Hacker News Reader

Comment by liranz | Hacker News Reader

This looks interesting, though I don't agree with the "most candidates cannot write code to simple algorithm" thesis.

I interview A LOT. Almost 100% of the candidates I receive cover the basic programming abilities. The real problem is that they don't know how to THINK.

I talk with them 20 min on the phone when they have to present a problem they had and how they solved it. Then I try to find out what they find interesting.

Giving a mini project is a good idea, but it does not scale well, and still requires a lot of resources top candidates will just not be willing to invest.

View on HN

There's a fence here - a bimodality if you will. I've been on both sides of it.

I used to do a large amount of phone screening for Amazon, probably over a hundred in all by the time I got sick of being a professional interviewer. I experienced much of what the author did: the bulk of the candidates had zero algorithms or data structures knowledge and could not program. Not "can't program well" or "can't program at a professional level", we're talking "I don't trust you with a simple shell script".

Now I work for a rather more obscure startup - we're relatively well known in the tech scene around here, but we're by a very long shot not a mainstream household name. The caliber of people that come through the door now is vastly different. I've interviewed some duds here, but have never interviewed someone in this position that just outright couldn't code.

The question I have for the author is who he hires for, and how much they pay. The low end of the software industry is plagued with people who have no business calling themselves programmers. It's really a vicious cycle - there's a large segment of our industry where the typical skill level is disastrously low, which causes ever more bureaucratic and ever more absurd filtering mechanisms to be invented, which further alienates programmers who have any real choice in employers.

Nowadays I'd be mildly insulted if given a FizzBuzz during an interview - but I used to do the same out of necessity, because 25% of my candidates couldn't pass it.

A simple programming task adds 20 minutes to the burden of the interviewee, costs you next to nothing and weeds out the folk who don't actually know how to program. If someone fails to print prime numbers below 100 in 20 minutes, there is something wrong with his skills.

And is this a buyer or a seller's market? If it's a buyer's market then you can give the interviewee as many tests as you like. If it's a seller's market... you'd better have something really good going for you!

That would seem premised on the idea that it is better to have someone barely capable in the position than to go without. In my own personal experience I'd rather be short handed than work with someone incompetent.

Your outlook is only correct if jerking applicants around with meaningless trivia questions and irrelevant tests actually, somehow, probably accidentally, results in the selection of world class developers.

If there were 20 plumbers and only 10 plumber jobs, I could get away with jerking them around forcing them to do crossword puzzles before hiring them to fix my toilet. That doesn't mean the selection pressure would magically result in the guy most likely to fix my leaky toilet; much more likely it would result in the guy best at doing crossword puzzles. Unfortunately I now have a concrete deliverable of a leaky toilet to fix, and the best crossword puzzle solver in the market trying to fix it. That isn't going to end well. Especially when the best leaky toilet fixer in the world knows he can refuse the crossword puzzle test and get a job via his rep or connections so he never even applies. So its not just a random selection of applicants, but a random selection of an inferior population.

I understand where you are coming from but personally I believe that asking a programmer to produce a small amount of code is certainly relevant.

Don't get me wrong, it would be possible to pick nearly worthless questions which prove nothing about a candidate but a well designed problem requiring a few dozen lines of code and no 'trick' or specific knowledge can tell you more about an applicant than many hours of trying to code on a white-board or doing silly logic puzzles.

That would indeed be a terribly way to hire plumbers! But happily, that is not anything like what the article is suggesting.

The article suggests (1) using a simple but relevant exercise and (2) using it as a screen, to select people for a real interview.

The analogy with plumbers would be something like asking them to list the tools they would need to replace a hot water cylinder. For someone who knows their stuff, it's trivial. But it will weed out the guys who are bullshitting you. The guys who can give a good answer, you ask to come in for a proper interview.

I actually prefer working on a programming assignment when applying for new jobs. Most of my side project code is undocumented with few tests, so getting a coding assignment gives me the chance to prove that I do include those elements in paid work.

1. Write program that prints prime numbers below 100.

2. ?

3. Profit!

Interviews are a two-way street. If you're looking to hire an experienced developer and the best you can do is ask him to do something that he'll probably never have reason to do on the job, you're sending a pretty strong message about your company.

If you're going to ask somebody to code (or solve a problem) as part of the interview process, there's nothing worse than being lazy in creating the task. "Print prime numbers below 100" falls under the lazy category.

If the best you can do is to set a trivial programming problem, then indeed, that is deeply unimpressive.

Happily, that is not what the article is suggesting, nor what anyone actually does. The article gives three steps: first, a trivial programming problem, second, a challenging programming problem (here delivered as homework), and thirdly, a face-to-face interview to check cultural fit.

The purpose of the trivial problem, as explained in the article, and explained by anyone else who has ever advocated this approach, is simply to weed out people who just cannot code at all, and to do that early and cheaply.

My current employer uses a trivial programming problem like this as an early screen (it comes after our recruitment guy has read their CV, but before we put them on the phone with a developer). Therefore, i, and many of my colleagues, are experienced developers who have been given a trivial programming problem to solve. It did not any of us a message which put us off the company. I have not heard of anyone being asked to do this test and losing interest in working for us as a consequence. On the contrary, i was pleased when i was given the problem as an initial screen, because it told me that the company was serious about hiring people who can actually code, which puts them head and shoulders above many employers!

> check cultural fit.

I see lots of references to "culture" in hiring comments, so I'm not trying to call you specifically out, but what does it mean? It always bristles me, like if your criteria is "white men only need apply" well just outright say it, don't tiptoe around it. As a suggestion to anyone on HN who does hiring, if for example you're not going to hire my wife because of her ethnicity or because she's a woman, just put it in the job ad and save everyone involved a lot of time and tiptoeing around "culture".

I read it as "people from Standford aged 19-26"

See that's what causes frustration about "culture BS". OK fine it means you won't hire my wife because of ethnic reasons and gender reasons. Or wait, does it really mean you won't hire me because I'm a perfectly good white Germanic guy but I didn't go to Stanford. Or WRT not hiring my wife, does it mean my daughter need not apply because of her gender is still a cultural mismatch, her being a girl and all that, or is it enough of a cultural match that her dad is a pasty white germanic dude so it could be worth her time to apply?

Its somewhat frustrating that "what defines a good developer" to a boss might be being a white male brogrammer from Stanford. But whats even worse is the uncertainty of not understanding what the criteria even is. Is it only whites need apply, or only men need apply, or both, or neither? That is what makes it even more aggravating. Its not 1950 anymore boys, its a multicultural world, so demanding a certain culture/demographic fit is going to result in epic fail if any competitor is smart enough to (gasp!) hire women or people of color or dare I suggest it, non Stanford grads.

Startups and people who hire should realize that they're turning off good applicants when they start talking about "culture" because people of color are going to assume you mean white people only, women are going to assume you mean hire men only, etc. As a business decision its just a bad move.

If a company doesn't want to hire me (or my wife or my daughters) because I'm not a Stanford grad from the right demographic, that's absolutely fine because I don't want to work for a company run by idiots.

In fact, the louder and more explicit they are about their dumbass culture the better, because it makes it easier for me to avoid wasting my time with them.

I outright ignore any job postings talking about "cultural fit"

Imagine hiring a manager who expected their reports to be in at 9am sharp, when the current company culture is to work whatever hours people want. Or someone who very strongly preferred a more methodological approach to development to a "ship fast and break things" approach in a company that values the latter. Or someone who wanted five weeks of vacation in a bootstrapping company where everyone was working 60 hour weeks (even if they were willing to accept less). Or someone who wanted 60 hour weeks in a company with a core value of work life balance.

In no case is the person wrong in what they want, but it might not fit in with the current culture, which can be bad for everyone involved.

That said - I agree that this can manifest itself as a never never land (the culture is drink till 4am on weekdays!), but I do still think that it is an important aspect of any organization, and properly applied can be beneficial.

LOL don't take it personally, but that is completely wrong.

Culture is not issued at your previous employer along with the company policies handbook, id card, and a dress code.

Culture is what you are, not how your immediately previous employer temporarily defined you while at work. My culture is I'm a northern European descended male living in USA, for my whole life I've been whats now called a "maker" for decades before being a maker was cool, religiously unaligned with any particular deity and live and let live with everyone else's imaginary man in the sky (and why is it always a "man in the sky" anyway?), educated in the classical "great books of western civ" sense, and my tastes in entertainment are vaguely European in outlook for example sex is great fun but violence is sick (I'm not a big fan of hollywood gorefest movies). And being a mostly German descended guy I have a pretty tidy front lawn in front of my house, that stereotype is often true.

Culture is not "$current_employer tolerates flextime so I could never work in the future anywhere else that doesn't have flextime because all human beings are permanently inflexible, oh wait, actually I've never worked the same schedule at two difference consecutive jobs so I guess it doesn't matter anyway"

Everyone knows 60 hour weeks means reduced productivity compared to 40 hour weeks, so I can slack off if I have to, to fit in with the other burned out sleep deprived zombies, but rest assured I'm only going to work like a demon for 40 hours and then surf the net for 20 because you get there the fastest by going slow.

"Culture" is a bullshit nonspecific term and i apologise for using it. It's a word the article used, and i was trying to stick to its language.

For us, it just means "will we be productive and happy spending the working day with this person?". Somebody could be a great programmer, but might not be someone you're going to get on with. They could be too quiet, too loud, too aggressive, too passive, too arrogant, too sensitive, too confident, too insecure, strangely obsessed with tubs of jelly, etc. Great programming skill is a great resource, but if a programmer can't get on with their colleagues, that resource cannot be exploited.

We would not rule you or your wife out on grounds of gender, ethnicity, nationality, religion, university, taste in literature, or the state of your lawn. If you for some reason wanted a job with us, and passed our technical tests, then all we would want to know would be that you could sit down and be an effective pair in programming, and stand up and be an effective participant in a discussion. Ideally, we would also like you to be good company at lunch or in the pub, but that's not essential.

Sounds like the word you (and many other hiring people) are looking for is "Professional" which pretty much meets all your requirements without borrowing an unrelated word. There is a strange aversion to it industry wide. Unclear why. My guess is some peculiarity in .gov regulation WRT overtime pay or maybe contractor status.

> They could be too quiet, too loud, too aggressive, too passive, too arrogant, too sensitive, too confident, too insecure, strangely obsessed with tubs of jelly, etc.

How about too picky? Hint hint.

> We would not rule you or your wife out on grounds of gender, ethnicity, nationality, religion, university, taste in literature, or the state of your lawn. If you for some reason wanted a job with us...

How generous of you!

The language one chooses to use can reveal a lot about one's thinking and here, "rule you or your wife out" and "for some reason" is, I would argue, very telling.

Yeah, i don't think we've ever had a problem with anyone being too picky.

> Interviews are a two-way street. If you're looking to hire an experienced developer and the best you can do is ask him to do something that he'll probably never have reason to do on the job, you're sending a pretty strong message about your company.

I know others will probably disagree, but I feel this way about brain-teaser questions. I remember one very specifically from a ex-Microsoft employee who interviewed me: "If you're alone on a deserted island and you have two unmarked jugs, one that can hold 3L of water and the other 5L of water, how would you make sure you had exactly had 4L of water?"

To his credit, he helped me through the question and I was able to figure it out.. but the first minute was spent trying not to be a smartass and ask "Is this a problem that programmers in this company frequently have?" (I did get the job fwiw)

Brain teaser questions are great at finding candidates who have heard the brainteaser before. They are useless at finding people who are good at coding. Even Google accepts this now.

I would ask why if I had 8 L of water containers why I would try to figure out how to have no more than half of that and just fill both jugs, then I'd say that's a stupid question and has nothing to do if I can code or not. I hate these stupid job interview things just be frank and lets talk about my skills and what you're looking for, not stupid brain teasers, fizzbuzz, or anything like that.

That's technically a memorization test or a trivia test, because its really easy if you remember the trick, and a total PITA if you don't remember the trick. If I recall correctly the trick is something along the lines of pouring a full 5L into a empty 3L and when the 3L is full there's 2L left in the 5L, so then you dump out the 3L then pour the 2L in the 5L into the 3L such that the 3L contains 2L of H2O, and then refill the 5L and pour the 5L into the 2/3 full 3L till its full such that about 4L remain in the 5L bottle. Ta Da, I've just proven I can memorize meaningless trivia, does that mean I am now a C++ developer or whatever they were interviewing for?

If you started life as a chemist this is kind of a lab joke as in its funny but if anyone in the lab tried something this wasteful of time and product or risky of contamination if you reuse the product, they're too stupid/lazy to work in your lab. The correct answer is supposed to be walk to the stockroom and ask for a 4L container.

Its about as useful as a practical programming test as asking who won the 1984 baseball world series, in other words a complete waste of time.

Fizzbuzz at least is an excellent test of the modulus operator (oh boy is that ever useful to me on a daily basis LOL) and a so so test of control block and conditional knowledge which is actually useful on a daily basis.

Reply on news.ycombinator.com