Some companies are more worried about puzzles and algortithms. Most/all are available online. Some people mug them up and crack interviews. How can it be taken as a measure of capability for problem solving?
- WalmartLabs ferrariWhat is a better measure? A "four hour" take home programming assignment?
- Nothing worse than doing a phone screen and then getting a HW assignment. As much as I love technical stuff, I like to do it at work and focus on hobbies and family at home. So having to do one of these is just a fucking nightmare. Often I just end up getting half way through thinking that the whole process is bullshit and quitting. It's probably why I can't join any top tier companies but whatever.May 7, 20183
- I guess that's an option. Part of the problem is my ego is probably unreasonably large and I take offense to these challeneges. These also tend to weight hiring based on skills learned in the past rather than potential. I can't know what I've never been given the chance to do, but I tend to be pretty good when I finally do get a shot. That being said, I'm sure there are candidates just as good as myself who already have the background so...oh well.
- So skills are things like languages. It could be that I'm just too slow in the language they're testing me in at that point in time. Potential is more difficult.to quantify but comes down to your ability to do analysis and be creative. It's a lot harder to learn that than it is to be proficient in a skill. Thus, the fallback to testable things like leetcode, and hw problems.
You don't have to agree with me on this, it's jsut my opinion.
- At the same time, independently of the test, you’re always going to have the problem of the person not putting the same energy level in work as in the interview, whatever the contents of the interview are.
After the interview, the company can always fire the people that don’t do their job properly.
If you just onboard anyone without an interview, you’ll likely get an even worse result.
- IBM / Eng6969696969moreThis is the best way, period. The alternatives:
a) trusting whatever is on a candidate's resume and only asking behavioral questions to determine culture fit (people lie on their resumes)
b) take home assignments (waste of time)
c) asking non-leetcode programming questions, which would probably be trivia questions (no one likes these because they're easy to look up on the job, so do not assess a candidate's skills correctly)
As you can see, none of the alternatives are desirable. With FANG companies, they usually ask questions you most likely haven't seen before on LC, so algos questions _are_ good interview questions. And even if all questions are LC questions, why do you not like that? Practice LC for a couple weeks/months and you'll net yourself double maybe triple your TC. What's not to like? (And no, that's not "gaming the system", that's just practicing and strengthening your CS fundamentals).
- Amazon AkjsbAt worst it takes dedication to Leetcode your ass off to get in somewhere you may not have otherwise and learn how to see solutions to complex problems. This shows for persistence. At best they got in with their natural ability to see solutions to complex problems.... They may still be lazy.