. They aren’t taught this, and they all do it.

There’s hounds rolling in stink. There’s sight hounds and smell hounds. There’s retrievers retrieving, being irresistibly drawn to water, and carrying around things very gently. There’s huskies being extremely energetic and vocal.

I could go on.

Do you really think that dogs that have been bred to fight other dogs to the death and bear enormous amounts of pain (game) before giving up are not dangerous? You’re mental.

Sure they’re sweet to their owners. That’s because people who breed animals for blood sports are not the kind of people who would have trouble immediately removing from the gene pool any of their animals that are disloyal.

It’s not like it’s just pitbulls. Dobermans are implicated too. They’re guard dogs but for humans rather than predator animals.

People with agendas can play all kinds of statistical games to show what they want to show. In the scientific world, these kinds of tricks stand out. That’s why any non-trivial summary statistic is useless without a large text explaining the methodology.

This is one of those things that is so obvious it boggles my mind that people even question it.

Of course dogs that are bred to murder are dangerous.