Hacker News new | past | comments | ask | show | jobs | submit login

It's not "fully subjective" by any sensible definition. Yes humans are the measurement device, and humans differ and are unreliable, but given a sufficient number of samples you arrive at more accurate numbers.

Nobody's going to give a paprika the same rating as a Carolina Reaper. You might have some give two very strong peppers the same rating, but then you might just need more people to test it.

In any case, the practical utility of the scale is mostly to get an idea of how spicy something is, it doesn't really matter much if something is 700K or 600K, it's enough to know that it's around 2x as strong as a Habanero.




Isn't there going to be a huge sampling bias? I certainly refuse to eat Carolina Reaper, and can't be the only one.


No, the Scoville test is based on a comparison of diluted samples. Trials are rerun with increasing dilution until the testers cannot detect the sample, and the level of dilution determines the score.




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

Search: