First of all, I am a fan of using leetcode style questions on interviews in large companies with a lot of positions to fill. It's provides a cohesive standard for interviewing that scales well. I believe it should reduce the amount of worthless hires. However, my post is about the notion that these kind of questions provide you insight into the candidates problem solving. Who came up with that bullshit and why do people keep parroting it? Let's say I ask you a leetcode medium/hard you've never seen anything like it before. There is no way that in 45 minutes you'll get it correct. You'll be showing me how bad your thought process is at trying to come up with something that took many people many hours to optimize. Then I will fail you for providing at best n squared algorithm. Now let's say you have seen something like it before. And you spent hours and days optimizing your thought process to get that answer. So all i see is your ability to memorize and recall something. If you were to face a new problem on the job, your thought process would be nothing like that. Plus I will pass you, even though it may have taken you a year to memorize that one question. So please don't say that it provides insights into ones thought process. It's simply a test to weed out people that can't comprehend something enough to memorize and recall it. The thought process you show will be the one you copied from other people.
Your problem solving skills are so bad that you don’t even know what problem solving is. It’s like someone who is asked 32 + 53 and being unable to do it because they’ve never had to add exactly those 2 numbers before. Then if someone figures out the answer you accuse them of having memorized it.
This
Except "multiplication" is usually a very basic CS concept covered in the first 2 years of undergrad. It's expected that you know how to do it. I think the most niche thing I've seen is like... a trie? Which isn't even niche. Or maybe some intricate DP solutions.
Google gave me problems I had not seen before, I solved them 🤟
Same here. 3 of them were dp.
Two issues: 1) You assumed people who came up with an optimal solution memorized it may be because that’s how you got in. 2) Passing someone who memorized vs who actually came up with decent solution is your problem not everyone else’s
What OP fails to realize is that in order to memorize a solution to a problem....wait for it...someome has to fucking solve it first. Learn to do that part OP.
@lchardisez I agree you can learn to solve it for the interview. But that is not your own unique thought process. You memorized how someone else solved it. And I bet you forget it in a few months. I'm not against this method. I just don't want to put the label on it that it gives me insight into your thinking process. That all of a sudden I can tell that you will be adept at solving all sorts of hard cs problems you've never seen before without looking them up or asking for help.
Coming up with a perfect optimal solution is not a requirement for passing an interview. Unless the question is simple it's often not expected you will which is why good interviewers are not only interested in a perfect solution. Time is the enemy in an interview and it's true you can only do so much in 45min. A lot of people have this notion of perfect solution or interview failed. Focus on effectively communicating a working solution and go from there on speeding it up. Interviewers are often more interested in the communication explation process. If you get a question you totally bomb. Understand why you failed it and don't repeat the same mistake. There are an infinite number of questions but a finite number of topics. As to if leetcode / grinding out these problems are an effective measure is a different issue. I believe there are plenty of smart / great engineers who would excel in most companies that may not be as good at test taking type situations but because companies are terrified of false negatives they make the test hard to prevent that
Im fine with saying the process is to gauge your communication skills. But I think people focus on "thought process". Your last point is really crucial. I know a lot of engineers with great thought processes who would fail these interviews without leetcode. Their thought process is amazing. But that's hard to see in a 45 min interview.
Mix in some divergent thinking tasks /problems. More difficult to evaluate but also more informative/revealing.
Did a round of interviews with 5 companies recently. Among 15 (most of them lc hard) problems I saw only 2 or 3 beforehand. Got offers from 4 of them. So it IS possible to solve problems if you didn’t see them before. Maybe you should practice more;)
So if I practice (look at) more lc problems, the better I will get at solving problems I've never seen before?
India
Yesterday
1357
Slavery has REVERSED! the US is the slave!!! Check out this dude who pays a personal trainer in India
Tech Industry
9h
544
Chloe Shih: "I’m the highest performer on every team, hands down"
Work Visa
Yesterday
5560
Last h1 attempt
Tech Industry
Yesterday
636
Best LCOL or MCOL city?
Tech Industry
Yesterday
777
Do you really think Amazon is that bad
You must be dumb. Dumb people say problem solving questions can only be memorized. Smart people figure it out from first principles.
So where are your papers on your brand new answers to new problems?
I’m serious. As an interviewer I ask a problem solving question that hasn’t leaked and I can easily tell apart good cs thinkers from bad