# Has there ever been a real-life equivalent to the "Kobayashi Maru" test? **URL:** \ **Category:** Factual Questions\ **Created:** [April 14, 2014, 11:28pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119 "2014-04-14T23:28:51Z")\ **Posts on this page:** 20\ **Page:** 3
**Author:** ![aaelghat](https://avatars.discourse-cdn.com/v4/letter/a/c5a1d2/32.png) [@aaelghat](https://boards.straightdope.com/u/aaelghat)\ **Post date:** [April 15, 2014, 2:11pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/41 "2014-04-15T14:11:44Z")
> [@Pixel\_Dent](#): > > Twenty five years ago I interviewed at one company and after a full day of technical interviews with lots of programming problems the last question was this… > > “Design two six sided dice so that the probability of all results from 2 to 12 are equal.” After playing with the equations enough to realize it was impossible, I ended up writing a proof that it was impossible and giving them that. > > It sounds cruel but it was a great company to work for (in fact the only company I’ve ever worked for. After leaving there 10 years later I went the entrepreneur route), had very high standards, and we ended up inventing something that most of you now use every day. I’ll bite - may I ask the name of the company, and what was invented that most of us use every day? ---
**Author:** ![Chronos](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chronos/32/134_2.png) [@Chronos](https://boards.straightdope.com/u/Chronos)\ **Post date:** [April 15, 2014, 2:15pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/42 "2014-04-15T14:15:47Z")
> [@](#): > > Quoth **Kobal2** : > > We don’t really do multiple-choice exams here in France for the most part, but in the few exams I took that used the format you’d get docked more points for answering wrong than for simply not answering the question. If you really applied yourself at being terrible you could wind up with a negative final score, even. The SAT works like this, too, so turning in the paper blank won’t actually give you the lowest possible score. The penalty isn’t actually all that large, though, so guessing completely randomly will, on average, give you approximately the same score as turning it in blank (and guessing not-completely-randomly, like ruling out an obvious wrong answer on each question, will on average still be better than nothing). The ACT, on the other hand, does not have a penalty for guessing, so if you’re down to a minute left and aren’t finished, you might as well just bubble in C for all remaining answers. And there was one math test I took in high school (that didn’t count for anything but prestige, unless you scored _really_ well) where if you turned in the answer sheet completely blank, you would get 60 points out of a maximum of 150 possible… But the national average was only 58. This even after the teachers all warning their students of this fact. ---
**Author:** ![Melchior](https://avatars.discourse-cdn.com/v4/letter/m/e5b9ba/32.png) [@Melchior](https://boards.straightdope.com/u/Melchior)\ **Post date:** [April 15, 2014, 2:28pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/43 "2014-04-15T14:28:05Z")
> [@Mr.Nylock](#): > > I should have said working in retail. I was being a little silly, but that is the first thing that came to mind. Lots of times in retail a store is very understaffed and the sales people try to make sure they can help everyone, and then someone wants to take up all sorts of time who is not very serious about buying something while a whole lot of other people are getting irritated, and no matter what you need to find a way to stay smiley and not let anyone walk out the door in a huff. In a huff? In a minute and a huff? [http://youtu.be/Dsw9jYU\_rJI](http://youtu.be/Dsw9jYU_rJI) [http://www.marx-brothers.org/whyaduck/info/movies/scenes/firefly.htm](http://www.marx-brothers.org/whyaduck/info/movies/scenes/firefly.htm) ---
**Author:** ![Melchior](https://avatars.discourse-cdn.com/v4/letter/m/e5b9ba/32.png) [@Melchior](https://boards.straightdope.com/u/Melchior)\ **Post date:** [April 15, 2014, 2:33pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/44 "2014-04-15T14:33:37Z")
> [@Pixel\_Dent](#): > > Twenty five years ago I interviewed at one company and after a full day of technical interviews with lots of programming problems the last question was this… > > “Design two six sided dice so that the probability of all results from 2 to 12 are equal.” After playing with the equations enough to realize it was impossible, I ended up writing a proof that it was impossible and giving them that. Uhm, you don’t even have to think about this to know that it is impossible. ---
**Author:** ![cjepson](https://avatars.discourse-cdn.com/v4/letter/c/a9adbd/32.png) [@cjepson](https://boards.straightdope.com/u/cjepson)\ **Post date:** [April 15, 2014, 2:39pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/45 "2014-04-15T14:39:54Z")
> [@RickJay](#): > > I am sure there are examples from the field of psychology where people are subjected to similarly unfair, unwinnable simulations. Yes, there have ben a lot of experiments in which the participants are told that they’re being tested on, say, math ability or spatial reasoning, but the test is actually set up so that they can’t succeed, and what they’re really being tested on is generally their response to the stressor. [Here](http://www.academia.edu/164194/Emotional_Intelligence_Personality_and_Task-Induced_Stress) is an example using an impossible anagrams task. ---
**Author:** ![Telemark](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/telemark/32/372_2.png) [@Telemark](https://boards.straightdope.com/u/Telemark)\ **Post date:** [April 15, 2014, 2:54pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/46 "2014-04-15T14:54:22Z")
> [@Really\_Not\_All\_That\_Bright](#): > > Doesn’t it take about 10 seconds to realize that there is no way to equalize the probability of rolling a two? Are you assuming each die has numbers 1-6 on the faces? You could have no ‘1’ and multiple '2’s on one die and a ‘0’ on the other. ---
**Author:** ![msmith537](https://avatars.discourse-cdn.com/v4/letter/m/d9b06d/32.png) [@msmith537](https://boards.straightdope.com/u/msmith537)\ **Post date:** [April 15, 2014, 3:14pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/47 "2014-04-15T15:14:47Z")
In consulting, we have what is called the “case interview”. Basically it’s a question like “how many flights land at JFK in a week” or “how many manhole covers are there in Manhattan”. Now the typical SDMB response is “those questions are stupid I would never work for such a stupid company that tried to trick you with something like that blah blah blah.” The thing is, the answer isn’t important. The case interview is designed to see how you take a complex problem and break it down. We look at your reasoning and assumptions. It’s my understanding tha the Kobayashi Maru served a similar purpose. It wasn’t so much “winning” was the objective (since you can’t). It was to get a sense for how the captain led under difficult circumstances. Wasn’t there another captain test in ST:TNG where Diana kept failing because she wouldn’t send holographic Jordi into the irradiated air ducts to pull some stupid lever that will save the ship? ---
**Author:** ![Quercus](https://avatars.discourse-cdn.com/v4/letter/q/7ab992/32.png) [@Quercus](https://boards.straightdope.com/u/Quercus)\ **Post date:** [April 15, 2014, 3:26pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/48 "2014-04-15T15:26:28Z")
> [@Pixel\_Dent](#): > > Twenty five years ago I interviewed at one company and after a full day of technical interviews with lots of programming problems the last question was this… > > “Design two six sided dice so that the probability of all results from 2 to 12 are equal.” After playing with the equations enough to realize it was impossible, I ended up writing a proof that it was impossible and giving them that. But if you are allowed to use an algorithm for getting the result that’s different from “Add the result of each die together”, (and can distinguish the dice, and are allowed to re-roll if necessary or, equivalently, allowed to return a ‘no answer this time’), it’s trivial to do. It’s not like adding A to B is the only way (or even the only natural obvious) way to get a third number from A and B. ---
**Author:** ![Really\_Not\_All\_That\_Bright](https://avatars.discourse-cdn.com/v4/letter/r/e8c25b/32.png) [@Really\_Not\_All\_That\_Bright](https://boards.straightdope.com/u/Really_Not_All_That_Bright)\ **Post date:** [April 15, 2014, 3:28pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/49 "2014-04-15T15:28:51Z")
> [@Telemark](#): > > Are you assuming each die has numbers 1-6 on the faces? You could have no ‘1’ and multiple '2’s on one die and a ‘0’ on the other. … and that’s why I’m not qualified to work in engineering. 😃 ---
**Author:** ![Machine\_Elf](https://avatars.discourse-cdn.com/v4/letter/m/82dd89/32.png) [@Machine\_Elf](https://boards.straightdope.com/u/Machine_Elf)\ **Post date:** [April 15, 2014, 3:32pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/50 "2014-04-15T15:32:08Z")
Obligatory [XKCD reference,](http://what-if.xkcd.com/84/) showing the execution of a Fermi estimate. ---
**Author:** ![kanicbird](https://avatars.discourse-cdn.com/v4/letter/k/5f8ce5/32.png) [@kanicbird](https://boards.straightdope.com/u/kanicbird)\ **Post date:** [April 15, 2014, 3:40pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/51 "2014-04-15T15:40:54Z")
> [@Pixel\_Dent](#): > > Twenty five years ago I interviewed at one company and after a full day of technical interviews with lots of programming problems the last question was this… > > “Design two six sided dice so that the probability of all results from 2 to 12 are equal.” After playing with the equations enough to realize it was impossible, I ended up writing a proof that it was impossible and giving them that. Very possible and simple: Design = one red and one blue: 1 -Role them 2 - Take the number rolled on the red die, add 6 if the number on the blue die is 1,2,3 3- reroll if the answer comes out to 1 (not really required given the OP’s guidelines) Or correspondingly one die standard numbered and the other numbered 6,6,6,0,0,0 and add them together ---
**Author:** ![kanicbird](https://avatars.discourse-cdn.com/v4/letter/k/5f8ce5/32.png) [@kanicbird](https://boards.straightdope.com/u/kanicbird)\ **Post date:** [April 15, 2014, 3:57pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/52 "2014-04-15T15:57:17Z")
> [@kanicbird](#): > > Very possible and simple: > Design = one red and one blue: > 1 -Role them > 2 - Take the number rolled on the red die, add 6 if the number on the blue die is 1,2,3 > 3- reroll if the answer comes out to 1 (not really required given the OP’s guidelines) > > Or correspondingly one die standard numbered and the other numbered 6,6,6,0,0,0 and add them together Another more complex answer may be to design dice with uneven sides for instance where 1+6 have a larger probability of coming up and 4 and 3 have the lowest. Maybe 😕 but that would be a much more complex solution and I’m down with my simple one above 🙂 ---
**Author:** ![Ravenman](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/ravenman/32/2929_2.png) [@Ravenman](https://boards.straightdope.com/u/Ravenman)\ **Post date:** [April 15, 2014, 4:14pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/53 "2014-04-15T16:14:44Z")
Except that your post #50 doesn’t actually accomplish the objective of having the two dice roll two 9% of the time; three 9% of the time, four 9% of the time, etc. Under your numbered scenario, it would be impossible to come up with the result of two, for example (because rolling 1 on the red die and 1 on the blue die would result in eight being the lowest possible total). ---
**Author:** ![twhitt](https://avatars.discourse-cdn.com/v4/letter/t/87869e/32.png) [@twhitt](https://boards.straightdope.com/u/twhitt)\ **Post date:** [April 15, 2014, 4:19pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/54 "2014-04-15T16:19:16Z")
> [@kanicbird](#): > > Very possible and simple: > Design = one red and one blue: > 1 -Role them > 2 - Take the number rolled on the red die, add 6 if the number on the blue die is 1,2,3 > 3- reroll if the answer comes out to 1 (not really required given the OP’s guidelines) > > Or correspondingly one die standard numbered and the other numbered 6,6,6,0,0,0 and add them together My first reaction was 2, 4, 6, 8, 10, 12, and 0, 0, 0, -1, -1, -1. The result is the same. It would work depending on the precise wording of the prompt; so would two dice labeled 13, 14, 15, 16, 17, 18. It’s only if the prompt is constructed very strictly to permit only the eleven unique results from 2 to 12, with no results out of that range, that this becomes a difficult task. ---
**Author:** ![kanicbird](https://avatars.discourse-cdn.com/v4/letter/k/5f8ce5/32.png) [@kanicbird](https://boards.straightdope.com/u/kanicbird)\ **Post date:** [April 15, 2014, 4:32pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/55 "2014-04-15T16:32:19Z")
> [@Ravenman](#): > > Except that your post #50 doesn’t actually accomplish the objective of having the two dice roll two 9% of the time; three 9% of the time, four 9% of the time, etc. Under your numbered scenario, it would be impossible to come up with the result of two, for example (because rolling 1 on the red die and 1 on the blue die would result in eight being the lowest possible total). I assume you mean post 51 since you state red and blue dice. Let’s use my second example, it’s easier but the same 1 die, numbered 1,2,3,4,5,6 1 (the 2nd) die numbered 6,6,6,0,0,0 (basically a coin flip, so I will simplify this to 2 possible outcomes 0 and 6) Role possibilities (1st die, 2nd die): 1,0=1 (can be discarded if you wish, take as the die rolled off the table, invalid role) 2,0=2 3,0=3 4,0=4 5,0=5 6,0=6 1,6=7 2,6=8 3,6=9 4,6=10 5,6=11 6,6=12 Those are the only possible outcomes, and if you include the role of 1 or throw it out (as dice sometimes have to be rerolled) the probability of all outcomes 2-12 are exactly even at 0.083333- if we allow 1 as a valid role or 0.9090- if we take a role of 1 as invalid. ---
**Author:** ![Really\_Not\_All\_That\_Bright](https://avatars.discourse-cdn.com/v4/letter/r/e8c25b/32.png) [@Really\_Not\_All\_That\_Bright](https://boards.straightdope.com/u/Really_Not_All_That_Bright)\ **Post date:** [April 15, 2014, 5:00pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/56 "2014-04-15T17:00:14Z")
_Roll_. ---
**Author:** ![Ravenman](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/ravenman/32/2929_2.png) [@Ravenman](https://boards.straightdope.com/u/Ravenman)\ **Post date:** [April 15, 2014, 5:17pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/57 "2014-04-15T17:17:42Z")
> [@kanicbird](#): > > I assume you mean post 51 since you state red and blue dice. > > Let’s use my second example, it’s easier but the same > 1 die, numbered 1,2,3,4,5,6 > 1 (the 2nd) die numbered 6,6,6,0,0,0 (basically a coin flip, so I will simplify this to 2 possible outcomes 0 and 6) > > Role possibilities (1st die, 2nd die): > 1,0=1 (can be discarded if you wish, take as the die rolled off the table, invalid role) > 2,0=2 > 3,0=3 > 4,0=4 > 5,0=5 > 6,0=6 > 1,6=7 > 2,6=8 > 3,6=9 > 4,6=10 > 5,6=11 > 6,6=12 > > Those are the only possible outcomes, and if you include the role of 1 or throw it out (as dice sometimes have to be rerolled) the probability of all outcomes 2-12 are exactly even at 0.083333- if we allow 1 as a valid role or 0.9090- if we take a role of 1 as invalid. If you have to adjust the data to come up with the outcomes you want, then the dice quite literally are not designed to achieve the outcome. For example, if I have blue and red socks in by sock drawer, and I want to pick out blue socks 100% of the time; setting up a rule that discards any pick that turns up a red sock means my system doesn’t work. ---
**Author:** ![glee](https://avatars.discourse-cdn.com/v4/letter/g/b5a626/32.png) [@glee](https://boards.straightdope.com/u/glee)\ **Post date:** [April 15, 2014, 5:29pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/58 "2014-04-15T17:29:07Z")
I recommend a Star Trek novel The Kobayashi Maru’ by Julia Ecklar, which covers how several Enterprise officers tackle the test. It wasn’t an actual test, but when one young pupil applied for a Chess Scholarship at my School, I managed to get him into a clock simultaneous display versus Kasparov! ---
**Author:** ![MaxTheVool](https://avatars.discourse-cdn.com/v4/letter/m/9fc29f/32.png) [@MaxTheVool](https://boards.straightdope.com/u/MaxTheVool)\ **Post date:** [April 15, 2014, 5:29pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/59 "2014-04-15T17:29:50Z")
> [@Ravenman](#): > > If you have to adjust the data to come up with the outcomes you want, then the dice quite literally are not designed to achieve the outcome. > > For example, if I have blue and red socks in by sock drawer, and I want to pick out blue socks 100% of the time; setting up a rule that discards any pick that turns up a red sock means my system doesn’t work. Sure, but if someone is asked the question and says “well, here’s a set of dice faces and slightly modified rules that comes very close, depending on how the question was phrased”, that’s presumably a more-likely-to-result-in-getting-hired answer than someone else who just tries sets of numbers for a while and then says “gee, seems impossible to me”, even if it is in fact impossible. ---
**Author:** ![kanicbird](https://avatars.discourse-cdn.com/v4/letter/k/5f8ce5/32.png) [@kanicbird](https://boards.straightdope.com/u/kanicbird)\ **Post date:** [April 15, 2014, 6:17pm UTC](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119/60 "2014-04-15T18:17:48Z")
> [@MaxTheVool](#): > > Sure, but if someone is asked the question and says “well, here’s a set of dice faces and slightly modified rules that comes very close, depending on how the question was phrased”, that’s presumably a more-likely-to-result-in-getting-hired answer than someone else who just tries sets of numbers for a while and then says “gee, seems impossible to me”, even if it is in fact impossible. Actually there were no slightly modified rules. I was asked to design a set of dice that would give equal probability for outcomes 2-12 I did exactly that, with throwing out the 1’s or including them, your choice, given the parameters I made every single one of them and every aspect of them exactly and without any compromise. I also feel like the poster who stated the solution as impossible should be thankful for getting the job despite the incorrect answers but with kudos points for fully explaining it. I also feel like the dimensions of the dice or weighting of them may still be able to be used (since we are allowed to ‘design the dice’) can be done that may allow standard rolling to achieve this. So in short 1 - if one can design the dice to roll all equally via weighting or non-cubic design, that would be the people who get man to walk on the moon 2 - if one can come up with the simple workable solution of renumbering one die 6,6,6,0,0,0 (or likewise), these are the people who can get Apollo 13 back to earth safely after the first group screwed up 3- if the person can only prove it is impossible then they should be thankful they were given the job if it is proven to be possible by group 1 or 2 or both. [Previous page](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119.md?page=2) [Next page](https://boards.straightdope.com/t/has-there-ever-been-a-real-life-equivalent-to-the-kobayashi-maru-test/686119.md?page=4)