> Meanwhile Teslas still drive straight into overturned trucks.

To be fair, so do people.

And people driving into things is considered to be a problem, not an insignificant quirk that's almost unworthy of mention.

In 2017, the last year we have data for, 10,230 accidents hit a stationary object, representing 30% of all fatal crashes.

Tesla has had a total of 5 fatal crashes in its history, and only one of them was hitting a stationary object (20%).

So objectively Tesla is doing better than humans.

You also need to figure out how many of those crashes are intentional suicides or a result of something else (such as someone having a heart attack and losing control of their car).

Basically, you would need to figure out how many of those accidents hits stationary objects on the highway that are the drivers fault. That's what Tesla cars have done. I don't think those stats are available.

How many miles driven (which is a probably more correct measure of exposure) are represented in the two cases?