[rifung]

If you consider how well someone does as how likely they are to be hired, then it makes sense that interviewers are the source of "truth", because it's their grade or opinion which will decide whether you get hired or not.

We certainly can speculate whether whose opinions are more correct or valid, but if we just objectively consider that doing perfectly means you get hired, then what the interviewer thinks matters and what the interviewee thinks doesn't.

I do think you raise an interesting question though. I do wonder whether, in a scenario where many interviewers saw the same interview performance, how varied their scores would be.

After all, most people fail at many job interviews before landing one they get, but is that because of variation in the performance of the interviewee or because of the differences in interviewers?

[p4wnc6]

Using the measurement, "how likely are they to be hired" seems like it's precisely the problem. "How likely someone is to be hired" is a totally erratic and unpredictable property of a bureaucratic hiring process. It's exactly not the sort of thing you want to use to define an absolute notion of skill or performance in a job interview.