Freeway accidents, due to their nature, are a lot harder to ignore and underreport than accidentally bumping or scraping into another car at low speeds. It's like using murder rates to estimate real crime rates because murders, unlike most other crimes, are far more likely to be properly documented.

> Driving on freeways would significantly lower Waymo's accident per mile rate.

Maybe. We don’t know for sure.

You seem to frame that a bit like Waymo is cheating or padding their numbers.

But I see that as them taking appropriate care and avoiding stupid risks.

Anyway as someone else pointed out they recently started doing freeways in Austin so we’ll know soon.

> You seem to frame that a bit like Waymo is cheating or padding their numbers.

Not sure how you read that. I'm saying Waymo was prioritizing safety.

Oh, sorry. I thought you were arguing from the other side, saying that Waymo only looked good because they were avoiding anything difficult.

Same argument, different sentiment.

I said they were avoiding the easy thing.