(This is the continuation of an article about the use and abuse of brainteasers in job interviews.)
Should an interviewer ever pose a brainteaser for a candidate? It
will probably surprise you at this point that my answer is
Yes. I see two valid uses for these questions:
shinemore on this less subject-matter-specific problem. When I do this, I'm definitely trying to
filter ina candidate who might otherwise be overlooked, not to
correct, I will let them run with it (although after a bit I might ask why they are headed that way and what they are after). I might learn something. At the very least, I will learn more about how the person thinks than I would by guiding him/her to the
correctanswer as I see it.
Actually, when interviewing an experienced candidate, I almost never
use a brainteaser
as such; more commonly I pose a hypothetical situation such as
trying to clarify a
vague requirement brought by a client. I try to frame this as
and to preface it with something like,
OK, a client/ product manager
has come to you and asked for this and you are trying to work out the
details. Keep in mind that they may not really have a clear idea of what
they really need and they may actually be asking for something that desn't
make sense. Then I try to give an initial statement of a
that is not insane, but also
not entirely well thought through, pretty much what I typically see from
a marketing department or a client. When I can, I try to draw the example
from an area in which the candidate claims expertise.
Along similar lines, if I pose a brainteaser in an interview, I first
say something like,
This next part of the interview is meant to try
learn something about how you think. Feel free to press me for any information you think would
be useful to solve the problem, and I'll do my best to give it.
also say explicitly to the candidate that, as
always when someone comes to you with a problem, I may not fully
understand all the ramifications myself, so you should
feel free to question my premises.
Also, if the candidate can't solve the problem, I would always tell him or her my proposed solution. Maybe my solution doesn't really work. Maybe the candidate has already thought of it and rejected it for a good reason. Maybe I'll find something out about the candidate if I give him/her a chance to poke holes in my thinking about the problem.
I am aware that some interviewers will pose a truly unsolvable problem
if they believe that
the key to a particular
job is the ability to handle enormous stress. There is
almost no chance that a person in a job interview will work out firmly
and confidently in a matter of minutes that the problem is unsolvable,
so all the interviewer will typically get to observe in this scenario is
behavior under exacerbated stress. (I have
never actually used this technique myself, and I don't believe I would be
comfortable doing so. Still, I would venture to suggest that if you
are going to deliberately ratchet up
stress in an interview, this should be done by a professional from a Human
Resources department, not by the
supervisor-to-be. The supervisor should get to play it straight and
establish a rapport with their potential new subordinate. The
interviewer had also
better know in advance what he or she plans to say if the candidate
calls the bluff and asks
Do you believe your problem has a
In any case, a brainteaser well-used in an interview is, precisely, a tool to learn something about how someone thinks. It's not simply a tool to see how clever they are. Frankly, computer professionals are usually pretty good at gauging how clever someone is, and we don't typically need to pose brainteasers to do it. Conversely, few of us are experts at learning how someone else thinks, and throwing brainteasers at them is not, in and of itself, sufficient to learn that. Learning something relevant about a job candidate from watching him or her try to solve a brainteaser requires something that goes totally against our usual bias as technical professionals: it requires letting go of whether the person is headed toward what we believe to be the right answer and instead observing whether they are headed along a process that will often be useful in our organization.This is particularly tricky, because often the most useful person for us to find would be one whose intellectual processes do not simply duplicate our own, so our insight into those processes will probably be limited.
Ha! So you think you can deal with this as a job candidate?
Sorry, I couldn't resist. I make no secret of the fact that I find
brainteasers very hard to deal with as
a job candidate in an interview, so maybe this is
leading the blind.
Sometimes I try to divert the subject, at least temporarily, to some of the points that are in this article, especially the point about wisdom vs. cleverness. After all, I'm looking for a good fit. Only a person who is open to that conversation is likely to be a good person for me to work for.
Ever since grappling with the Pirate, I make a point of asking for a little chill-out time before I start trying to solve this kind of problem. So far no one has said no. If they did, that would be a real indication I didn't want to work for them or with them. You might not want to be so explicit, but at least take a deep breath and give yourself a few seconds before starting.
Something I've never done, but I'm beginning to think is a good idea,
and I might do it next time this arises: if the interviewer hasn't set a clear context, you might try to do so youself. Say something like
OK, I've had the experience of people deliberately setting me an impossible problem in an interview and I've had the experience of someone being really bothered that I was probing around for hidden premises. Could you please set a context here and tell me if you are just looking for a clever solution, or if you want me to approach this the way I would if a client came to me with a feature request, or what?
Last, but not least (and a lesson I need to learn properly myself),
don't just skip past a naive but wrong solution without mentioning it explicitly. Say,
such-and-such looks almost like a solution, but I see a flaw in it. Maybe this is the solution the interviewer has in mind, but they didn't see the flaw, maybe it just gives them some clue to what wheels are turning in your head, but I don't see how it can hurt.
In any case, if someone posed me a problem I already knew, I would
stick to a policy of don't lie, don't fake being someone you aren't, etc.
Bad jobs are easy to find, and I believe that the only way to find
a good one is to be real in the interview. However, a more cynical friend of mine has said that while she'd be upfront with a person she was going to work directly for, if someone more peripheral tossed out a brainteaser and she knew the answer, she
...could do a very realistic rendition of figuring it out on the spot [...]
'I've heard of this kind of problem before, I think
it's probably like xyz, let's see if it is, blah, blah, blah. Oh, it's
like with children, there's always one that you can't leave alone, so
you bring it back and forth until you can keep it safe from itself and
each other. Blah, blah, blah. Would that work, no. How about this.
Oh, that's the answer?' I wouldn't whip it out all at once, because [...]
he'd get tossing them out one
after another and I'd want the interviewer to run out of
Of course, next time this arises, I now can say,
should ask me that, I wrote an article you might want to
Copyleft: With appropriate notification and appropriate credit, non-commercial reproduction is welcome: contact me if you have any desire to reproduce these materials in whole or in part.
My e-mail address is email@example.com. Normally, I check this at least every 48 hours, more often during the working week.