I voted yes because look at googles engineering level
In theory if someone is dedicated enough to spend a lot of time learning and grinding leetcode, they should be capable of dedicating the same amount of effort to learning the skills necessary for excelling on the job? But in reality it seems like most people stay long enough to write something meaningful on their resume for that job before grinding leetcode again and hopping somewhere else.
Yup. This is exactly the problem. Job hopping to optimize tc since they are good at LC. Cant blame them and understandable since people are greedy. Nothing on capability since they are capable of grinding LC.
So admit it. Most software engineers are easily replaceable. No matter where they work
It acts as a THRESHOLD not a correlation. I.E Leetcode questions cannot differentiate the top 2% from the top 15% of candidates. But it is very good at weeding out the bad ones, I.e the difference between the top 40% and 15%. This is why Google doesn't find any impact on job performance. Because it didn't track anybody who didn't make the bar, only those who did and were hired.
Very well said. Exactly what I’m thinking.
In the past Google has hired some people who didn’t meet their bar for this reason, and they found that those people had the same performance distribution as the rest of Googlers who did meet the bar as @The Plague said above.
Leetcode based selection is why we have so many engineers with bad attitudes and suffering from brilliant jerk syndromes.
I’m confused. You don’t just LC. There’s system design and previous experience interviews. Why is everyone on blind saying it’s all people memorizing leets? How is that even realistic? You get better at leets by literally learning algorithms and get good at SD interviews by actually learning system design. Also, you are very limited to the amount of stuff you can assess on an interview.
Agreed. I would do far better away system design I think than I would at leetcode. Only 2 yoe employed as a software engineer, 4.5 if you count all the stuff I did idependently before getting a job, but I've yet to implement a BFS or DFS of anything. It's far more valuable to understand basic software engineering principles to write and maintain actual production code. I think that theres only been one time I've had to revisit algorithm complexity so far (And I'm in research). The rest of the optimization I've had to do is based on query execution and managing memory. 20-40% of my job in any given week is soft skills (focused on requirements). 20% mentoring/training. Only 40-60% writing code. Of that time most of it is "how do I build these classes best?" Not "how can I improve this implementation to get better than O(n log n) that I get from the sort? Can I rewrite it to avoid sorting in O(n) time?"
You won't pass if you don't do perfectly at the multiple coding assesments. These other hurdles are more to work out if you have the brain capacity for semi-audible communication.
Why do companies ask absurdly hard leetcode problems that dont seem reasonable within 45 min context?
No in fact one can argue the opposite. We have people who are in it purely for money (and they are only good at interview) instead of craftsmanship and deep passion for engineering
This is what i see at linkedin. Many engineers dont even understand how http work, design principle etc.
That makes sense. I knew a guy like that. It's kind of annoying because he had to be handheld so much.