Hacker News Reader

I feel like the designers of the test have to take more blame here. Having a test that was "just show us once and that's it" seems like a really bad test design.

Randomly testing vehicles after they've been produced should have always been part of the testing strategy. Shouldn't have been relying on a random researcher to catch them.

Would be like testing an Athlete once for steroids then never testing them again, ever.

View on HN

Would not have caught this. Every time the vehicle is randomly tested it would have entered cheat mode.

that isn't a very random test then.

"Random test" means taking a production vehicle of any age, and testing it in real world conditions. Basically putting measurement devices over the exhaust and driving it on the highway, not in a lab.

It's less scientific but also so much more scientific if you expect the object you are measuring to be cheating in lab conditions - and they should be expecting that.

Reply on news.ycombinator.com