Weeds out people who didn't memorize an answer to a well known test. But it's more efficient to discuss a real world code problem as this allows you to weed out people who can't solve problems or code either.

For Christ sake. There was never any need to memorize fizz buzz, that's the whole point of the damn question. It tests an applicants basic math and logic capabilities (in this case modulo and for loops).

> For Christ sake. There was never any need to memorize fizz buzz,

Wrong. Fizzbuzz is the epitome of a hiring test that is gamed by rote learning and adds absolutely nothing in terms of decisive info on wether s candidate is competent or not.

Your argument is like arguing in favour of using hello word in as many languages as possible and then claim that there is no need to memorize them because that's the point of the question.

I have big issues with impractical interview questions that favor interview prep over actually practical skills. That said, anyone who can program should be able to implement fizz buzz in 10-15 minutes given a description of the problem even if they had never heard of it.

hmm interesting... i haven't done a technical interview in 15 years so i just tried fizzbuzz (leetcode was the first thing that popped up on google)... took me about 15 minutes. i was caught a bit off guard, there are a couple of little subtle gotchas that require some actual thought. i'd say it's a pretty good test without requiring knowledge of data structures or CS level algorithms.

that said i would only want to do it on a terminal, not on a whiteboard. i'd probably fuck that up somehow.

FizzBuzz exactly as printed in an example? An easily game-able memorization problem.

The concept of FizzBuzz? If you are a programmer you should be able to write a simple program with arithmetic and string manipulation. There's no reason an interviewer needs to ask exactly the FizzBuzz - but just something a bit like it. Asking the exact question from the blog is pretty lazy anyway.

if you need to memorize fizbuzz answer - I have a very very bad news for you

If you need to use it to test candidates, bad news is you can't tell rote memorizers from problem solvers.

Then you say "but let's do buzzfizz instead" and let the memorizers start crying?

Or just pose a real world problem? Like every technical interview I've ever had. Giving me a problem from a catalogue is letting me know you're boring - I'll be polite but I won't accept your offer.

you wont accept, but there will be a line of thousand+ fresh college graduates and H1Bs behind you, all willing to grind leetcode and accept 200k+ offer

"Weeds out people who didn't memorize an answer to a well known test."

For the love of god, please Google what fizzbuzz is.

People who need to memorize a solution to that have no business coding.

> For the love of god, please Google what fizzbuzz is.

Yes, please do google what fizzbuzz is. You'll eventually stumble upon resources on how to memorize fizzbuzz, which is such a Hallmark of the test that it's even explicitly mentioned in the C2 wiki on fizzbuzz.

The test can't tell if you can think through a problem or just googled technical interview common questions and memorized some answers: Crammers can pass as easily as problem solvers can.

Exactly, I started doing it since I know if they can do literally 5 lines of super basic code. It should take 30 seconds. Why would I ask someone to write kernel code or other domain topics if they can't write fizzbuzz??? The replies to your comment are exactly what I meant lol.