Hacker News Reader

Comment by jarjoura | Hacker News Reader

Of course it does, it's not a gender problem, who said it was? How many of you have heard your friends and colleagues tell you, "wow I'm so lucky I got through that interview, there's no way I would pass it if I did it again today."

You shouldn't walk away from a good interview feeling like you nailed it. An interviewer should push you to your limits and the more you know, the more they should push. It doesn't mean you failed, it means you reached the limits they could push you in the time you had.

If you're rejected from the role, most likely you didn't reach the bar they needed you to be at, or you failed at the behavioral parts. Or, the interviewers you had weren't properly calibrated, which isn't to say anything about you other than bad luck.

View on HN

> An interviewer should push you to your limits and the more you know, the more they should push.

Granted, this is HN, but this viewpoint is very tech-engineer focused. Most jobs aren't like this at all. Chicken-parts processor interviews aren't like this, para-legal interviews aren't like this, firefighter interviews aren't like this, teacher interviews aren't like this, etc. For the vast majority of jobs out there, your certifications, references, and resume are more than enough (especially in right-to-work states in the US).

> If you're rejected from the role, most likely you didn't reach the bar they needed you to be at, or you failed at the behavioral parts. Or, the interviewers you had weren't properly calibrated, which isn't to say anything about you other than bad luck.

Hanlon's Razor states that the mostly likely explanation is not conspiracy, but stupidity. Most of the time, your interviewers are just dumb/inexperienced/busy/assholes.

Chicken-parts processor interviews aren't like this, para-legal interviews aren't like this, firefighter interviews aren't like this, teacher interviews aren't like this, etc.

Nor were tech interviews (usually), until about 10 years ago.

Nor are they many other places. I've never had an interviewing experience like this in Norway, for instance. We don't have a Google, Facebook, Apple or Microsoft, but we are reasonably ambitious when it comes to creating software and automating tedious and labor-intensive work.

> Hanlon's Razor states that the mostly likely explanation is not conspiracy, but stupidity. Most of the time, your interviewers are just dumb/inexperienced/busy/assholes.

What are the odds that everyone in your interview is dumb/inexperienced/busy or an asshole versus the odds that it's just you?

> What are the odds that everyone in your interview is dumb/inexperienced/busy or an asshole versus the odds that it's just you?

In a lot of tech interview loops, all it takes is one idiot or one asshole to torpedo your candidacy.

What are the odds that everyone in your interview is dumb/inexperienced/busy/asshole [and not you]?

The commenter wasn't saying that everyone they talked to was in that category. But rather that one person they talked to who basically tanked the whole interview for them. You know, the one who reported back everyone about what an uninspired dolt you are because you didn't answer that "Tell us about your dreams" question with quite the level of gushing enthusiasm they were expecting.

So what are the odds that you'll run into at least one person in that category, in a random tech interview these days?

When we take into account how toxic and culturally dysfunctional the tech industry in general has become in recent years -- I'm speaking specifically of it is huge vanity problem (from which this "we only hire the best" nonsense stems); combined with its insatiable addiction to bullshit and the cargo-culting of bad practices (and its worship of dumb and/or useless interview questions specifically) --

actually, fairly high.

>What are the odds that everyone in your interview is dumb/inexperienced/busy or an asshole versus the odds that it's just you?

Let's assume that your interviewer is selected at random from the population and that traits are independent from each other. Let's further assume that 50% of people are 'dumb', that 30% of people are inexperienced, that 90% of people are 'busy', and that 50% of people are assholes. Honestly, I think these estimates are kind to most people in corporate America, but we're going from a random sample of the population. You can play around with the odds yourself and see what you get.

Now, my probability math isn't the best these days, but I can give this a go:

The odds that your interviewer is all four of the listed traits is ~7%.

Writing out the matrix for all the probabilities is going to take a while and really isn't formatted well for HN, so I'll skip most of them and just focus on the dumb assholes. So the odds that your interviewer is:

- dumb, inexperienced, and an asshole is ~8%

- dumb, 'busy', and an asshole is ~23%

- dumb and an asshole is ~25%

Now, that's just a single interviewer. You typically get a few rounds of them. Let's say that we have three interviewers to go through. What are the odds that at least one of them is a dumb asshole? Well, again, my probability math isn't the best, but I think that you just add the events together. So, the odds that you have at least one dumb asshole to be interviewed by is ~75%. Again, though, probability math is a head-trip to me and I always have to relearn it when I do it.

Now, you asked what the odds were that the person being interviewed is the one at fault. Again, we have to assume that we are being picked at random from the population. I'll use the numbers as above and assume they are randomly distributed in a person likewise.

The odds that both you are all four things (dumb/inexperienced/busy/asshole) is still just ~7%, and likewise for the rest of the permutations. For the sake of brevity, let's assume that the interviewee is dumb asshole and that the person giving the interview is NOT a dumb asshole. The probability that they are not a dumb asshole is ~75%. The odds that you, the interviewee, are a dumb asshole are ~25%. Again, though, probability math is a head-trip, I think the odds are then just multiplied to give the odds of that particular interaction at ~19%, or just shy of 1/5th of the interviews.

Now, again, you have to go through three of these rounds of interviews, each time going through the independent probability math. Again, my probabilty math isn't the best here, but I think the odds that you are a dumb asshole and that the interviewers are NOT dumb assholes is ~56%. Again, assume random distributions of traits and random sampling of people and that my distributions of traits are accurate.

Hope it helps!

EDIT: Crap! You asked a question I did not answer (sorry): what are the odds that ALL the interviewers have NONE of the four listed traits and you have at least one of the traits. Well, the odds that a random person does NOT have ANY of the listed independent traits is ~2%. That ALL three of these hypothetical interviewers would not have ANY of the listed traits is ~0.000054%. Now, that you have NONE of the traits is ~93%, but lest focus on the dumb assholes again. So the odds that you have at least the dumb trait OR the asshole trait are ~50% ( Thank God I chose both those percentages to be the same!). So, given that you have at least one trait, and that the interviewers have NONE of the traits, what are those odds? They are ~0.000027%. Again, though, my math is not always the best, so please re-check.

A for effort.

> Now, again, you have to go through three of these rounds of interviews, each time going through the independent probability math. Again, my probabilty math isn't the best here, but I think the odds that you are a dumb asshole and that the interviewers are NOT dumb assholes is ~56%.

This is what I meant to ask, but I should've worded it differently.

> Well, again, my probability math isn't the best, but I think that you just add the events together. So, the odds that you have at least one dumb asshole to be interviewed by is ~75%

Why would you undertake a several-paragraphs-long probabilistic analysis if you are aware of your inability to do so?

It shouldn't be so hard to grant that one could harbour doubts about one's ability yet one could go ahead and make an attempt to do something. Also, I'd not read it as "inability" from the wording chosen.

Chicken-parts processor interviews aren't like this, para-legal interviews aren't like this, firefighter interviews aren't like this, teacher interviews aren't like this, etc.

Doing this sort of interviews on programmers probably wouldn't work even if current programmer interviews do suck for both parties and they're mostly embarrassing, too.

I'm guessing here and might be wrong but all these professions have a rather fixed set of limited skills they have to know in order to qualify. Teachers have to teach a certain curriculum: on top of that they can be bad teachers or good teachers but basically what they have to achieve each year for each class is known and predicated. Firefighters have their training and as a baseline it's sufficient to be a good team player and just execute. Firefighters will have to improvise at times, and they will have opportunities to make some really good judgement calls based on their experience alone, but that's sort of extra. You're still a good firefighter if you just stick with doing it by the book.

Now, the same for programmers would be something like "Must know Java and we'll hire you". This didn't end very well in the IT bubble twenty years ago. Knowing Java inside out doesn't make you a good programmer, not even a good Java programmer, there's so much more to that. Would you hire an artist based on "Must know how to paint in oil, watercolor, and draw in charcoal"? That's where the real skills begin, not suffice.

If you're a Java programmer it doesn't mean you can just sit down and program anything for any company: there are hundred domain-specific skills and questions, all sorts of CS theory of which you know one or two fractions based on what you've worked on before, and a very specific mindset that makes a good programmer for a specific role. To some extent you can be a generalist and learn a bit of this and that but especially for heavy specialists it could take years to transition from one specific subdomain to another.

All programmers are cathedral builders but they will all make very different cathedrals and they are continuously requested to make cathedrals that have not been built before. Then there are other jobs where they follow the blueprints and produce a cathedral one after another, meeting the same requirements each time.

I'm not saying programmers are superb to these other professions when it comes to complexity and depth. It's just that a programmer's job can be about anything: the number of problem domains is endless and more so as the world becomes more software than hardware. You can do so much with computers and programming that whatever baseline there exists alone doesn't yet take you anywhere as a programmer.

There's the fizzbuzz level of a baseline, there's the skillset, buzzwords and qualifications baseline, and there are a lot of other baselines that need to be met. That's why you need to figure out a way to interview programmers that takes all this into account.

There's no single interview that could be used to qualify people for being hired as a programmer because the requirements are never fixed.

Many of these people interviewing wouldn't pass their own interviews if the roles were reversed. The bar should be high, but not so high that you leave feeling demoralized.

As an interviewer, I've been in this situation. I hate interviewing other people and felt really guilty about it, but I didn't have control over what I asked. I've also been in situations where I wouldn't be able to answer the questions I'm asking if I hadn't just learned or reviewed them for the sake of interviewing itself. It's so stupid.

I wonder if I would. I joined my current employer "through the back door" so to speak, that is, without the interview or the 5+ years of experience they asked (as an apprentice) and have been taking technical interviews ever since. But am I actually competent enough?

I mean I've been working there for nearly 8 years now and haven't gotten fired yet so it should be alright. But still. [/impostor syndrome]

And given the possibility some interviewers experiencing their own imposter syndrome, around being qualified to assess qualifications, perhaps that could explain some of the questions too.

But the interviewee is just a cog in the machine. Maybe it's unreasonable to require these people too are top notch at their game.

>An interviewer should push you to your limits and the more you know, the more they should push.

Disagree. An interviewer should be trying to find out if you are technically and socially capable of performing the job.

So the interviewer should settle for just good enough and not try to get more for his money ?

Still is capitalism here...

"Getting more for you money" means hiring someone overqualified. That's always been considered a bad idea because of the generally accepted (and quite reasonable) belief that people who are overqualified will go somewhere where they're merely qualified, but paid a lot more.

So, yes, most interviewers are looking for the best fit.

As someone who interviews people I can speak for myself. No I dont look for best fit.

I look for someone who will be able to grow inside the company and fill gaps. Like to see them do smart things they are hired to do without being told what to do.

Your response is precisely why I said "best fit" and not "minimally qualified." Your definition is the best fit for your company.

However, if you're seriously arguing that you'd hire an MBA for the front lines of a minimal proficiency call center "so they can grow into the company," then you're exaggerating.

I wonder if I'm the only person who thinks an interviewer should assess whether the person can do the job, and that an interviewer pushing someone to their limits seems like an inappropriate activity for the employer-employee assessment process.

I feel like there are two different conceptions about what it means to "do a job" that are co-existing within the economy, and a lot work-related offense and disgruntlement occurs because of mismatched expectations.

In one conception, "a job" means that your boss sets out a list of tasks for you to do, and you do them, consistently and reliably. As long as you do what you are told, you will be paid the salary agreed upon. If you do it long enough, you will be promoted. Under this conception, there is no reason for a job interview to do anything other than ascertain whether you are qualified to do the tasks listed in the job description.

In the other conception, "a job" means that you're responsible for looking around the organization, identifying ways to either generate more revenue, make customers happier, or eliminate inefficiencies, and then go do them. Your boss's job is to remove roadblocks and introduce you to coworkers that can help you. The specific tasks you will be working on are unknown at the time of hire, and it is part of your job description to discover them. If you generate value for the organization, you will be promoted. Under this conception, it makes sense for a job interview to push you as far as possible, because if it turns out you're more qualified than the position, the position can be extended so you have more impact and get correspondingly higher compensation.

When people with the first conception apply to jobs with the second conception, they wonder why their employer demands so much of them, why there are no clear guidelines, and why they don't advance while people hired after them shoot right by them and end up becoming managers. When people with the second conception apply to jobs with the first conception, they get bored, step on other peoples' toes, make political enemies, and end up getting kicked out.

Adding to the complication, your boss is many companies is far more likely to hold the second conception of what a job is, because that's how he got promoted. Good managers can adjust their managerial style to make both types of employees productive, but bad managers assume everyone is just like them and complain about how employees are so lazy these days, while those employees complain about how their manager doesn't know what he wants.

Also complicating things, the first type of job is dying out. In many cases, they're being replaced by automation, and the jobs themselves disappear in favor of machines. In other cases, their boss is being replaced by an app, and those jobs become contract positions for Uber or Postmates or DoorDash or Amazon.

Great response. To add more details... the second type of managers respond very poorly to when their workers have no idea what is going on and can’t reach targets/goals. In this case they need more leadership.

The first types of jobs Should have good leadership where the complex parts have already been decided and all that’s left is the implementation.

The issue is what happens when a worker for a type 2 manager is smarter and knows exactly what to do? That employees best course of action in their perspective is to get their boss fired and take their role since they are more capable.

When the type 2 manager gets frustrated due to incompetence that becomes scary since they lash out at their underlings for incompetence.

Type 1 jobs are more relaxing on the psyche as you don’t have to think about

This may have been the best comment I have read on HN. Sounds like the foundation of a blog post.

Thanks, this is awesome, you put into words what I’ve been thinking about many times before but not so clearly!

Better yet, the interviewer should have experience in said job, as to better evaluate a potential candidate.

I'm pretty sure, most of us have come across folks who pretend to know something, you ask them easy, little, but esoteric questions, that prove whether or not they even have knowledge in something.

At least for big firms, this isn't out of the realm of possibility. Small companies... I get it. They have to then base it off experience, references and gut feelings.

> "wow I'm so lucky I got through that interview, there's no way I would pass it if I did it again today."

That's because most people wouldn't pass their interviews if you gave it to them again.

Unfortunately a lot of interviews these days are passed by grinding on LeetCode (that's not to say your experience doesn't get displayed in other forms) and then forgetting about these patterns & problems after.

What makes you imply that those who grind on Leetcode then forget?

I can confidently say that none of my technical interview practice has ever transferred to a useful skill in the job role, but I can confidently say it has helped me land jobs. Not using it definitely makes me forget it.

You don't forget, but you do fall out of practice doing what are essentially competitive programming exercises.

When you're in a situation in front of a whiteboard with 15-45 minutes to craft some algorithm it makes a ton of difference whether or you've been practicing.

Because these skills are use-it-or-lose-it. They will be LRU'd out by the skills you will actually use every day. I guarantee at your next job you will not have to roll your own red-black tree from scratch...

> Of course it does, it's not a gender problem, who said it was?

Many people have said this.

I've always been interviewed by multiple people - I think 12 was the high, and 4 was the low. When I do end up getting the job, I always wonder... which of them said yes and which said no?

12 people! What were they interviewing you for, a role as the code messiah?

This was spread out over three days: one phone screen and two separate days of in-person interviews. It wasn't even for a really amazing job (we're not talking FAANG here). You know - because we're in such high demand?

Truly the code messiah. He rose after three days of interviews.

That does seem unusual. Who does 2 days of onsite interviews for an engineering job?

After all that, they didn't offer me a job - I figured bringing me back for a second round of interviewing must be a formality but nope, they never called back.

The also do bar raiser rounds on top of those 12 rounds.

That's not that unusual. 15 minute recruiter screen -> 1 hour tech screen -> 5 hour onsite (5 parts w/ 2 devs each) can do it.

I've probably exceeded 20 before if you count lunch and meeting VPs/directors toward the end.

I still don't get it tbh. Our interview process is 3-4 1 hour talks at most - introduction / initial interview (phone or face to face) with 1-2 people (manager / HR), technical interview / code review (2 people), then a final chat with the CEO and finally a contract offer / negotiation. People can have a signed contract within a day. And we are actually very demanding of our people.

>> And we are actually very demanding of our people.

That explains it. If the job is hard, people just faffling around don't even bother to apply.

You need 12 adversarial grilling rounds of interviews so that you can hire what you claim to be the best and put them to build JSONs.

That sounds close to a dozen people involved in the process if teams double up on the 1 hour sessions.

What your describing (pushing them the more they know) is adaptive testing. Adaptive testing is great because it allows you test a wider range of skill to a higher accuracy. It's horrible for interviewing because it's very open to bias and you get inconsistent results. Unless you're doing a computerized multiple-choice test given to thousands of people, you probably shouldn't use adaptive testing.

Reply on news.ycombinator.com