5 ms·

I agree with the need to assess one's ability in Fermi-style problem solving, but I find fault with using non-sensical information for the topic. Seriously, gol

I agree with the need to assess one's ability in Fermi-style problem solving, but I find fault with using non-sensical information for the topic. Seriously, golf balls on a bus? Why not ask about a relevant problem to be solved? For me, an interview is a two-way street, and a company can most certainly put itself out of consideration with these types of questions. The biggest issue I find with these esoteric-problem-analagous-to-something-relevant types of interview questions is that they usually send a negative signal about the company. I immediately think of gamesmanship, trick questions, watching candidates squirm, or about a dozen other things that I find have nothing to do with the entire reason I might consider joining a company. While the interviewer's intentions sound very noble, I find the means to the end more risky than useful.

EwanG 14y ago
I've found that usually once I get asked one of these, that's the end of the interview. Why? Because I epitomize "think outside the box". You see, there are a number of different bus styles - are we talking a BlueBird full length or their Handicapped Plus series - ways to fill the bus - after all if I am allowed to tamp and pack I'll get a lot more of them in and we need to account for that - and of course golf balls are not all the same size - though they are close but when you're talking literally thousands of them the differences add up. Another favorite of mine is the "scored" bar of gold to pay someone's salary. First off, real gold is kind of soft. I guarantee that if I have the tools to build something I need to pay for, I can cut the bar of gold. Even if I didn't, gold can be dissolved by a number of different acids. You see what I mean?
- yummyfajitas 14y ago
  Hearing your clever answers is the entire point of asking the question.
- Luc 14y ago
  > I've found that usually once I get asked one of these, that's the end of the interview. Why? Because I epitomize "think outside the box". You see, there are a number of different bus styles [...] That's IMHO the wrong approach. Everyone who's thought two minutes about this problem realizes these issues. None of these are a show-stopper to making a reasonable estimate. It would be completely acceptable for you to say 'Okay, I'll assume a sort of schoolbus like in the Simpsons, let's say 12 rows of seats, we'll need about 2m of vertical space plus some extra for when the driver goes over a speedbump so tall people don't hit their head', etc. The point is whether you can make a quick estimate, so that when in a real work situation you are asked to do something tricky, you can work out whether it's feasible or not.
  - EwanG 14y ago
    Really? Because the times I've tried to "seriously" answer such questions it became apparent the person asking was looking for "THE" answer. In fact I was halfway into the effect of having seat belts included when the interviewer started gently feeding me "the" answer they were looking for. Why? Because they felt that there really was a "right" answer that folks could agree upon. Not surprisingly, said company folded two years later as I gather their client didn't agree...
    - Luc 14y ago
      Ah, in that case, you're quite right to walk away! It's just meant to see if someone can 'put their arms around a problem', a rough way of weeding out people who can't make a move until they have every answer, a blunt tool. Sounds like you used it appropriately as a blunt tool to weed out the company :)
  - v0cab 14y ago
    How do you know that it's OK to make assumptions like that? How do you know that a quick estimate is acceptable? Is this level of accuracy acceptable? How do you know? I can't get the point of this. The interviewer wants to know how many golfballs fit on the bus, but he doesn't even know what kind of bus? He hasn't decided what things are on the bus, and how many of them, and whether there is a driver, and what kind of track he's going to be driving the bus on? And he probably doesn't know why the golfballs are being put on a bus at all? And he doesn't care as to what kind of accuracy this estimate is being made, as long as it just seems I've gone through some kind of motions? This is ludicrous. It doesn't resemble any real-world task.
    - Luc 14y ago
      Yes it does! 'We have an idea for a game where you walk around a real-looking representation of London, with realistic traffic and people on the side walk. We'd like to have this on the iPhone, PS Vita and Nintendo 3DS. We want you to write the engine for it.' So, what do you do, start developing and hope for the best? The fact that it's balls in a bus and not polygons on a screen is not a negative. It makes clear that the point is not to test your knowledge of the RAM of each platform, the point is to see if you can reason while facing unknowns. > I can't get the point of this. The interviewer wants to know how many golfballs fit on the bus [...] No, no, no. The point is to see if you give up (a large proportion do), or if you make bad assumptions (e.g. hand-wavy 'Lets say there's 1000 golf balls in a cubic meter').
- sanderjd 14y ago
  Yes. And what's wrong with any of those things? It's a give and take - the interviewer will say "Ok, let's stipulate it is full length and you can't tamp them" and "Oh, that's interesting, assuming some sort of variance in size, estimate a range instead of a number", etc. That's one of the reasons made-up questions are better than "real" questions, the imaginative tit-for-tat is a good thing, and is harder with "real" questions.
  - v0cab 14y ago
    Is the bus in space? Is it a toy bus? Is it a bus for aliens? Can I assume the bus is perfectly spherical? Is the bus a ghost? Do I have a ray gun that can shrink and enlarge things? Are the golf balls made out of water? Do kids keep coming along and stealing golf balls from the bus until it is perfectly full? etc.
    - sanderjd 14y ago
      Interesting, how would that affect the answer? No, it is a real bus, like the ones they use for humans in public transport. Interesting, if you're curious because aliens might be a different size, no, it is a bus for human-sized things. No. Nope. No, they are normal golf balls. Maybe, after we estimate the number, perhaps we can estimate the time-to-fill based on various rates of child-miscreant golf-ball theft. etc.
      - v0cab 14y ago
        > Maybe, after we estimate the number, perhaps we can estimate the time-to-fill based on various rates of child-miscreant golf-ball theft. This comes across as a kind of punishment for asking stupid questions.
        sanderjd 14y ago
        Is that a good thing or a bad thing? If you're really trying to show your outside-the-box thinking skills, you're gonna get as good as you give and it could be fun. If you're just being a jerk and trying to frustrate me then yeah, you should just politely stop the interview and go home.
- masklinn 14y ago
  > Even if I didn't, gold can be dissolved by a number of different acids. Which acids other than aqua regia? Mercury is not an acid, and alkaline solutions (of potassium or cyanide) are the opposite of acids.
planetguy 14y ago
Seriously, golf balls on a bus? Why not ask about a relevant problem to be solved? Asking about an irrelevant problem makes it clear that this question is about your logical process and not about your knowledge related to this job. It's a way of isolating variables. The actual golf ball/bus question is a bit too much of a cliche to actually use, though. I'd rather ask how many raccoons you could put in a battleship.
- sanderjd 14y ago
  YES. Getting a sense of core thought process and problem solving skills in an interview is the hard part. Isolating variables is important.
- malandrew 14y ago
  What variables are you trying to isolate and why? And how does asking about golfballs on a bus or raccoons on a battleship effectively isolate those variables? What was the thought process used that led to that solution?
- jroseattle 14y ago
  Oh, I get the point the interviewer wants to reach -- using size estimation questions to gain an understanding of one's logic process. But as a method of achieving that goal, I've found it's a poor instrument. Suppose candidate #1 is unimaginative and cannot offer a reasonable response. Suppose candidate #2 says "well, a golf ball is such and such size, and there is empty space around each ball, and the bus is such and such size....." What, exactly, have we learned here? In my experience, nothing. Having gone through this cycle as both a candidate as well as hiring manager (not my choice to use this line of questioning), the real-world results I've found is little to no correlation between candidate responses and accurate assessment of logic process. The false positive/negative outcomes were significantly higher, and the success rate for candidates into the positions we hired them for were mixed. Basically, these questions offered us nothing in finding candidates who would be successful in the position for which we were hiring. If I were to ask a question such as this, it would be to gauge one's reaction to the question (assuming it's a stressor) and how they react. The response is nearly irrelevant. But I don't like the tone the question sets between myself and a candidate, so I just don't resort to this as a means to achieving that goal. There are better ways to get at that information.
mlavorel 14y ago
Fermi trained his students by asking them questions that were not related to physics, like the number of piano tuners in Chicago (D. Hubbard, 2007).

EwanG 14y ago

I've found that usually once I get asked one of these, that's the end of the interview. Why? Because I epitomize "think outside the box". You see, there are a number of different bus styles - are we talking a BlueBird full length or their Handicapped Plus series - ways to fill the bus - after all if I am allowed to tamp and pack I'll get a lot more of them in and we need to account for that - and of course golf balls are not all the same size - though they are close but when you're talking literally thousands of them the differences add up. Another favorite of mine is the "scored" bar of gold to pay someone's salary. First off, real gold is kind of soft. I guarantee that if I have the tools to build something I need to pay for, I can cut the bar of gold. Even if I didn't, gold can be dissolved by a number of different acids. You see what I mean?

yummyfajitas 14y ago
Hearing your clever answers is the entire point of asking the question.
Luc 14y ago
> I've found that usually once I get asked one of these, that's the end of the interview. Why? Because I epitomize "think outside the box". You see, there are a number of different bus styles [...] That's IMHO the wrong approach. Everyone who's thought two minutes about this problem realizes these issues. None of these are a show-stopper to making a reasonable estimate. It would be completely acceptable for you to say 'Okay, I'll assume a sort of schoolbus like in the Simpsons, let's say 12 rows of seats, we'll need about 2m of vertical space plus some extra for when the driver goes over a speedbump so tall people don't hit their head', etc. The point is whether you can make a quick estimate, so that when in a real work situation you are asked to do something tricky, you can work out whether it's feasible or not.
- EwanG 14y ago
  Really? Because the times I've tried to "seriously" answer such questions it became apparent the person asking was looking for "THE" answer. In fact I was halfway into the effect of having seat belts included when the interviewer started gently feeding me "the" answer they were looking for. Why? Because they felt that there really was a "right" answer that folks could agree upon. Not surprisingly, said company folded two years later as I gather their client didn't agree...
  - Luc 14y ago
    Ah, in that case, you're quite right to walk away! It's just meant to see if someone can 'put their arms around a problem', a rough way of weeding out people who can't make a move until they have every answer, a blunt tool. Sounds like you used it appropriately as a blunt tool to weed out the company :)
- v0cab 14y ago
  How do you know that it's OK to make assumptions like that? How do you know that a quick estimate is acceptable? Is this level of accuracy acceptable? How do you know? I can't get the point of this. The interviewer wants to know how many golfballs fit on the bus, but he doesn't even know what kind of bus? He hasn't decided what things are on the bus, and how many of them, and whether there is a driver, and what kind of track he's going to be driving the bus on? And he probably doesn't know why the golfballs are being put on a bus at all? And he doesn't care as to what kind of accuracy this estimate is being made, as long as it just seems I've gone through some kind of motions? This is ludicrous. It doesn't resemble any real-world task.
  - Luc 14y ago
    Yes it does! 'We have an idea for a game where you walk around a real-looking representation of London, with realistic traffic and people on the side walk. We'd like to have this on the iPhone, PS Vita and Nintendo 3DS. We want you to write the engine for it.' So, what do you do, start developing and hope for the best? The fact that it's balls in a bus and not polygons on a screen is not a negative. It makes clear that the point is not to test your knowledge of the RAM of each platform, the point is to see if you can reason while facing unknowns. > I can't get the point of this. The interviewer wants to know how many golfballs fit on the bus [...] No, no, no. The point is to see if you give up (a large proportion do), or if you make bad assumptions (e.g. hand-wavy 'Lets say there's 1000 golf balls in a cubic meter').
sanderjd 14y ago
Yes. And what's wrong with any of those things? It's a give and take - the interviewer will say "Ok, let's stipulate it is full length and you can't tamp them" and "Oh, that's interesting, assuming some sort of variance in size, estimate a range instead of a number", etc. That's one of the reasons made-up questions are better than "real" questions, the imaginative tit-for-tat is a good thing, and is harder with "real" questions.
- v0cab 14y ago
  Is the bus in space? Is it a toy bus? Is it a bus for aliens? Can I assume the bus is perfectly spherical? Is the bus a ghost? Do I have a ray gun that can shrink and enlarge things? Are the golf balls made out of water? Do kids keep coming along and stealing golf balls from the bus until it is perfectly full? etc.
  - sanderjd 14y ago
    Interesting, how would that affect the answer? No, it is a real bus, like the ones they use for humans in public transport. Interesting, if you're curious because aliens might be a different size, no, it is a bus for human-sized things. No. Nope. No, they are normal golf balls. Maybe, after we estimate the number, perhaps we can estimate the time-to-fill based on various rates of child-miscreant golf-ball theft. etc.
    - v0cab 14y ago
      > Maybe, after we estimate the number, perhaps we can estimate the time-to-fill based on various rates of child-miscreant golf-ball theft. This comes across as a kind of punishment for asking stupid questions.
      - sanderjd 14y ago
        Is that a good thing or a bad thing? If you're really trying to show your outside-the-box thinking skills, you're gonna get as good as you give and it could be fun. If you're just being a jerk and trying to frustrate me then yeah, you should just politely stop the interview and go home.
masklinn 14y ago
> Even if I didn't, gold can be dissolved by a number of different acids. Which acids other than aqua regia? Mercury is not an acid, and alkaline solutions (of potassium or cyanide) are the opposite of acids.

planetguy 14y ago

Seriously, golf balls on a bus? Why not ask about a relevant problem to be solved? Asking about an irrelevant problem makes it clear that this question is about your logical process and not about your knowledge related to this job. It's a way of isolating variables. The actual golf ball/bus question is a bit too much of a cliche to actually use, though. I'd rather ask how many raccoons you could put in a battleship.

sanderjd 14y ago
YES. Getting a sense of core thought process and problem solving skills in an interview is the hard part. Isolating variables is important.
malandrew 14y ago
What variables are you trying to isolate and why? And how does asking about golfballs on a bus or raccoons on a battleship effectively isolate those variables? What was the thought process used that led to that solution?
jroseattle 14y ago
Oh, I get the point the interviewer wants to reach -- using size estimation questions to gain an understanding of one's logic process. But as a method of achieving that goal, I've found it's a poor instrument. Suppose candidate #1 is unimaginative and cannot offer a reasonable response. Suppose candidate #2 says "well, a golf ball is such and such size, and there is empty space around each ball, and the bus is such and such size....." What, exactly, have we learned here? In my experience, nothing. Having gone through this cycle as both a candidate as well as hiring manager (not my choice to use this line of questioning), the real-world results I've found is little to no correlation between candidate responses and accurate assessment of logic process. The false positive/negative outcomes were significantly higher, and the success rate for candidates into the positions we hired them for were mixed. Basically, these questions offered us nothing in finding candidates who would be successful in the position for which we were hiring. If I were to ask a question such as this, it would be to gauge one's reaction to the question (assuming it's a stressor) and how they react. The response is nearly irrelevant. But I don't like the tone the question sets between myself and a candidate, so I just don't resort to this as a means to achieving that goal. There are better ways to get at that information.

mlavorel 14y ago

Fermi trained his students by asking them questions that were not related to physics, like the number of piano tuners in Chicago (D. Hubbard, 2007).