that graphic does actually kickass.

Let me guess: That's the collective wealth the entire dealbreaker staff brought to the office today.

Let me guess: That's the collective wealth the entire dealbreaker staff brought to the office today.

Also, do semis, boats, lawnmovers etc use gas, or do they just run on puppies and happy thoughts?

Only idiots continue to ask interview questions such as "How many trees are there in Manhattan." This a futile line of questioning because any intelligent candidate will recognize this as a type of Fermi problem, named after the physicist Enrico Fermi. The classic Fermi problem, posed by Fermi himself, is "How many piano tuners are there in Chicago." Now, once you have studied Fermi problems in general, it is trivial to answer a Fermi question to the satisfaction of the interviewer. The ability to "answer" Fermi questions does not say anything about how well a person will perform at a given job, and only idiots make hiring decisions based on this type of stuff. Most of this idiots asking these questions have probably never never even heard of Fermi. They might as well be making hiring decisions by asking "If you could be any animal what would you be?"

I find your mathematical abilities to be incredible! Would you mind calculating how many books there are Hamilton College library? ps i love you.

This is crap. All you're doing is saying that Livingstone has 3000 people per gas station, and if that's true of the country as a whole then since the population is 300 million there must be 100,000 gas stations overall. All of this bullshit about number of cars per person is utterly irrelevant. If I was an investment banker grading you, I'd give you an F.

Ah, yes. "If you were an investment banker". From what I understand you're a blogger for Portfolio, so we already know you have questionable judgment, but I'm curious about your finance background. How much banking and finance experience do you have exactly?

No one uses these questions anymore. These questios are completely useless unless someone is so unprepared that the've never heard it before and if they're that unprepared they're screwed anyway. Seriously, this is the type of thing we used to ask each other freshman year of college when we first heard about "tough banking interviews".

They don't use these questions for banking interviews. Only consulting, hence "only idiots continue to ask interview questions such as..."

"Namaste"? Anyone who employs a Yogic greeting at the start of her posting must be a piker. "Because we've got a lot of time on our hands." Indeed.

Ah yes, my banking and finance experience. Is exactly the same as my formal training in economics. Which is zero. You got a problem with that?

Or you can just spend 5 minutes in Google and get the correct answer. Apparently the only people using Fermian logic in the modern information age are I-bankers in coming up sales revenue forecasts that are always several orders of magnitude too high.
http://sbdcnet.org/industry/gas_stations.pdf

"Only idiots continue to ask interview questions such as "How many trees are there in Manhattan.""

"No one uses these questions anymore. These questions are completely useless unless someone is so unprepared that the've never heard it before and if they're that unprepared they're screwed anyway."

Questions like that are standard in management consulting interviews.

Felix--you just can't comprehend the complexity of that question unless you are a real live investment banker. Ooh my nipples get hard just saying those two words together. Investment bankerr. Man, they really are God's gift to us. Ensuring our markets stay "efficient" all while asking nothing more than the GDP of a small nation for their tireless work.

felix the dip wad: the whole deal is an estimate, so pack sand, buddy.

felix: she didn't do anything wrong. she might've added some superfluous steps but the calculation and the answer (which is really just an estimate, all Fermis are) are corret, too.

Sorry, my bad. I should have realized that dividing cars by cars is like Dana "The Blessed" Vachon swapping the dollar into itself: something without any visible utility to the naked eye, but something for which any self-respecting investment banker should be able to charge a coupla hundred grand for, in any event.

Iceman: Management consultants do ask questions like that, but i was referring to "real" jobs. Also if you go to a pre interview recruiting session the management consultants will tell you how to answer those questions... Bain uses the same fucking cases and brain teasers every year, as I recall.

How many gas stations there are in the US?

What, do they not ask why manholes are round anymore? oh yes, that's a fermi question too..

I have good smart-alek answers for all of these, but they didn't ask me any of this when i interviewed as Microsoft, which is what the URL below identifies them as, so all the time thinking up those smart alek answers went to waste. I'm really not into hard work or long hours, so I wouldn't have wanted that job anyways.

http://64.233.167.104/search?q=cache:1WF40l-DmEUJ:www.sellsbrothers.com/fun/msiview/question.htm+interview+questions+manholes+round&hl=en&ct=clnk&cd=1&gl=us

oh yeah, there are actually 3 cars per 4 people in the US, you danged hippie.

:)

Felix, you missed the point (as usual). Bess did a good job with her estimate. The whole point of a Fermi estimate is that you break it down into variables that MAY BE WRONG but can be refined as needed. For example, Bess seems to have randomly gussed that there is 1 car per 3 people in the US, and she used her knowledge about the number of people and gas stations in her town of Livingston, NJ. Maybe the 1/3 guess is wrong, but the POINT is that with some relatively inexpensive statistical surveys, we could get a pretty good estimate of how many cars per person America has (you could just cold call housholds and ask about their family size and number of cars). This sure as hell beats trying to count every gas station in America, one by one.

Dear god. Did I EVER accuse Bess of being WRONG about the number of cars per people? No, I believe that was someone called "beanspants1". My point was that you could "refine" her 1-car-per-three-people estimate to a point where there were 100 cars per person, and you'd still get the same result for the number of gas stations in the US. You might as well throw in an estimate for the number of ficus plants per household as well, for all the difference it would make to the final result.

Felix, try to understand.

Number of gas stations = (number of persons) * (number of cars per person) * (number of gas stations per car)

If you change any of these three variables then you will get a different result. What you are trying to is that Bess could have built a different model with fewer variables:

Number of gas stations = (number of persons) * (number of gas stations per person)

The problem here is that "number of gas stations per person" is not as relevant as "number of gas stations per car." To illustrate, we could also model this problem as follows:

Number of gas stations = (number of pigeons) * (number of gas stations per pigeon).

Now if you get these two variables measured correctly, then you will get the correct number of gas stations. But since pigeons are irrelevant to gas stations, the model will have no predictive power as the numbers of pigeons and gas stations change over time.

I hope this makes it clear why number of cars per person is an important variable in any model of the number of gas stations a country has.

I think we're getting somewhere here!

But the thing is, Bess didn't actually model what you say she modeled. In fact, she modeled something more like this:

Number of gas stations = (number of persons) * (number of gas stations per person) * (number of cars per car).

If she estimated (number of cars per person) independently from how she estimated (number of gas stations per car), then what you're saying would make sense. But she didn't. She derived (number of gas stations per car) from her estimate of (number of cars per person).

I hear you, Felix. Still, not so egregious a crime, Bess' inefficiency--or casuistry, depending on how you want to spin it. I give you both 5 demerits for caring far too much about such a silly thing. Now please, will you two make up so I can get back to enjoying both of your sites without any attendant guilt?

I was asked how many gas stations in Manhattan, my response was not enough!
The interviewer agreed and said the correct answer was 50.