http://elsmar.com/jpg/Google_Job_Puzzles.jpg (http://money.cnn.com/2007/08/29/technology/brain_teasers.biz2/)

I'm not sure this is the best forum for this, but...

Want a job at Google? Try these brainteasers (http://money.cnn.com/2007/08/29/technology/brain_teasers.biz2/) first. Google, Microsoft, and eBay are looking for engineers who can think on their feet. Here's how they find them.

I've moved this to "discussions" since it is not a job offer per se.

Certainly puts a new slant on a "stress interview" if you only have 60 seconds to avoid being a bug in a blender.

To make a discussion of this article, let's have some comment on any or all of these points:

requiring a candidate to prove his ability to think on his feet and out of the box with puzzles
Contacting folks NOT on a candidate's reference list to get a full picture
running a candidate through multiple interviews with each interviewer having the power to "black ball" the candidateI have some very strong views on these points, but I don't want to "taint" the thread just yet. Let's see some other comments.

At it's worst, these psychological and behavioral test are better than nothing as a pre-employment screening devices. The power or relevance of these test really depends on the user - what are they trying to achieve or find out.

Generally, I equate the ability to think on ones feet to common sense. Common sense plus a strong academic foundation should get one somewhere! I am not sure if there's a correct answer for these puzzles but I think a logical answer vis-a-vis the situation should be a good guage of the candidates ability to get out or survive a bad situation.

I think contacting people not on the referee's list makes one look unethical. There are alternatives.

Running candidates through multiple interviews are more relevant for sales or communication type of jobs. I think having some good man and bad man in the interviewing team makes the assessment more interesting. The candidate has to know how to handle a losing situation (like a rejection) nicely also.

Good comments Harry. :agree:

I harbor a special dislike for psychological tests. Still, I can understand why an employer would want to use them. My question for Google would be (if they cared): "In what job performance aspect does the applicant's ability to play with these toys accurately predict success?"

I get the problem solving objective. However, understanding learning styles and intelligences as I do, I know that problem solving abilities can be exhibited on paper or whiteboard--visually and not kinetically.

I don't know if Google uses more than one kind of test and who has designed it; who interprets the results and how it's done. Soo, while admitting I don't have the complete story I'll leave my remark with my impression: :rolleyes:

Q: How many golf balls can fit inside a school bus?
A: Not enough information:

http://elsmar.com/jpg/schoolbus3qj7.jpg

http://elsmar.com/jpg/thomasschoolbusco6.jpg

Q: You're shrunk and trapped in a blender that will turn on in 60 seconds. What do you do?
A: Sprout wings and fly out, which is just as likely to happen as being shrunk and put in a blender.

Q: How much would you charge to wash all of the windows in Seattle?
A: One million dollars:

http://elsmar.com/jpg/mediumdrevil1jl1.jpg

Excellent examples of poor "operational definitions."

These sorts of problems (particularly the ones like ping pong balls on a bus or washing windows in Seattle or how much water flows out of the mouth of the Mississippi river) are often called "Fermi problems" in honor of a 20th century nuclear physicists who used to spring them on grad students during the final oral exam before being awarded a PhD (talk about high-pressure!).

They show:
a degree of practical knowledge (what are the dimensions of a school bus?)
a degree of mathematical skill (how do you calculate the volume of a school bus)
an ability to deal with uncertainty & approximations
an ability to think on your feet under pressureI think all of these skills are applicable to most higher level jobs. The answer is not really the goal, but to see the thought process.

For example
A bus is tall enough for an adult to walk in, so it is ~6' tall
A good-sized bus seats 60 students in 10 rows, and each row would have to be about 3' from the previous, so 10 x 3 = 30' long inside. (3' might be a little long, but there is also space for the driver & door)
a 3' wide bench would seat three kids (or two adults) and another 1.5' for the aisle = 7.5'
A pingpong ball is ~1" in diameter, so each one will have a volume of about 1 in^3
The volume of the bus is ~70in x 360in x 90in = ~ 2,400,000 cubic inches = 2,400,000 ping pong balls.
There is will be some space taken up my the seats, but the spheres call also pack a little more densely than cubes, so those factors will approximately cancelThat is off the top of my head. I could also use knowledge about the typical size of semi trailers to estimate the dimensions of the bus. I just googled the size of a ping pong ball and found it is a little bigger than I estimated (1.38"), which would cut my answer by more than half, to about 1,000,000.

This is still twice what the article estimates. I would challenge them to defend their values. As I mentioned earlier - it is the thought process (and general knowledge) that are important - not the exact number.


Tim F

Enrico Fermi: Enrico_Fermi

These sorts of problems (particularly the ones like ping pong balls on a bus or washing windows in Seattle or how much water flows out of the mouth of the Mississippi river) are often called "Fermi problems" in honor of a 20th century nuclear physicists who used to spring them on grad students during the final oral exam before being awarded a PhD (talk about high-pressure!).

They show:
a degree of practical knowledge (what are the dimensions of a school bus?)
a degree of mathematical skill (how do you calculate the volume of a school bus)
an ability to deal with uncertainty & approximations
an ability to think on your feet under pressureI think all of these skills are applicable to most higher level jobs. The answer is not really the goal, but to see the thought process.

For example
A bus is tall enough for an adult to walk in, so it is ~6' tall
A good-sized bus seats 60 students in 10 rows, and each row would have to be about 3' from the previous, so 10 x 3 = 30' long inside. (3' might be a little long, but there is also space for the driver & door)
a 3' wide bench would seat three kids (or two adults) and another 1.5' for the aisle = 7.5'
A pingpong ball is ~1" in diameter, so each one will have a volume of about 1 in^3
The volume of the bus is ~70in x 360in x 90in = ~ 2,400,000 cubic inches = 2,400,000 ping pong balls.
There is will be some space taken up my the seats, but the spheres call also pack a little more densely than cubes, so those factors will approximately cancelThat is off the top of my head. I could also use knowledge about the typical size of semi trailers to estimate the dimensions of the bus. I just googled the size of a ping pong ball and found it is a little bigger than I estimated (1.38"), which would cut my answer by more than half, to about 1,000,000.

This is still twice what the article estimates. I would challenge them to defend their values. As I mentioned earlier - it is the thought process (and general knowledge) that are important - not the exact number.


Tim FOf course, there is no mention of whether there are seats left in the bus nor even whether we are stuffing it while there are kids (how about adults?) This reminds me of the item a few days back about Niels Bohr estimating the height of the building with a barometer. Are they interested in the answer per se or in the cleverness in which you attack the problem? Probably a big part of the cleverness is how well you defend the "posits" or assumptions as you set up a "ball park" solution to the problem (How many major league baseballs to fill Tropicana Field?) With or without the seats? include the concession area and locker rooms or only the playing field and stands? etc.etc. etc.