Comment by guhcampos

I can't really explain why, but to me they don't feel the same at all.

I've always finished exams absurdly quickly. I used to finish 60 min exams in like 20 min and sleep the rest of the time when the professor didn't allow us to leave, because I had likely spent the previous night drinking and partying my way through college. I'd usually ace most of them.

Thing is, at those times we were working with freshly acquired knowledge that you've been practicing a lot. That's not the case with most leetcode interviews. As a senior/staff engineer, I'm not using double pointers to perform little math tricks on a daily basis, I'm troubleshooting cascading timeouts on distributed systems. I'm not worried about how fast I'm going to iterate over a batch of a couple thousand records on a list I'm worried about how much latency I'll accrue if I rely on a primary source of truth instead of hitting a cached answer.

Code interviews don't measure experience or competence. I don't even think they measure stress as the article mentions. To me, they just measure how much leetcode you practiced for that interview. Nothing more.