Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Why would you normalize by number of vehicles sold?

If a large-volume vehicle gets recalled, it should be a bigger news story than a niche car.



It sounds like they're talking about normalizing by number of models (SKUs) not number of units.

But in the case of Tesla, I would still want to know normalized since those niche cars are being trumpeted as mroe revolutionary and better than everyone else.


Actually their post suggested both normalizations (notice the word "or"), however both are flawed.

"Being trumpeted as [more] revolutionary" still has nothing to do with number of models and/or units. If you only give extra attention to certain brands because of their marketing, just say so. There's no need to connect it to unrelated factors, and there's no need to introduce 'normalizations' that don't normalize for anything.


Maybe you wouldn't - my thought was that (to exaggerate) if company A produces 10 vehicles and issues 5 recalls, whereas company B produces 1,000,000 vehicles and issues 10 recalls, then company B isn't doing twice as badly as company A despite issuing twice as many recalls, because they've manufactured 100,000 times as many vehicles, and we probably need to factor that in somehow. Likewise the number of vehicles affected by each recall.

As the other commenter said, I also think number of SKUs is important as each one represents a different design and BOM.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: