Anybody who gave an answer of 5 or above was an immediate no-hire :-) Because nobody is more than a 4 out of 10 on knowledge of C++.
Anybody who gave an answer of 5 or above was an immediate no-hire :-) Because nobody is more than a 4 out of 10 on knowledge of C++.
You'd definitely hire programmers with the same sense of humor I suppose.
I do agree that asking to rate one's knowledge of C++ is an obvious trap question. Any reasonably seasoned person would probably laugh and respond with another question, "... which part?" But yeah they'd have to share your sense of humor.
There are many ways to structure interviewers to hire clones of yourself.
Answer: We hired interns the same way. Is that really so bad though?
How would you rate yourself on a scale of 1 to 10 on C++: 1 being you know C, and 10 being you are Bjarne Stroustrup or Bruce Eckle. Personally, I started out as a 4 back in 1996, and have monotonically fallen ever since.
And not only that -- it betrays the mathematical (and everyday practical) ignorance of the people who apparently think of themselves as wily and clever for asking it.
It doesn't matter how weird and fuggly that particular language is. The only obvious, intuitive and reasonable way to approach the "How do you rate yourself?" question is always in relative terms, compared to the pool of candidates applying for the job (or perhaps compared to other engineers at the company).
Which either way means a fricken bell curve, obviously and by definition. For which 5 is the median, and the only acceptable answers are "7,8,9".
But instead someone on your time decided they had to be cute and turn it into a "Guess the punchline to the snarky joke question I came up with this morning" question.
This is Intel's idea of culture, apparently.
On the other hand, if the scale is from 1 to 5, it implies a bell curve with a low standard deviation. Especially if the scale is measured in stars. So if you took an Uber to the interview and nothing unusual happened, the driver should obviously get a three-star rating.
At one point, I would have interpreted a scale from 1 to 10 in terms of an RPG I was playing at the time. 5 meant a beginner who had been focusing on that thing, 6-7 was basic competence, and 9-10 was an expert, but not a particularly unusual one.
Hazing is an initiation rite. This is a barrier to entry.
They turned away anyone who wasn't deemed humble enough.
They were assholes but that's not hazing.
The practice of playing tricks on somebody, or giving them very unpleasant things to do, sometimes as a condition for entering a fraternity or sorority.
So it can also be a "barrier to entry". The fact that Intel isn't technically a fraternity or sorority but is basically acting like one (in regard to this team's interview process) is but a small side detail here.
We also have, from Wiktionary:
To oppress or harass by forcing to do hard and unnecessary work.
Like requiring them to discern and accommodate the interviewer's oblique sense of humor as a condition for entry, for example.
And we didn't ask dick C++ questions either. After somebody has told you they are only a 3 or a 4 out of 10 in C++, you don't have to ding them if they don't know your favorite language feature, or the latest cool thing.
I've been dinged in several interviews because I didn't happen to know the 20% subset of C++ which that particular group used. Every group only uses about 20% of C++, and every group has a different subset. I think it's highly unfair to expect anybody to be a 10 out of 10 in C++.
I would much rather be in an interview where the interview valued my ability to accurately determine my limitations, than if somebody asked me some esoteric, trick c++ question, and then concluded that I didn't know C++.
Quite plainly there's no rational reason for someone to suspect that in this particular case of a standard boilerplate HR question ("How do you rate yourself?") that they should expect your secret trapdoor answer ("4 or below") should apply rather than the usual, automatic "7-9 is the only right answer").
I'll take your word that it wasn't intended as a joke. But it was intentionally oblique and slippery. All that you accomplish with a "filter" like this is to flush a lot of otherwise potentially perfectly decent candidates. And if they were to find out after the fact that this was where they "failed" your interview (after taking a whole day off work, driving down there, and going through all the other interview rituals) -- they'd be perfectly justified in applying whatever 4-letter words came to their mind in regard to that company and its vaunted "culture".
But we didn't ask dick C++ questions. Like list all 200 ways you can initialize a variable in C++. And, once somebody has told you they are 4/10 in C++, you can't ding them if they don't know your favorite C++ feature, or if they haven't memorized the next version of the standard.
IMHO, interview questions should not be about what the candidate knows. At all. When I went to Intel, I knew nothing about chip design or EDA software.
Which was OK, because chip design methodologies were all obsoleted by Moores law, and the techniques needed to fix it hadn't even been invented yet.
What is important is what you can learn and how fast can you learn it.