Why interview brainteasers are on the way out – even at Google

You don’t have to be a genius to work out why brainteasers are dying.

 

Imagine you have eight balls of equal size. Seven of them weigh exactly the same, but one is slightly heavier. How would you go about finding the heaviest one using some scales and just two weighings?

 

If you’re thinking about taking six balls and putting three on either side of the scales, congratulations – that’s half the answer. You’d then have three sets of balls and would know which group the heavier one was in (the final weighing will be enough to find it). But don’t get too excited – you’re in possession of an increasingly useless skill. 

 

Cracking such devilish brainteasers might have helped land you a job at Google or elite consulting firms like McKinsey and Bain a decade ago, but recruiters have been turning their backs on the questions faster than a Microsoft engineer could estimate the number of petrol stations in Manhattan (roughly 40, apparently). Laszlo Bock, the senior vice president of people operations at Google, told the New York Times last year that the firm now considers them a “complete waste of time” as an assessment tool. Why?

 

Interview brainteasers were designed to intuit information about how candidates approach thorny intellectual problems. So-called “structured estimation questions” – how many trains are there on the London Underground? –  are a good example. The ideal answer begins by guessing the number of routes – say, 10 to 12 lines, with two branches per line, making 25 a nice round guess for the number of distinct routes. 

 

Next, estimate the number of stations per route, average journey time per station, the round trip time for a route, and how frequently trains stop at stations, in order to get an idea of the number of trains required on each route. Consultancy Oliver Wyman’s model answer ends up with 20. Finally, multiply this by your estimate for the number of distinct routes (25) to get 500 London Underground trains. 

 

For analytical careers, the ability to break down a seemingly impossible problem into its component parts, and then work through them using a logical structure, is highly prized. 

 

It’s not clear, however, that brainteasers actually measure this. According to Block, Google’s internal analysis found that being able to solve questions that would get even the best minds scratching their heads wasn’t a good indicator of future success. In fact, the questions “don’t predict anything,” he says. “They serve primarily to make the interviewer feel smart.”

 

The problem, according to Maria Konnikova, who writes about science for The New Yorker, is one of context.  Brainteasers measure a candidate’s ability to come up with a plausible-sounding piece of reasoning on subject matter they know nothing about. This is something very few of us are likely to be faced with on a daily basis, making the tests largely redundant beyond the interview room. It’s almost a “gotcha” technique, she says, with the open-ended nature of the questions leaving an awful lot to the interviewer’s discretion.

 

Worse, a 2012 study at San Francisco State University found that interviewers weren’t often sure how to make sense of candidate’s answers, and that otherwise qualified candidates were actually put-off by puzzle-style questions.

 

---

Hone your reasoning skills

Free

 

Brainteasers may be on the way out as assessment tools, but they still make for excellent pub conversation. Logic Games, part of developer Andrea Sabbatini’s collection of time-killing, mind-stretching puzzle apps, will help to keep your reasoning skills sharp. There are 98 different games included in the app, and they come in a variety of difficulty levels (there are over 6,000 puzzle levels in total). Your progress is autosaved, and each level comes with hints in case you get really stuck.