I am sick of LeetCode-style interviews

I understand the frustration, but at the end of the day I think they do serve a purpose. They&#x27;re not great at that purpose, but they&#x27;re good enough and they generally produce false negatives (smart people fail) rather than false positives (you hire an idiot). And this makes sense, because a false negative is low cost (maybe you spend 50% longer interviewing candidates) but a false positive is high cost (you hire an idiot and have to spend months establishing that, firing them and starting the hiring process again).<p>The side effect of this though, is that it&#x27;s an extremely painful process for the applicant. Even if you&#x27;re a great software engineer, you&#x27;re going to freeze up, miss something, have a bad day, and fail a few interviews. Now for most software engineers failure is an unusual and harrowing experience, so they react really badly to it. <i>Especially</i> ina scenario where it&#x27;s difficult to blame anyone but yourself. But just don&#x27;t! It&#x27;s fine! just move on, there&#x27;s plenty of jobs and the interview process is a blunt tool, not a final evaluation of your worth in life.

I once had an interview that I really liked. It was for a bank.<p>They give the candidate some messy, done-in-a-hurry code (but not purposefully obfuscated) and ask them to refactor it to the best of their ability. The interviewer sits next to the candidate and talks to them throughout the whole process. It&#x27;s pretty much pair programming, but the candidate has the initiative.<p>I found it to be a breath of fresh air after all the leetcode interviews. Of course, they have the luxury of doing this, because they know what exact technologies they are using and they don&#x27;t need to measure candidate&#x27;s ability to learn new tech, etc. In their situation they are more interested in topics like whether the candidate can produce maintainable code, name things properly or communicate with co-workers.

Leetcode style interviews probably serve two functions:<p>1) A way to suppress wages and job mobility for SWE. Who wants to switch jobs when it means studying for a month or two? Also, if you get unlucky and some try hard drops an atomic LC hard bomb on you now you have an entire company you can no longer apply to for a year.<p>2) A way to mask bias in the process while claiming that it’s a fair process because everyone has a clear&#x2F;similar objective.<p>Meet someone who went to your Alma mater? Same gender? Same race? Give them the same question as everyone else, but hint them through it, ignore some syntax errors, and give them a strong hire for “communication” when they didn’t even implement the optimal approach…<p>Or is it someone you don’t like for X reason? Drop a leetcode hard on them and send them packing and just remain silent the entire interview.<p>To the company this is acceptable noise, but to the individual, this is costing us 100s of thousands of dollars, because there’s only a handful of companies that pay well and they all have the same interview process. Failing 3 interviews probably means you’re now out $200-300k of additional compensation from the top paying companies.<p>I’ve interviewed for and at FAANGs. I can’t believe the low bar of people that we’ve hired, while simultaneously seeing insane ridiculous quad tree&#x2F;number theory type questions that have caused other great engineers to miss out on good opportunities.<p>Someone will reply to me “if you know how to problem solve you will always pass.” Ok, come interview with me and I will ask you verbatim one of those quad tree&#x2F;number theory&#x2F;inclusion exclusion principle questions and I’d love to see you squirm, meanwhile another candidate is asked a basic hash map question.

The important thing to screen a software engineer for is not knowledge but the strength of their problem solving and ability to build mental models of problems.<p>Whiteboard interviews are a good way to evaluate these skills - assuming the candidate hasn&#x27;t seen this particular problem before and the interviewer understands that this is what they&#x27;re supposed to look for.<p>Goodhart&#x27;s law applies - the measure becomes a target and it ceases to be a good measure.<p>People who have weak problem solving skills or poor abilities to form mental models still want high paying software jobs so they &quot;grind leetcode&quot; until they can pass these interviews.<p>Then over time these grinders get high enough up they start becoming interviewers themselves and they don&#x27;t understand that they&#x27;re even supposed to screen for talent, they conduct the interview like a memory test since to them that&#x27;s what it was.<p>In spite of this, there&#x27;s no better interview type I&#x27;ve seen proposed given the constraints most companies are under (don&#x27;t waste candidate&#x27;s time, have a consistent rubric that HR likes, etc.) so for now we&#x27;re going to keep using it.

A friend once in the middle of a technical interview stopped speaking, stood up and went out. They ran after him, &quot;what&#x27;s wrong, you&#x27;re doing fine!&quot; and he told them it depresses him too much and he&#x27;s not interested anymore. Sad, given that this is a fun profession for most of us (I assume)

5 years later, I still refuse to do them and I don&#x27;t want to put myself through that process ever again. I got bullied, harassed, laughed at and taunted by people who use this as a one-hour game of kicking people around and I&#x27;m not interested in entertaining them any more. I have the following disclaimer both on my LinkedIn [1] and GitHub [2] profiles because I had plenty of surprise-leetcode interviews, so I want to make it clear up front what&#x27;s acceptable and what isn&#x27;t.<p>&gt; I have a personal policy against any type of live coding or online coding tests during interviews and I don&#x27;t enjoy or engage in any form of competitive programming. Otherwise, I am happy to work offline on coding assignments with reasonable goals and deadlines and to have in-depth technical discussions about software architecture and design as well as relevant technologies.<p>[1] <a href="https:&#x2F;&#x2F;www.linkedin.com&#x2F;in&#x2F;mtodor" rel="nofollow">https:&#x2F;&#x2F;www.linkedin.com&#x2F;in&#x2F;mtodor</a><p>[2] <a href="https:&#x2F;&#x2F;github.com&#x2F;mihaitodor">https:&#x2F;&#x2F;github.com&#x2F;mihaitodor</a>

I see these posts a lot, and I sort of disagree with where they are coming from so I&#x27;ll play devils advocate. Leetcode is used to select for one of (or both of these) 2 things for the average IC hire:<p>1. Do you work hard (i.e a person who studies many problems, memorizes solutions, and regurgitates then)<p>2. Are you smart (i.e a person who doesn&#x27;t do any leetcode but knows the fundamentals and can sythesize past knowledge well enough to answer a new question)<p>Now in a vaccuum, my takeaway from someone who doesn&#x27;t pass a leetcode question is that they are more likely to not be either of those (given no other information about them). In my opinion, solving at least one leetcode easy - medium question (maybe with some direction from the interviewer) should be the standard for any role where you need to code.<p>And realistically, companies index less on leetcode skill as you become more senior or apply to more niche roles. And plenty of companies out there have bug squash rounds, system design, behaviorals, take homes, trivia, background specific questions, etc. Leetcode is usually not the be all end all for hires above entry level. Leetcode is not meant to magically select for the best engineers. Its simply another signal for the hiring committee to use.

Another interesting point about Leetcode-style interviews I&#x27;ve not seen raised: it&#x27;s a form of intellectual hazing.<p>It&#x27;s been proven many times that the severity of an initiation ceremony significantly boosts the commitment of those admitted to the group.

The last few interviews I&#x27;ve run, I&#x27;ve shown the candidate sample code and asked them to review it. We have samples that have at least one bug &#x2F; problem per line. Then some modeling task (whiteboard). We have some simple exercises (fizzbuzz style) to figure out if juniors actually can junior and as a springboard into discussing topics like runtime complexity, memory layout and stuff. This is all highly interactive, depending on how well the candidate does. The idea is to figure out if the candidate somehow faked themselves through screening as well as to somehow get a feeling for cultural fit...

I&#x27;ve interviewed developer candidates who seemed to know a lot of theory, but when it came to write some code, they failed miserably. They could not implement fizz-buzz. Senior Java developers who could not import HashMap without looking it up.<p>It&#x27;s kind of like math. If you need to multiply 562 * 1041, you should use a calculator. If you need to reach for a calculator to tell me how much 3 * 7 is, I will doubt that you are an expert.<p>Many companies overdo the leetcode interviews. A live coding interview that shows that a person can do array manipulation, implement flow control, and use basic data structures shows me that you have the muscle memory that only comes with experience. If you need to study graph-traversal algorithms in order to pass, it&#x27;s probably too leet.

Such interviews are made to legally discriminate against old developers.<p>A freshly graduate student has more chance to still know this or that algorithm, and prepare for that.<p>An old&#x2F;experienced developer has dealt with plenty of real world complex problems. So much that his attention is absorbed by them, not by what is the best way to search for an element in a tree.

I&#x27;ve interviewed lots of developers and recommended &#x27;hire&#x27; to about 15 people in the last several years and not once did we get it wrong.<p>All it takes is a short problem to solve at home, which can be easily googled and one hour architecture interview, where we discuss the technical architecture of a hypothetical &#x27;real world&#x27; service.<p>It takes about 20 minutes to determine if the candidate has the experience with the technologies listed in the resume. Little experience is not necessarily a show stopper.<p>True that the language we hire for (Clojure) filters out a lot of people ahead of time, but knowing the language is not exactly what I look for..<p>What I look for is &#x27;passion&#x27; - does the candidate love programming and can the candidate articulate technical issues with ease.<p>The other question I try to answer - will the candidate enjoy working in our team and will we enjoy working with him&#x2F;her.<p>Smart people will shine in a certain way, even if they bomb specific questions - they have an opinion, they try, they ask the right questions.<p>I&#x27;d be sad if we missed on some of the people in our team because of automated (inhuman) puzzle interviews.<p>We&#x27;re not looking for a problem solving machine, we&#x27;re looking for a partner to create something great together, someone who shares our passion for hacking and someone who we&#x27;d love to work with.<p>I share TFA&#x27;s opinion that leetcode-style interviews are not the way to go and I hope the industry comes back around and focuses on the human side more.

I&#x27;m happy that Amazon have a horrible leetcode style interview practice. In my interview I had to design an algorithm to detect certain combinations of fruit and veg bought from Amazon fresh, which would trigger fruit-machine style payouts. If I hadn&#x27;t said &quot;f this&quot; and did something that actually interests me, I would probably have a huge mortgage sitting in rainy ol Blighty designing user hostile experiences like Prime cancellation. I&#x27;m poorer, happier, healthier and mentally iron-clad doing my own thing, making just enough money and travelling to where life has a reasonable cost.

Isn&#x27;t the entire point of memoizing this to demonstrate you are able to study&#x2F;grind rather than prove intelligence? Different companies&#x2F;managers have give me different reasons for the usage of this style of interview but it was usually to &#x27;assess ability to get through rough stuff&#x27; and &#x27;assess cognitive abilities&#x27;.<p>Idk anymore either and I refuse to conduct these interviews. I generally prefer somewhat realistic problems that are multi-disciplenary but have ties to real world applications related to the role.

The market is so bad, and most interview sucks one way or another. The main reason is interviewer have tens or hundreds application for same position.<p>I have just went through one interview that have 4 rounds.<p>1. half hour talk 2. 2 hours coding for several crud api. 3. system design project ( I spend 2 days on it. they told I have about a week on it) 4. another half hours with CTO.<p>and then got rejected. (The interviewers does not bother to send a reject letter. I asked them they told me the position is closed).

Since 99% of people have to study hard to pass LC (medium&#x2F;hard), it effectively acts as a selector for employee conformity. People who play by the rules imposed upon them, who work long hours if corp wants them to etc.<p>All the talk about diversity, but no diversity of thought. Only LC chasers. I genuinely believe this decreases innovation massively.

I hear you!<p>I quit doing engineering work for companies, but back when I did, I&#x27;d take some complex code I&#x27;d written to interviews. I tell the interviewers that I&#x27;m happy to step through my code and explain it during the interview, but that I wasn&#x27;t interested in wasting hours of my weekend writing a werpderzle or answering gotcha questions.<p>Some companies agreed and I ended up working for them and some companies didn&#x27;t, so I&#x27;d end the interview. In my experience, companies that over-complicate the interview process also over-complicate their day-to-day work, and I don&#x27;t have any interest in working in that sort of environment.

I wish I could, as a candidate, ask leetcode-style questions to my interviewers, too. If I am to work with them, they should prove to me that they can pass them, right?<p>Interviewers tend to forget that they are in a dominant position: they have nothing to lose, they are the ones who take the decision, and they are the ones who chose the questions. I have been in such interviews where I completely failed, but where I am absolutely certain that I could have made the interviewer fail as well were the roles inverted. It feels a bit hypocritical to fail someone who did not actually do worse than you would have done...

A good interview is an absolutely rare thing. I can&#x27;t name the chances, but throughout one&#x27;s career it&#x27;s going to be close to amount of jobs one&#x27;s going to land. More likely less than that number.<p>From my experience you can say it&#x27;s going to be good or bad right from the start. 5 minutes into an interview and you can already see where&#x27;s it&#x27;s going to end.<p>I don&#x27;t understand when companies try to give test assignments to senior devs. Even if your GitHub or whatever profile is not rich with projects you can evaluate person&#x27;s professional level during the conversation.<p>And I also hate stupid riddles during the interviews. But there was an interesting case. I got a couple of verbal puzzles that I should have solved right at that time and I failed. But I got the job anyway. So I guess if people in the company are common sense driven it helps to choose hiring strategies that work for both sides.

Everyone thinks they are google, if you are like 90% of tech companies and have a crud app, maybe a couple mobile apps and a few backend systems and admin tools, a radical idea is to involve your engineers in the interview design process and tailor your interview to the actual job they will actually be doing. Front end engineer and your stack is react? Pair with them on fixing a UI bug, backend engineer in rails, pair with them on a adding a basic endpoint. You get the point. If an engineer pairs for 30 min with someone I can pretty much guarantee they will have a strong sense of how they work and whether they know their stuff. I really don’t know why we treat this like it is some impossible task to determine if someone can do the job, just try doing the fucking job.

I don’t think anyone is saying in a traditional SWE role data structures and algorithms aren’t useful things to have some me amount of mastery about (higher levels required for higher levels of seniority maybe)… but it’s just hilarious and also sad that we’ve, as an industry, not been able to move away from these logic puzzles disguised as programming tasks that ostensibly should mimic a typical task but serve only to filter out groups of people.<p>Of course I don’t know the replacement I just know I loathe doing these things and would rather learn more about what the day to day would be like, what’s the team like, the manager, the growth potential, etc. I am all proving my skills but can I do so in an environment that is more congruent with my actual work?

There is an absolutely vast array of stuff you need to know to write software including networking, hardware, software, database, operating system, multiple languages, mutiple tools, standards, testing, processes, and on and on and on and on. It goes on forever.<p>BUT one of the things I&#x27;ve pretty much <i>never</i> needed is leetcode.<p>That says something about recruiting processes that evaluate you on your leetcode.<p>If you can&#x27;t come up with an interview proecss that evaluates a developer against the bajllion things that developers have to do, then it says YOU the interviewer don&#x27;t know what you are doing.<p>I think leetcode interviews are lack of imagination on the part of the interviewer.<p>Being good at programming does not make you good at assessing the capabilities of other programmers.

&gt; &quot;they don’t reflect the actual responsibilities of a software engineering of a job.&quot;<p>Exactly. I refuse to do them. Oh, your business is solving basic problems in computer science? I though you were an e-commerce platform that needed cleaning up your APIs and moving to a more efficient DB?

What you&#x27;re tired of solving algorithmic riddles that have very little to do with what you&#x27;ve been doing the last decade, all this under pressure with a shared screen in front of two bored interviewers? I don&#x27;t understand how you can get tired of it its so much fun!

I feel like a large number of software engineer positions interview as if they are looking for computer scientists.<p>The distinction is subtle, but important in terms of what is needed for the role.<p>Most software engineers are going to be asked to solve problems and actively avoid recreating algorithms and libraries from scratch while doing so: it represents added time, risk of bugs, and complexity to do so for minimal gain in many cases.<p>For most software jobs, the more important role when it comes to algorithms is in choosing which algorithm is most relevant to a task, not implementing it from scratch. To do so is often wasting your companies&#x27; time.<p>It&#x27;s unpopular, but most software engineering (including most of my own career) is closer to a plumber than a scientist. And you know what? We provide a lot of value as plumbers, as unsexy as it sounds.<p>It&#x27;s for these reasons that I think leetcode is an unsuitable test for most software engineering positions.

&quot;Leetcode questions are the worst form of interview questions, except for all the others&quot; - Churchill<p>I too used to hate leetcode questions, but after conducting 100s of interviews as an interviewer, leetcode style questions are the best way to test intelligence and coding skills ive found in general.<p>However, leetcoding interviews must be paired with system design and topgrading in order to have balanced interviews. Leetcoding alone is incomplete but i find it essential.

Leetcode is about proving you can give up huge chunks of your time to learn about the current industry fads, useless to life though they may be: This is unfortunately a vital work skill in the software development industry...

The only winning move is not to play.

I see LeetCode as a pure numbers game. Keep applying until you get easy questions and don&#x27;t get hung up on your failures.<p>I work at a FAANG but if I had another interviewer I probably wouldn&#x27;t.<p>For me it&#x27;s the same process as the green card lottery: apply to your target company every year &#x2F; six months (cooldown period) until you eventually get it.

The unfortunate reality is that If you don&#x27;t do this you&#x27;ll hire people who can&#x27;t program

Over the past 25 years, the only thing that has ever reliably identified successful candidates for my hiring was having 2-3 people (at least one &quot;cross-departmental&quot;) rate how much they wanted to work with the candidate, leaving it entirely up to them to determine how they arrived at that rating, and only hiring the highest rated people. I still performed superficial screening&#x2F;filtering of resume&#x2F;cv but there are very few roles in very few businesses where the &quot;leetcode-style&quot; of interviewing tells you anything relevant whatsoever.<p>The industry only does it because it FEELS LIKE it tells you something relevant, and it FEELS LIKE you and your business must be important if this is how you interview people.

I have a friend who stopped being a dev and went into management because of this practice.<p>Makes sense, he is older and has two daughters.<p>I even considered leaving the field myself.

It seems to me that while passing a leetcode question shows that you might know a thing or two, not passing one doesn&#x27;t really show the opposite. You could be the smartest &#x2F; greatest coder ever and you just got a really unlucky question... Or it could be that you&#x27;re not great when you are nervous, etc...<p>Also, a lot of the time it feels like the answer to most leetcode problem is some sort of trick.... and if you think of it you are golden and if you don&#x27;t it sucks to be you. Most of the time I&#x27;m trying to write code that is NOT tricky... So in some ways I think the job itself will train you to seek simple clean solutions as opposed to tricks.

In my experience, there is a strong correlation between good programmers and the people who can solve these types of problems.<p>It&#x27;s a kind of audition.

My biggest concern about Leetcode is that it’s not actually that relevent to the day to day job of being a software engineer and has more to do with being a computer scientist which is an adjacent but different set of skills.<p>I’ve worked mainly on scientific software which requires a mix of domain knowledge and programming skills. I doubt many people who I’ve worked with would pass leetcode without spending a lot of time studying, because they primarily come from Physics&#x2F;Engineering backgrounds, but that doesn’t mean that they are not good software engineers, that the software they’re writing is not performant, or that they can’t offer anything valuable.

My problem with LC-style questions is that it becomes a race to the bottom.<p>At some point, you end up with questions that either lucky people or savants are able to answer. Yes, that might be an exaggeration, but it&#x27;s not too far off the mark.<p>Applicants discovered, long ago, that if you simply grind every questions on LC - you will likely get asked some of those questions wherever you interview.<p>Companies fight back by asking harder questions. Before you know it, a standard interview consists of LC hard questions.<p>And with the advent of LLMs, it is even harder to sniff out candidates if the interviews are remote.<p>At some point you&#x27;ll just have to re-do the whole process, and built it up from scratch.

Jerry Weinberg used to run a discussion forum for experienced software development consultants, aggregating many years of experience at many companies building significant projects.<p>Their consensus opinion was that a &#x27;work sample test&#x27; was the most reliable means of assessing whether potential hire would be a good fit. Bring the person in and pay them to work with you, in your environment, on tasks parallel to what is needed, for a short period (anywhere from half a day to a consulting engagement) and base your decision on what you learn.<p>It is a really strong positive signal about a company if you see them doing this.

Could anyone provide an example of what Leetcode problems are too hard or unrealistic for hiring&#x2F;interviewing purposes?

Managed to solve this problem by going into business for myself. Have applied to some potentially interesting positions, but got batted away as not good enough at coding quiz questions. If I ever feel the need to be treated disrespectfully by people with two or three decades less experience then maybe I&#x27;ll put my resume out there again. Probably not, though. Kind of interesting that I know many like me who have decades of proven experience but little interest in modern interview games and end up going different ways. Lots of value being left on the table, but there it is.

Well give how well the FAANGS all turned out, evil-money generating society destroying hell holes... I imagine it actually goes like this:<p>Im going to randomly and silently draw from one of 3 topics that I understand, but nothing that I don&#x27;t and would need to collaborate with you to solve. It must be something I have memorized so not to risk looking less smart than the interviewee. Good we have now hired the 10,000th copy of me, the company can now continue to produce the SAME garbage across 10,000 people. I have succeeded and can now be promoted to level 5.5 and 2&#x2F;5th.

My frustration at this type of interviews is more philosophical. In broad strokes, for some reason the employers desire and require a significant amount of effort and training from potential employees so that those could (in theory) then solve complex problems for the company; but at the same time, employer&#x27;s management _usually_ absolutely refuses to invest any noticeable resource in researching, designing and maintaining the processes they should be accountable for, including the recruitment.<p>Which leads to the current state of IT&#x2F;high-tech recruitment where things are, in most cases, so mismatched between ends and means, it&#x27;s not even funny. It might be very entertaining to look at how people hiring for senior+ infrastructure roles (SRE, DevOps, what have you) try to use &quot;leetcode&quot; style stages - unless you are a desperate job seeker. (and no, I don&#x27;t think an SRE or cloud infra engineer shouldn&#x27;t be able to code, I just don&#x27;t think leetcode-style tests are at all relevant as tests for that)<p>And at the end of the day you end up with an engineering org where in the intake funnel they ask you to [competitively] solve variants of the knapsack problem, but once in, you end up solving it in the form of &quot;how do I slice and dice a set of tickets within a completely irrelevant, misaligned and misunderstood &quot;Agile&quot; model so that can best pack SPs into a sprint and <i>also</i> do some actually useful work&quot;.

LC questions are a shitty way to qualify 3 traits that employers very much look for (and have proven to be correlated):<p>* Available time to prepare (esp. when holding a job) = available to work long hours<p>* Self-motivate themselves for a long time (esp. to do such bullshit work)<p>* Mental capacity to remember all the material (esp. given its useless outside of the interview loop)<p>The flip of the coin are people who can&#x27;t put in the time due to e.g. family obligations, don&#x27;t have the motivation, or don&#x27;t have the mental capacity.

They&#x27;re fine to filter fresh university graduates.<p>But if you have any sort of track record (previous employment&#x2F;projects you can talk about, active github account), they&#x27;re just insulting. They shouldn&#x27;t be wasting your time if they could just have a look at the things you accomplished.<p>That there&#x27;s companies who make all interviewees take them just means there&#x27;s not enough competent software developers with sufficient self respect to just refuse that kind of nonsense.

I agree it&#x27;s not a great way but a good compromise. When you think of Leetcode you can be thinking of hard questions. But this is not the case for many of us.<p>Not all of us get a chance to work with a great candidate pool. So, whenever someone fails to provide a solution given an integer list, find the pairs that sum is equal to N, I don&#x27;t care how many technologies they have worked with or how well they can talk about it, I don&#x27;t feel comfortable working with them.

All the worst kinds of engineers, the kinds of people that talk big and deliver small and slow, the people I know I’ll need to go and fix code for, they are always the ones that work on leetcode the most. These are the false positives that people pretend don’t get through with leet code interviews, the kind of people that don’t understand how node async actually works but can reverse a tree optimally off the top of their head.

I think the fundamental problem is that in SW engineering discipline, we judge skill according to knowledge of tools (like programming languages) instead of knowledge of the application domain. There are interesting historical reasons for this, and it is a double-edged sword. But I think it would be healthier for the profession to acknowledge that application domain knowledge can be at least as useful as tools knowledge.

What exactly about Leetcode interviews can you Google, except the actual solution to the actual question (which you can probably Google for non-leetcode interview questions too)?<p>I always thought that while Leetcode interviews are problematic because they don&#x27;t represent most of the work you&#x27;re going to do, they&#x27;re much better than interviews where you&#x27;re expected to namedrop specific API functions or design patterns.

Leetcode interviews are in general just incredibly poor signal-to-ratio; especially with the advent of LLMs making it easier than ever to just cheat on the problem. I would just generally assume that any company with that as an initial barrier isn&#x27;t actually looking to hire anyone at all, but for the sake of &#x27;appearing&#x27; to have employee growth.

Honestly after interviewing lately, I prefer Leetcode-style interviews to doing some takehome project. I’ve failed two assessments now where my project fulfilled the functional requirements set out, but I was rejected because I didn’t use their conventions (which were not specified) or they didn’t like the structure of the code. Not only that, after investing a few hours of my free time to complete said project. I just get vague, hand-wavy reasons as to why I’ve been rejected.<p>Leetcode-style any day, over that.

Leetcode style interviews are not great, but I much prefer them to take homes. However, some FAANG+ companies have a nice system (for internship), where they give a very small codebase (~100ish lines) to work on with some errors &#x2F; possible improvements and you go through those in the interview.

There&#x27;s a way you end the feedback loop that encourages leet code, the people who endorse it. Openly commit that your focus in hiring evaluations will be on real world problem solving specific to each service team and that each service team must maintain 3-4 hacking challenges that are direct representations of their work. For most of us that&#x27;s &quot;perform a code review of this junior engineers undocumented code&quot;, &quot;add some specific functionality to this API&quot;, &quot;add functionality to this API without breaking contracts and show me how you ensure it&#x27;s not broken&quot;.<p>That creates variety in the question bank, standardizes evaluation, and will only encourage engineers who really want to work on your stack and product combination.

It&#x27;s true that nobody is expected to balance a binary tree as part of the job but the point of these questions is to see how you approach the problem and how you communicate your solution. Given that you can&#x27;t perfectly predict how someone will do at the job, employers use leetcode problems as a proxy. Even those who memorize leetcode solutions must also memorize how it works and understand the solution. Given that the problem is random and you&#x27;re likely given multiple interviews, it&#x27;s unlikely you&#x27;ve memorized the exact problem and solution without cheating. If you&#x27;ve memorized enough solutions that it&#x27;s likely you&#x27;ve seen the problem before and you can understand the solution enough to present it, then you&#x27;re in the 0.1% and deserve to pass.

In my experience, the best kind of assessment for Software jobs is a reasonable take-home project that involves a private repo that you fork, work on adding a feature or fixing a bug, and then send a PR.<p>It&#x27;s much closer to day to day work than code challenges are, it doesn&#x27;t involve the stress factor of an interviewer staring at you while you work on a leetcode, and you can have a followup technical interview to discuss the take-home in more detail.<p>The most recent take-home experience like this that I saw was with Clipboard health, and I thought it was a great experience. If I remember correctly, they used Woven Teams to automate the process.<p><a href="https:&#x2F;&#x2F;www.woventeams.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.woventeams.com&#x2F;</a>

I lead tech interviews for E5&#x2F;6 level engs where I work.<p>Everyone knows how pointless it is. But it’s become more like paying the tax.<p>For the record, I typically give away the a-ha moment early on and see how the candidate is able to take the suggestion and work through a problem with me.

&gt; Especially since I know for a fact they don’t reflect the actual responsibilities of a software engineering of a job.<p>This is irrelevant. No interview can accurately reflect <i>actual</i> job behaviors.<p>Job interviews are a proxy. Either they produce false positives&#x2F;negatives or they don’t. It doesn’t matter that the proxy isn’t 1:1 with day-to-day.<p>I sympathize with job seekers dealing with leet code interviews. But I’m somewhat over the million blog posts that say the same thing. I’d be much more interested in hearing from the other side. Hiring is a nigh impossible task. People rarely share reproducible alternatives. Only hand wavey gut checks. It’s unfortunate this is the best we’ve come up with. Hiring is hard!

My biggest beef with this style of interview is that someone with commitments (family, sports, hobbies) outside of work, I have no time to study the format, spend months practicing questions, practicing the art of speaking my thoughts in a clear way while actually having time to myself to think and process, understanding the gotchas etc.<p>luckily, I have experience as a programmer and in a previous career, and have so far managed to mostly avoid it as there are enough companies where I am that I’ve been hired elsewhere. But it does make me wonder if I’ve either missed good opportunities, or am limiting myself for the future.

He&#x27;s right to be sick of it. There&#x27;s no sense in them.<p>How often do you find an employee who if you&#x27;d just asked them an LC, they would not have gotten the job that they have proven themselves incapable of?<p>The times I&#x27;ve met a programmer that we shouldn&#x27;t have hired, it could not have been detected by LC. This has tended to be things like &quot;guy does not want to work on this problem&quot; or &quot;this guy is suggesting things that are obviously bad designs, and doing it with an aggressive style&quot;. Rather than &quot;if only we&#x27;d asked him to implement Dijkstra we would have found out and not hired him&quot;.<p>LC also doesn&#x27;t ask what you want to know. I don&#x27;t need to know whether you can solve the Towers of Hanoi. I need to know that if there&#x27;s a problem we need to solve, and it is a disguised ToH problem, you will recognize that there&#x27;s an algo to be found, you will eventually find it&#x27;s this one, and you will eventually find the solution and code it up in a sensible way that helps the team in the future.<p>LC problems tend to be too short to decide whether someone actually has the skill that I really value, which I guess is gumption. We have a codebase that&#x27;s become crusty over time, will you take the initiative to clean it up, or will you just solve the little LCs that present themselves and add that to the spaghetti?<p>The interview form that&#x27;s worked for me over the last 20 years is simply to have a long technical discussion, meandering across a bunch of technologies and problems. You can&#x27;t prepare for it, and you can&#x27;t waste your time preparing for it. Even if you say you touched a bunch of stuff, when you run out of opinions, I will know the depth of your experience. If you have an opinion, you will also know what the orthodoxy is on that area, and you can explain why you agree or disagree. If you are disinterested, it will be clear.<p>Now of course this isn&#x27;t going to satisfy people who want something they can call standardized. Maybe a pair of twins will walk in with the same skillset, and one of them veers down one path and the other down another, so that the conversations have different questions. But I would wager that I&#x27;d find the same result.

Especially in this new era of AI assistance - these leet code problems with clear definitions and repeated online solutions are becoming less important to solve from memory or study.

The key point for me is - what does the company value? Don&#x27;t forget that you are interviewing the company too.<p>Does this company value grinding at solved problems, rather than actually solving new and interesting ones? That&#x27;s a dead end. I&#x27;d rather talk about how I spent my time building real stuff in prod. That&#x27;s generally how I respond to leet code challenges, pointing out the existing libraries and tools that already do it and discussing their tradeoffs. Pretty good way to find out they&#x27;re looking for an actual engineer or a code monkey.

I personally fight against this trend by almost entirely indexing on problem solving approach when I give leetcode interviews at my BigTech. For me at least it&#x27;s possible to get a strong hire recommendation even if you only get a brute force solution. I only care that I think you could get to a solution eventually, and are professional in the way you get there.

The best interview I ever had was with GitLab, where they make a realistic PR&#x2F;MR (but one that is obviously not actual code they&#x27;d use for anything, for the ones worried about doing free work for them) and you have to review it. It&#x27;s exactly what I do at the actual job at least 50% of the time, and I&#x27;m sure from their side it gives them a lot of room to look into how the candidate works&#x2F;communicates.<p>They then ruined it by still having a leetcode style interview afterwards which I found monumentally stupid, but oh well.

Joel Spolsky, whose influential articles started the Leetcode-style interviewing, agrees we need something better:<p><i>&quot;But today Spolsky has mixed feelings about his guide — and its impacts on hiring in tech. “So this idea that you’re going to have a rigorous interview where you bring people in on the whiteboard and you say, ‘Show me how to copy a string while deleting the letter Q. In C, on the whiteboard.’ That was a big step up, I think, from, you know, ‘What college did you go to and who do you know and where did you work?’” [...]<p>“It’s a great way to hire 10 developers. It’s a very bad way to get developers in a scarce environment where you’re trying to find the people that might be good.”<p>A lot of good programmers end up getting rejected — while, even worse, companies end up hiring people who are good at passing tests, but underperform in the real world. “I think that method is definitely state-of-the-art 1995. It was even good for 2005. For 2018, I think you need a better system, and I think it’s probably going to be more like an apprenticeship or an internship, where you bring people on with a much easier filter at the beginning. And you hire them kind of on an experimental basis or on a training basis, and then you have to sort of see what they can do in the first month or two.” </i><p>Article: <a href="https:&#x2F;&#x2F;thenewstack.io&#x2F;joel-spolsky-on-stack-overflow-inclusion-and-how-he-broke-it-recruiting&#x2F;" rel="nofollow">https:&#x2F;&#x2F;thenewstack.io&#x2F;joel-spolsky-on-stack-overflow-inclus...</a><p>HN discussion thereof: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=39748546">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=39748546</a>

<i>I don’t really have a solution to this problem, I just know it’s a problem.</i><p>Neither does anyone else. Everything else we’ve come up with either has worse issues or is extremely expensive to operate.

Leetcode test would be fine, if they were formalized with automated scoring and nobody giving you hints&#x2F;silent treatment. Unfortunately, proponents of &quot;normal&quot; interview process will tell you that it&#x27;s necessary to observe engineer working on a problem to know how good they are.<p>Maybe they are right. After all, I&#x27;m not an expert on the matter of recruitment. But from my experience, I felt far better when the process was not &quot;personal&quot;, even if I didn&#x27;t get accepted in the end.

I’d like to know how other professions are hired. Do CPAs get challenged with hard tax problems and no access to reference material? Do plumbers replace a toilet in an interview?

That would be OK for top-end jobs where one makes $1M but nowadays everybody is subjecting devs to this torture even if they offer $40k just because the big boys are doing it.

&gt; I don’t really have a solution to this problem, I just know it’s a problem.<p>The solution is simple: refuse to do those interviews. I have, and I wish more engineers would. It&#x27;s one of my screening questions: do you do whiteboard style quiz&#x2F;brainteaser interviews? If the answer is affirmative, I politely (and sometimes impolitely) decline.<p>I&#x27;ve written a book, edited another, contributed to plenty of open source, including Golang, started a few companies, and been around the block. I&#x27;m more than happy with a take-home or with going over code I&#x27;ve previously written, but doing leetcode in a &quot;live coding session&quot;? I&#x27;m good, thanks.<p>For FAANG and other huge companies, it&#x27;s probably a decent heuristic (though I doubt even that), but smaller startups or non-tech companies doing this nonsense is a clear red flag. I mean, there&#x27;s entire <i>books</i> dedicated to gaming the software engineering interview. The signal-to-noise ratio is probably so terrible, it&#x27;s essentially just professional hazing at this point.

Interviews are incredibly noisy signals. It is far easier for someone to blag about their AWS experience than for someone to bluster through an unseen leetcode problem.

It&#x27;s weird to me because I haven&#x27;t had a single person ask me Leetcode questions, and I never opened Leetcode once in my life. And yet I interviewed plenty and got job offers aplenty. Who asks these questions? I&#x27;m I somehow simply self-selecting out of these types of companies via some proxy feature I&#x27;m unaware of? Or maybe they stop asking these questions when you&#x27;re more senior and not fresh out of school anymore?

I&#x27;ve found most of ukranians as well as russians love this style of interviews. I&#x27;ve also found them to be bad contributors to teams I&#x27;ve worked in.

I think the &quot;algo genius&quot; type of developer is a harmful myth of the industry. In most of the SE jobs you simply need dedicated, motivated persons who love the product and are able to organically grow to what the team needs. Places idolizing leetcode are in the stressed underorganised &quot;scale to the moon&quot; phase, where there is a need for individuals(!) to solve whatever is thrown at them at speed.

They are just bad and it’s an indication of laziness in the team, or that the organisation is hiring centrally and you could end up doing anything. Both of these are negatives imho. Developing a relevant technical interview aligned with what the team does is a project in itself and takes significant time to do and keep up to date. People don’t want to invest the time, but from experience it’s definitely worth it.

How does this look outside coding? Do we know of any other professions where there are whiteboard &#x2F; leetcode style interviews?

This is a place small companies can get an edge, I feel. They don&#x27;t have a zillion applicants, so they can be more hands-on in the interview process, and the interview process itself isn&#x27;t clogged with years of cruft. They don&#x27;t have to rely on &quot;good enough&quot; weedouts like Leetcode.

it sucks, but honestly having a very clear metric for success is significantly better than the wishy-washy interviews that predated this. i&#x27;d much rather grind and be prepared to just regurgitate some algo&#x2F;code and explain it than deal with the color of the sky being wrong, interviewer woke up on wrong side of bed, i was coerced into doing a &quot;project&quot; that equated to just unpaid work, etc.<p>whether or not you succeed as an engineer generally falls down to whether or not you know how to code (pretty much anyone is good enough at this for most jobs), can you communicate effectively (covered during the behavioral portions of interviews), and do you work hard.<p>honestly i think it sucks, but in comparison we&#x27;re pretty lucky. i think a lot of people in this world would be happy to grind a few hours a day for a couple of months to get paid 250k+ base

When I read posts like this I&#x27;m always surprised and confused. I&#x27;ve had quite a few interviews for software jobs in my life but not a single one was leetcode style or with a whiteboard... Is this just a cultural difference between countries? (I live in germany) or was I just lucky?

God I hate these interviews.<p>I&#x27;m 45, focus on Web development, have worked in-house on teams building ecomm carts&#x2F;solutions for multi-million dollar per ear local companies, in design agencies focusing on custom $100k or templated $5k Wordpress websites, and have freelanced for everything in-between.<p>During COVID lockdown, I got burnt out at my existing employer, resigned, and took several months of. When I started looking for work again, one potential employer was a Wordpress &#x27;freelance&#x27; company that gave me a LC test as a second round interview.<p>I BOMBED - like a 27%. I&#x27;ve been building sites and web applications for 15+ years and had zero idea what I was even being asked in any of the 3 questions on the &#x27;test&#x27;. The company ended the interview process and advised I study Leet Code for 1 month and come back to re-test since my experience was outstanding for what they were looking for.<p>I did and got an even WORSE score: 13%. I spent way to much time on trying to cram useless (for my daily work) programming techniques to pass a useless interview test when I was told I had the experience they wanted - I&#x27;ve built custom WP solutions&#x2F;themes&#x2F;plugins since wordpress first rolled out.<p>TLDR: With 15+ years experieince in Web development, I still couldn&#x27;t score higher than a 27% on a LeetCode test and got halted in the interview process for a position I was told I was WELL qualified for.<p>LeetCode tests for non-overly-technical positions is awful.

There is a lot of academic research on interviews. I&#x27;d like to see one of these blogs actually find and reference it sometime. Instead we get a lot of why someone &quot;feels&quot; something works&#x2F;doesn&#x27;t work, but no reason to believe they are correct.

Small feature request based interview?<p>We ask candidate to pick any OS project out of few dozen related their skillset. There are predefined feature request&#x2F;improvement we got prepared for each one. I&#x27;ve been experimentimg with it, and it feels more &#x27;natural&#x27;.

I am oddly fine with leetcode interviews. Mostly because they are predictable, and you can study them over time.<p>They obviously are useless to measure how good a candidate is, but that is not my problem. I didn&#x27;t pick the song, I just have to dance to it.

I quite like them. You know what sort of question you’re getting and they’re really not that hard. Practice. Get good at walking trees. Answer too slow? Sort the list first, use two pointers, use a hashmap, max heap…

Interviews cover collaboration compatibility. Not enough time to quiz you on tech knowledge. Leetcode-style interview represents your ability to work a problem collaboratively with a team

The solution is simple: since past behavior is the best predictor of future behavior, just let the few top candidates do the (paid) work for a few day&#x2F;week&#x2F;months and then keep only the best candidate.

Hot take: I like leetcode-style interviews and I&#x27;ve spent a lot of time in them on both sides of the fence (tens of hours as a candidate, hundreds as an interviewer).<p>&gt; But yet, these interviews quiz me on things that I can easily Google that I may not know off the top of my head. It’s absurd.<p>I don&#x27;t work at Google so I&#x27;m lucky to have freedom in the way I interview engineers and usually I try to squeeze out the quiz aspect as much as possible.<p>If a candidate doesn&#x27;t know the algorithm for a given problem and can&#x27;t quickly come up with one, I usually proceed to give a detailed explanation for it. In the end the task is reduced to &quot;write between 6 and 10 lines of code that do X in your favorite language&quot; and yet a majority (80-90%) of inbound candidates for software engineering fail at it (for non-inbound sources a picture can be very different of course).<p>Contrary to a common view, I believe that this approach actually imitates my daily interactions with a developer quite well and the correlation between interview performance and job performance is very strong in my experience.<p>And interestingly I&#x27;ve seen many people who shine in oral interviews (tell me about your experience) and then fail to produce any code at all, both during the interview and at work. So even if I wanted a good alternative approach to hiring of engineers I haven&#x27;t seen it so far.

The soft engineer is too much,but the job count is too little.you can easily google answers for these problems,but they like to test you by this method.

I wonder if its just to see how someone handles stress.

I&#x27;ve only heard of this crazy stuff with weeks of unpaid interviews and tests for programming jobs. Is this common in any other fields?

Basic coding problems involving data structures and algorithms are table stakes for a programmer. This whole thread is embarrassing.

I have so far not done a full Leetcode myself and I dont consider it an especially good approach.<p>But I want to share a couple of thoughts.<p>If these interviews in general, follow a form of template as to what they contain, what to expect, how answers should be presented, then studying it, while time consuming is not that hard.<p>It can function as a form of &quot;proof of work&quot;. That you took the time to prepare yourself for the interview will be points in your favor. Like putting on some nice clothes and brushing your hair. and putting together nice CV.<p>The second part is to perform under stress. I hate this part of modern interview processes. They want to give you something difficult, then spinkle with some rudeness and abrasive comments.<p>Will the candidate work under pressure or not?<p>Coming from the consulting world, perhaps that is more common in that sector than others.

I am leaning towards LC style interviews over a take home assignment that is meant to take “no longer than 2 hours” to complete

Sour grapes;I&#x27;ve never ever had problems with LC style interviews, and much prefer them over behavioral BS PM-style ones,

I am good at leetcode interviews. So are a lot of people reading this. For the most part, this is for the same reason: at some point we sat down and practiced until we were good at them. It’s a useful skill to have and I recommend to people on job searches that they learn it.<p>Would I recommend it if there weren’t so many leetcode interviews out there? No, I wouldn’t. It’s not a useful skill for actual programming; it won’t make you smarter. It will, however, establish that you have a lifestyle that affords you the kind of free time to devote to it.<p>So yeah, I’ll happily jump through these hoops. Honestly, I’m the personality type that finds it fun. But it’s a lousy way to assess my abilities as a dev.

This person doesn&#x27;t even spell check their writing. Imagine the code reviews six months out.<p>&gt; Kuberentes

If I am interviewing someone and if they ask me if they can google X, if it is easily googlable, I would just tell them what it is. Or let them google it. But I do not expect them to spend minutes googling it.<p>It is really unlikely, given a leetcode hard level problem to solve within 25-35 minutes and you can google your way out of it unless you are googling for the solution.

And I am so sick of „I am so sick of leetcode-style interviews“-style posts. Yet here we are. How much more can this dead horse be beaten? Aren’t you tired of discussing this topic with the same arguments each time?

I agree. Thankfully, you don’t have to take them and you don’t have to give them!

Same here. I miss the days when you were given a take home or, god forbid, a real world problem, something the company has run into that you can work through with an experienced interviewer.<p>Some of the worst interviews I&#x27;ve had I&#x27;ve felt like I was a lab rat, just being evaluated for specific traits&#x2F;reactions. The worst part is, like you said, every Tom, Dick, and Henry does this, its somehow become some kind of lazy industry standard hiring practice, hence the explosion of these interview aid services. Some people will argue it serves a purpose, but I really don&#x27;t see how. It&#x27;s like memorizing an answer sheet and doing well on a test. Is that really the type of problem solving you&#x27;re looking for? Don&#x27;t get me wrong, I do think its important to understand foundational concepts and algorithims, but like.. it&#x27;s also the first thing thats abstracted away for good reason in the real world. Anyways its shit.<p>The worst part is even with my experience I still feel like shit after these calls though I should know better.

I used to (still kind of) hate Leetcode-style interviews. But now I see it as a lesser evil, like Churchill&#x27;s description of democracy.<p>A while ago I posted a Ask HN question about the usage of Pseudo-IQ test which is <i>very</i> common in Swedish interview process. Pseudo-IQ test in the sense that it is almost always just Raven Matrix test + maybe math or personality test. I tend to score pretty well on those tests and it takes maybe 1h tops, but it&#x27;s annoying and stressful either way, especially knowing what ideological foundation it comes from....<p>Seems like the answer from the Ask HN point towards that those tests <i>used to be</i> common in US&#x2F;SV&#x2F;Software position interviews too, but it then got kinda replaced by Leetcode-style interview.<p>And sure, having a person staring at you commanding you to only use your brain to solve a toy problem is <i>definitely</i> not how you work. But it in combination of a bit of back-n-fort traditional interviewing is definitely more informative to the employee (and fair) than some pseudo-IQ test.<p>I just hate when Leetcode is used as *the first screening* it&#x27;s such an asshole move that show high level of disrespect towards the applicants time. Maybe you can get away with as FAANG, but small companies that do this to me just signal extreme arrogant (or delusional&#x2F;dysfunctional) HR or C-suite.<p>The best coding interview I&#x27;ve had is just the format &quot;You&#x27;ve received a PR from a junior dev. The code compiles without error and pass the unit test. Would you pass&#x2F;reject this PR and motivate why&quot; and it often requires a better understanding of the language (such as difference between modern C++ and C++12)

the solution is to start your own company and not do them<p>be the change and so on

there are plenty of people who would nail leetcode. but they are not even allowed for interview due to visa&#x2F;immigration restrictions.

I still believe that the Leetcode style interviews are the great equalizer that provides opportunities for folks coming from non-elite Universities to try their luck with FAANG.

Don&#x27;t hate the player, hate the game.

Nobody dislikes these sorts of interviews more than me. I&#x27;m not sure they&#x27;re particularly useful for anyone and I _know_ they don&#x27;t show off what I do well for employers particularly well. That said, prepping for interviews is just one of those things I&#x27;m willing to do for money. Is it worth $25K a year to me to spend an hour a day for three months working leetcode problems prior to interviewing? Probably.<p>I mean, is it a great way for companies to hire? No, not really. Is it how many companies hire? Yes. Do I occasionally learn a few things while doing this sorta prep? Occasionally. All-in-all, the expected value from a bit of prep seems worth it.

Don&#x27;t hate the payer, hate the game

I&#x27;m maybe more cynical, I believe that these types of interviews don&#x27;t filter out less smart people, but filter out people who don&#x27;t object to the processes of Big Tech. I think they mainly serve to pass through people who aren&#x27;t going to question the system, and won&#x27;t speak up when during their day to day trenches they&#x27;re implementing ethically bankrupt software. Intelligence has nothing to do with it, it&#x27;s about ensuring the capitalism machine keeps going without any revolt.

happened to me in an interview:<p>them: implement reversing a linked list<p>me: your company spends a lot of time reversing linked lists?

I&#x27;m almost as sick of reading about them.<p>&gt; I don’t really have a solution to this problem, I just know it’s a problem.<p>Yeah, exactly. You don&#x27;t have a better answer and nobody else does either.

I share the exact feeling as you do, it&#x27;s a lot of bs that has almost no value. None of those problems reflect what business needs to solve; everything today can be Googled and fixed with straightforward code.<p>Usually, the computing part isn&#x27;t the part that prevents a business from scaling more (servers are cheap).

yeah classic FAANG bs from consultancy leaders who can not solve 1 leetcode problem and don&#x27;t @ me with they don&#x27;t need to because they can&#x27;t make business decisions either even with fancy economy degrees and then layoff half the staff. If your CEO was not a techie the company is doomed anyway. So FAANG pushes leetcoding and everyone does it now, I just learned the 100 most common for interviews by heart and dabbled into 1000 leetcode problem. Absolutely only need to put in work and has 0 skill tbh it reminds me when I had vocabulary tests in school for other languages, just learn it by heart and score high mark. No skill of the daily business life involved at all. Landed a job that way. Improvise, adapt, overcome.

&gt; Especially since I know for a fact they don’t reflect the actual responsibilities of a software engineering of a job.<p>I had one job working on a compiler that fairly regularly had leetcode-style algorithms problems, even novel ones.<p>But yeah, most don&#x27;t.

I am still not sick for this. (╥‸╥)

Imagine what Franz Kafka could do with the Software Engineer interview process. It practically writes itself.

here here

leetcode is ok. I am so sick of visa&#x2F;immigration restrictions

Word.

Let me put this straight. You are just lazy and those interviews are intended to filter out lazy engineers. Training for coding interviews is not a problem. You just need to spend 1-2hr daily and solve couple of hundred leetcode tasks total for less than a year. After this you&#x27;ll not have any problem with this type of interviews at all. Additionally they boost up your coding skills and problems solving skills. You do not realize it yet because you did not train.