[imauld]

Because they all have jobs. And if they are looking for a new one they can pick where they want to go because by virtue of being good engineers it's easy for them to find jobs since everyone is looking for good engineers.

And if they aren't looking, why not? My best guess (aside from being happy where they are) is that the engineer interview process is broken. People don't want to go sit in a room and grilled on CS 101 level questions for 4 or 5 hours. It's stressful, it's a poor predictor of performance, and it's most likely not what they will be working on. So they avoid it.

Just my opinion though.

[dcminter]

I like this Spolsky article that - as well as supporting your point - suggests that there's inevitably a lot of dross in the remaining pool of candidates. Anecdotally this chimes with my experiences from both sides of the interview table! https://www.joelonsoftware.com/2005/01/27/news-58/

[neverminder]

The "CS 101" is bullshit, completely agree. Curious thing though, I think this is more of a thing in US, because in my many interviews I never got pestered with that over here in UK (London), although I should note that I never applied to one of the big 4.

[imauld]

I believe most if not all of the big 4 do algo heavy interviews. So it's my opinion that people know this and then look at their success and say "That's how we need to interview too." Ignoring the fact that probably has little to do with their success.

At a previous company we did the classic algo heavy, big O, technical questions and we suffered for it.

[pacala]

From the other side of the interview table, how do you tell apart engineers from bulls__t masters, who take credit for their team's work without ever having a positive engineering contribution?

[sidlls]

Have them talk about the problems they solved at their current and previous employers, and get to some technical depth while doing so.

This doesn't mean asking him to implement some CS algorithm from scratch on the board or doing a Big-O analysis.

It means asking him about the specific methods and how/why they worked and going into some depth. His former employers' "company secrets" aren't a good excuse for hedging here, either: with few exceptions he should be able to use a contrived, but representative system, data, etc. to explain if he truly contributed to the solution.

[Waterluvian]

Trust people's honesty. Fire frauds fast.