Hacker News Reader

Comment by meowtimemania | Hacker News Reader

I think for large companies where they want to reduce the number of false positives, leetcode style questions work well.

View on HN

Was contacted by Facebook recently for a data engineering position and was told I had to grind some leetcode to prepare for the interview process. I just could not be bothered and I can not be the only one who made that decision. My guess is that companies relying on leetcode style interview are also missing on some very capable engineers who just don't want to grind leetcode outside working hours.

But that's not a problem. These companies pay so much that they have too many candidates applying all the time, so missing good engineers is not a problem, but filtering out false positives is.

Though, I don't agree leetcode is a good interview practice.

Yup! And they work well as a filter for candidates too. If a company leans heavily on these then you know it’s probably a dehumanising corporate hellscape.

> And they work well as a filter for candidates too

I agree. A couple of years ago I was asked by two companies to solve leet code problems even before there was a screener meet-and-greet interview. They were quickly crossed off my list.

There are many issues with this approach, starting with using the same tools to evaluate vastly different candidates, e.g. a college kid would likely outperform a senior guy in Leetcode questions, so that offers no proof of competency.

Has this ever been tested properly? Has anyone ever given a failed candidate the job and monitored how they did compared to candidates who passed the LC?

That's an opinion with no data backing it up. I guess you didn't read the article either before commenting?

Reply on news.ycombinator.com