Hacker News Reader

Comment by bredren | Hacker News Reader

I have a very talented SWE friend who only left a now-public unicorn because they were recruited by a former colleague at a startup they knew wouldn't involve leetcode in the interview process.

I have another very talented SWE friend who has been doing leetcode exercises off and on for months and is struggling to start doing applications even though they do not like their current work and are undercompensated.

They fear they will fail algo / data structures and yet this person has an excellent public track record of quality contributions and comms in FOSS.

From what I've observed leetcode screens are problematic enough that they are causing market inefficiency.

View on HN

leetcode difficulty is vastly exaggerated. people study 4 whole years of their lives in college to increase their chances of making money, why not spend around 5 months on leetcode which can perhaps get you much farther than anything from college?

i'm not particularly smart, in fact i struggle at work routinely, but i know what i have to do clear interviews, so i do the grind and it has worked well for me. (to a certain definition of "well")

now that i think of it, it has only worked well for me because actual talented swe's don't put effort into interviews the way people like me do.

People who are good at the grind are good at the grind. These are also often the people who succeed in life; intelligence is vastly overrated.

I suspect some people want to make time but have family or other obligations.

Adding new project that must be executed on as a prerequisite to being seriously considered for a job when you are obviously capable can be a hard sell.

Especially given programming challenges are not that fun for people used to delivering features.

I know that for talented SWEs, having a leetcode interview sink an interview process it can be invalidating.

Companies that rely on single algo/ds interviews also discourage potentially great employees when they totally ignore evidence of capability.

The company wants candidates to grind ds/algos, but it won't compel its own employees to look at public contributions.

> From what I've observed leetcode screens are problematic enough that they are causing market inefficiency.

Totally agreed, I know of many similar situations like this.. however at the same time the question is whether there’s any other system that’s better in practice.

“Capitalism is the worst system, except for all the others” etc.

You want a controlled trial or natural experiment where two large successful software companies experiment with two different approaches to hiring over a number of years holding all other factors equal. That’s hard but may happen at some point in our lifetimes.

Reply on news.ycombinator.com