Note: This depends on hiring style. Some companies have a bar and everyone who meets that bar is hired. Other companies are hiring for a position that gets one person...

These results may be "curved".

If you give the same easy test with no pressure and everyone is sufficiently good, then it was a waste of everyone's time.

In such an approach, you do something that is sufficiently hard that you're able to identify the people who are the better performing ones along with a reasonable idea of how well they work under pressure (because that will happen at some point).

Is it testing the right thing? Probably not. However, we still haven't found "the right thing" that is able to scale, respectful of the time of the candidate, not too intensive on the interviewer, and minimizes biases in hiring (same set of questions to all candidates, same grading scale).

> Is it testing the right thing? Probably not. However, we still haven't found "the right thing" that is able to scale, respectful of the time of the candidate, not too intensive on the interviewer, and minimizes biases in hiring (same set of questions to all candidates, same grading scale).

Maybe because the concept of a test is flawed? It seems to me like every company that uses a whiteboard test gives out the same tests anyways.

What are they really trying to test? If someone understands common algorithms and data structures? I dont see why that couldnt be determined by a simple interview with an engineer. A test might be useful for a HR recruiter who doesnt have a computer science background, but those tests are usually conducted by other engineers anyways.

The whiteboard thing just seems to me like an arbitrary hazing ritual.

>implies that somebody can produce quality results in 50 minutes…

This is more a random fluke as much as sampling bias. Akin to someone either getting lucky or someone who is coming from a very similar environment. This isn’t a measurement that guarantees someone is competent, rather it’s a guarantee that someone is coming from just as incompetent an environment as you have.

You know the company is on a good trajectory when the new hires are a little less incompetent than the past hires.

Or they just hire the guy they knew through their network. A lot of "how to beat the tech interview process" advice is about how to break down the door with a battering ram when you don't have the key.

Very few people hired at big tech enter through their network without a technical screen.

Speaking only of ICs:

From principal engineers down, there's a technical screening. Distinguished engineers? Maybe. But then you're speaking about Guido van Rossum, James Gosling types.

This is basically the advice I give everyone I know when it comes to things like this, "If you're going through the front door, you've already lost."

My experience has been that this applies to recruiting, hiring, BD, sales, fundraising, etc. Anywhere that there's a systemic principal-agent problem.

> If you're going through the front door, you've already lost.

I've heard that advice in various forms but I must say (anecdote incoming) - in my own 30 year career, the best jobs I've ever had were the random ones that I just applied for and the worst ones were the ones that somebody I knew from before recruited me in from outside.

How many jobs have you had? You talked about the best and the worst. Can you comment on the relationship between (y-axis) quality of job and (x-axis) strength of personal connection before you joined. If you plotted a scatterplot, what relationship would you find?

I've changed jobs ten times since 1992. As it works out, exactly half of them I was recruited into by an internal reference and the other half I found on a job board like Monster or LinkedIn. By far, the two best were relatively large organizations that I applied to cold, and the two absolute worst were actually the ones I was most actively recruited into by former co-workers. I suspect that the problem in both of the latter cases was that they worked so hard to bring me in that they ended up overselling me - I'm good, but I'm also a human who has two kids to look after. On the other hand, if I'm "just" the best candidate from the pile of resumes, expectations seem to be reasonable.

hmm, I guess I can say a reasonably good one and the worst one ever that almost made me quit programming came about this way. The same guy recruited me in both places.

This is fairly true. The easiest side-door if anyone is wondering is a recruiting firm. At least with those there is already a somewhat simple contract in place that helps get rid of poor people.

I am wondering if the entire interview process should just be pay them for 2 weeks and see if they stick.

I was a hiring manager at Facebook, and even when there was someone we specifically wanted to hire, they still had to do the same coding interviews that a new college grad has to do.

I have known of many positions that stayed open for years until the position or company disappeared.

I have done that in the past, and I've failed miserably in the past as well.

Am I on the top of my game right then and not stressed about anything and the questions out of the random million possible questions that could come up are ones that I just happen to have had on my mind recently? Then I ace the interview.

The last time that happened it was in a big international consulting company, they then put me through 5 other interviews afterwards, each one telling me I would hear something back in the next couple days.

The last one I got told couple more days, they didn't call me back for nearly two weeks. Then they call: Hey Bryan, we would complete the offer process now hope you're still interested?

Uhm unfortunately I got an offer last week for a 6 month consulting project, sorry but I had to take it because they said they couldn't wait to see how this turned out.

on edit: clarification.

Or their standards drop enough to hire someone

Or they have have a really high hackerrank score and know the answer already.