Hacker News Reader

Comment by danielweber | Hacker News Reader

Eehhhh, maybe. You shouldn't talk bad about your current/former employer.

On the other hand, surely everyone working in any kind of sysadmin role has a story of things going sideways.

I'm still worried about the pop-psychology of "well, good candidates should have <stories>, and he didn't have a <story>, so he's not a good candidate." Just substitute other words in for "story" and you get more of a gut-feel that often doesn't translate into real world success.

View on HN

> On the other hand, surely everyone working in any kind of sysadmin role has a story of things going sideways.

I want to agree with this but there are some big caveats.

1. What if you don't expect this question and thus don't trawl through memories ahead of time? It's easy to feel put-on-the-spot and have your mind go blank.

2. What if you're really talented and worked at a place with rigorous process and thus all the stuff going bad gets caught in test or QA or whatever? Nothing really goes sideways in that situation.

3. There are a lot of situations where things go bad not because you've made a mistake but because management's made a mistake. Not wanting to talk about those situations is understandable for a couple of reasons. Not talking bad about the former employer is one, so is not wanting to get all worked up about former management screwing you in a job interview.

4. You could have been on either end of the situation described by a previous HN story: http://mikehadlow.blogspot.co.uk/2013/12/are-your-programmer... If you're on the "lazy" team you might have been a damn-fine sysadmin with no horror stories. Or on the "hard working" team with plenty of disaster stories and a questionable level of talent.

I'm not saying that having disaster stories makes you a bad sysadmin, nor am I saying that not having disaster stories makes you a good one. But I am suggesting that the correlation between disaster stories and skill level probably isn't p = 0.99 and might be more like p = 0.5 or 0.2 or maybe 0.07.

Even if you find that the correlation is 0.7 (pretty decent) are you going to be interviewing enough people such that you can be sure that you don't end up with sampling error?

If you need to hire 10 sysadmins and you interview 200 then perhaps you'll do a decent job of identifying the good candidates. But if you only need to hire one and you only interview 10 you don't have the law of large numbers on your side. http://en.wikipedia.org/wiki/Law_of_large_numbers

> 1. What if you don't expect this question and thus don't trawl through memories ahead of time? It's easy to feel put-on-the-spot and have your mind go blank.

Well now aren't we just walking about interview skills? Interviewing poorly leads to not getting job offers, and if that's why he didn't have a better example off the top of his head, it's due to his poor interview skills that he wasn't hired.

> 2. What if you're really talented and worked at a place with rigorous process and thus all the stuff going bad gets caught in test or QA or whatever? Nothing really goes sideways in that situation.

I don't think these places exist. I work for a Fortune 500 with an absolute ton of process due to the sector we work in. I haven't even been here 2 years, I'm not even in sysops, and I can think of one or two things that would have made a good interview story about something "going sideways."

1. Any good interview process should try and see through the interview skills and look more at the job candidate's actual skills that they're being hired for. A company's inability to do so isn't a referendum on the candidate but rather on management. Unless of course the company is hiring the candidate to be interviewed professionally.

2. I said a rigorous process, not a ton of process. There are good processes and there are not good processes. The fact that you work at a Fortune 500 company doesn't automatically make your processes good.

If Fortune 500 companies were so much better at executing processes than startups surely they would also be better at executing "finding new businesses" processes and thus startups would never get off the ground. The established businesses with far more financial and human resources would immediately swoop in on any opportunities.

But surprise! The world doesn't work that way. Being big doesn't mean being right.

> 1. Any good interview process should try and see through the interview skills and look more at the job candidate's actual skills that they're being hired for. A company's inability to do so isn't a referendum on the candidate but rather on management. Unless of course the company is hiring the candidate to be interviewed professionally.

This makes sense if the company is desperate to fill the role or needs tons of time. In many company's hiring is a process where false negatives (kicking someone out because of warning signs) is a better safety than a false positive (hiring someone and having them turn out to be crappy / toxic).

Hiring is hard but firing people is even harder. Some companies have the bandwidth to manage people who are underperforming and raise up their general level of performance and some don't have that luxury.

That being said, defining valid false negatives versus ridiculous bias IS HARD. There are things that are bad from a cultural fit perspective, but if you can only say hes a bad cultural fit and not give more details, if not to the candidate then at least to the rest of the hiring team, then you are not doing your job.

Good candidates should prepare for interviews. Interviews typically require that a candidates are able to reflect on previous successes and failures and provide insights so a hiring team can review their candidacy. Failure to do show shows either:

1. lack of understanding of professional working environments (ie. interview structure)

2. lack of preparation for the interview

3. lack of ability for self-reflection

4. actual lack of any experience with which to construct even a lame story.

Or, failure could show that the candidate has an understanding that how he handled "previous successes" and "previous failures" is almost certainly irrelevant to the job being applied for. Unless the new company's line of business is substantially similar to the previous company's, a candidate's experiences, success, failures, and how he handled them are at best useful as an aggregate measure of "stuff he's done." When the line of business is substantially similar, the candidate's prior experience at the similar company is only marginally more applicable, because actual differences in the internal operations figure more prominently. It doesn't indicate, necessarily, any thing at all about his personality, how he will do his job with the new company, or really anything else other than that "candidate has worked places."

Human interactions are complicated, stochastic things and I can't help but think that interview processes that include "describe a success/describe a failure" in order to "get a feel" for the person are a hackneyed attempt to distill these complicated things into overly simplistic checklists and scores.

Reply on news.ycombinator.com