If they were less safe than human drivers, one would there would be more accidents where Waymo was at fault than another party.

Having personally gone through several years of AV collisions reports to California a few months back, that wasn't the case. Indeed, in most of the small number of cases where Waymo was at fault, there was human driver in control of the vehicle.

The were a four cases where the Waymo was at fault. One hit a chain strung across parking lot. There was a minor collision with a parked car. One ran over and killed a dog. One hit a vehicle while trying to reverse.

There were dozens where the other party was at fault (lots of rear ends).

Now that you put it that way, it's probably tough to get in an accident with a waymo without being found at fault. How often does a car stop suddenly and unexpectedly, get rear ended, and get found at fault?

Never. If you rear end someone because you were following too closely to stop without hitting them, you're at fault. There's a reason why the three second rule is taught.

There are some rare exceptions like if someone's tail lights are out, but nothing that would apply here.

Waymo worked with Swiss Re a reinsurer to evaluate the data. I think they have a paper that is going through peer review.

https://waymo.com/blog/2023/09/waymos-autonomous-vehicles-ar...

They have been submitting data to the DMV for some time. Do you believe they are falsifying their data?