I remember failing a Facebook SWE interview a couple years back after a few months of studying "Cracking the Code" and some other textbooks - I blanked out on a

I remember failing a Facebook SWE interview a couple years back after a few months of studying "Cracking the Code" and some other textbooks - I blanked out on a rather basic question in retrospect ("reverse a linked list in place). It never really occurred to me to cheat, at the time I just figured Facebook had a high precision and low acceptance method of hiring - I could probably do the job pretty well but I didn't show high enough "precision" in the interview so ¯\_(ツ)_/¯. I can tell an anecdote - when I took my qualifying exams in graduate school, every year there would be a different exam committee drawn from the faculty staff and their job was to come up with new questions. When I studied I looked at past exams going back 20 years and there was only a single question that was somewhat similar to a prior one - so cheating was quite impossible in the situation. People also underestimate the time and effort it takes to setup and design testing that is difficult to cheat. It's effort on the part of the testers and very hard to scale. I used to design and interview for DS positions at my old job and I only had a question bank of 2-3 long form problems, in a case study style that I would use on all the candidates. The interviews are exhausting and it's a lot work because I have to personally "escort" each candidate through the problem and assess them qualitatively.