All along I’ve been planning to report repair rates per vehicle, rather than per 100. My reasoning is that doing the usual thing and reporting rates “per 100 vehicles” yields larger numbers that can give the impressions that repair rates are higher than they are. Sure, if people think through it they’ll realize that 62 repairs per 100 cars is the same as 0.62 repairs per car, and vice-versa. But the human mind doesn’t always operate so rationally–62 sounds much larger than 0.62.

But I’ve been receiving feedback that people have trouble thinking about tenths of a repair trip. So I’ve decided to give “repair rates per 100 vehicles” a shot after all.

The previous set of results remains in the old format, for those who don’t remember it or never saw it.

Is this better or worse? If you have a strong opinion either way, please let me know.