In my job, I often have to prepare engineers for grueling orals exams for us to win Government contracts. Often the Government’s approach to selecting a winner once they have downselected to three or so companies based on price and technical capabilities is to give you a “sample task” representative of the problems they face today in their program.
We are typically only given two week’s notice and sometimes only a week’s notice as to when they want to do these exams and I will have to take six subject matter experts in esoteric engineering fields and build a cohesive team in a short amount of time. While I have a specific technique for doing this that I cannot fully divulge here. I can tell you that as part of the practice questions, we do indeed create an impossible scenario to test the team out under stress. It’s not truly ‘impossible’ for them to answer, but there is no way they will have all the information they need to answer it properly, nor will they have enough time to provide all the details needed.
The trick is for them to realize they are missing the data, make up numbers, justify why what they made up is quasi-valid based on what is known about the situation presented, and answer as much as they can in the limited amount of time they have. Most people feel like they failed after they get this question, but I always tell them afterwards that was the point of the exercise and it is very effective in getting people who don’t like to improvise or ‘guess’ (like engineers) to do exactly that.
Some of the bio-chemstry and genetics exams I had in college were like this as well - too many questions and not enough time on the exams. Every single person walked out with some of the questions unanswered and felt like they failed it. Then you’d find out afterwards that because it was graded on a curve, you still got an ‘A’ even with a 70%. At least with my ‘impossible’ question, I let everyone know that was the point of the exercise. The teachers who gave those exams let you sweat through several days wondering if you should drop the class because you failed the exam only to find out you passed.
Fractions. They said nothing about fractions! We could map them. You say, "Here is a map that translates results into the numbers 2-12. The interviewers look puzzled at first at the numbers. Especially at the part that says, “On rolls of 34, 35, and 36, exclaim ‘takesey-backseys’ and re-roll.”
This appears to be closest to the actual question of the OP and I can provide anecdotal support for it. A buddy of mine became a cop and in the police academy they were given a problem in which they were faced with a bomb with a two minute timer being discovered next to the operating theater where an important witness was undergoing a medical procedure which, if interrupted, would kill him and the bomb could not be defused in any way.
The cadets were rated on their ability to think through the problem, (and the speed with which they recognized that the situation was hopeless).
Was this more of a Dungeon Master role playing game scenario( “I try to defuse the bomb.” “Your attempt to defuse the bomb fails.”), or was this a live action course where the bomb was a physical prop that made a “boom” noise when it went off, and the functionality was sealed inside very securely? What if the candidate threw the bomb at the wall, or made a wall of tables, chairs, beds, miscellaneous hospital equipment, his hat, shirt, shoes, and whatever else he could find between the bomb and the patient? Did the test administrators recalculate the blast radius and vectors or was that not allowed? If it wasn’t allowed, that would seem like a true rigged scenario.
“I spit on the bomb.” “Nothing happens.”
“I kick the bomb.” “You stub your toe. Nothing happens to the bomb.”
“I pee on the bomb.” “Nothing happens. The Chief sees you and demotes you.”
“I fire a bullet at the bomb.” “Your gun jams.”
“I clear my gun jam, reload, and fire a bullet at the bomb.” “Your gun jams.”
It was a thought exercise in which every proposal to defuse the bomb or move the patient was countered with an explanation that the effort would fail.
(E.g., the bomb was placed in a container with motion sensors so that any effort to move the bomb caused it to explode. The motion sensors fed through the trigger circuit on a dead man’s switch so that disabling the sensors triggered the bomb.)
I was not in the exercise, so I do not have every detail and I am not going to try to second guess the mechanics (or electronics) of the scenario. My point was merely to provide anecdotal support that there are agencies that do use a “Kobayashi Maru” exercise to test their students.
I’m still not following you. I’m not saying you’re wrong (I’m sure that I am), I’d just like to understand this.
Say there is a lake that holds 50Ml of water. The water level may fluctuate but over a year the amount of water added via run-off from nearby hills, and the amount lost via evaporation and drainage is about the same.
Then it turns out it’s feasible to dig out material from the lake.
So we remove 20Ml of earth. Now the lake still holds 50Ml, but the water level is much lower. For the water level to return back to where it was, it must gain more water than it loses but why has the run-off/evaporation-drainage equation changed?
Or, are we saying that the flooding itself is actually an important part of how the water level is set? The lake gains more water than it loses every year, and floods. If the lake were dug deeper, it would gain water year on year because it doesn’t have the flood “mechanism”, until the water level is back where it was.
I figured out a way to do it with 8 results, not 11.
One die is numbered cylindrically (one side and one opposite side is left blank on both dice, so 4 sides only are used) from 1-4. The other die is numbered 5-8 cylindrically. Glue the two dice together at a 45 degree angle on blank sides so that when rolled, only one die can be face up while the other die is on the edge.
I think the 0-1 and 13+ answers are just cop outs.
I’m sure I’m not explaining it well. This was decades ago. But there was a lake that flooded in heavy rains. The answer that some teams came up with in the past was just to dig the lake deeper, like that would keep it from flooding. But no, people wanted the surface of the lake where it was, so if you dug the lake deeper, they’d just pour even more water in it to get the water level at the desired, um, level. So then it’s at the level it’s always been before anyway the next time it rains hard and the the lake floods the neighborhood houses.
Solution: drain the lake; install a drain plug in the center, with a long pipe that leads to a lower part of town; refill the lake. When it starts to rain heavily, pull the plug.
Or have the piping lead to a pumphouse; pump the excess water to a water tower.
There does seem to be something missing from this. My take on it is that to get ‘the impossible situation’ to come about the watertable would need to also be at the lake’s level (or close enough), then digging it deeper would not matter (it would just refill), although it could work if you break through a impermeable layer of earth to a dryer subsurface soil to draw down the local water table - but that would depend on such a sub surface condition existing. There are wells that sometimes can be dug for exactly this purpose.
Additional solutions would involve delaying/stopping run off from getting into this lake, catchment basins, and water diversion channels on the hillsides that drain into this lake. Plants planted on the hills also can help in some situations.
Making it wider could also help if the water table is only locally that high and the lake could be extended to a area with a lower one, again the right subsoil conditions and topography would need to be present.
I wonder if this might have arisen as a corruption of the puzzle of designing two nonstandard six sided dice [with positive integer faces to be added in just the usual way; no trick question stuff] so that the probability of all results from 2 to 12 was equal to their probability if the dice were standard. (Aka, Sicherman dice)
It makes some sense to me. Presumably, you want a pilot to continue attempting to aviate even in the middle of a situation he believes to be unrecoverable. After all, he may be mistaken, his instruments may be faulty, etc. A good pilot should die attempting to recover from a situation that he believes to be unrecoverable, even if that belief is correct, and the choices he makes should be rational until the moment of his death. Testing this in a simulator is a not-bad idea, though one can never generate the fear of a real-world unrecoverable problem.
Again, though - the real world has a lot of problems that appear unsolvable. Some of those problems actually are unsolvable, some are just very, very difficult. If you’re running a military or quasi-military organization, you want officers who will try their darnedest to solve problems until they’re dead or incapacitated, even if they genuinely believe the problem cannot be solved.
Label the sides of one die 1 through 6, as normal. On the other die, label three of the sides 1, and the other three 2. Roll the dice and multiply the results. There’s an equal 1/12 chance of rolling each result from 2 to 12.
There’s also a 1/12 chance of rolling a 1, but the question didn’t say that you couldn’t have other results – only that the probability of all results from 2 to 12 were equal.
Another possibility is to give each die six blank faces. That way, the probability of all results from 2 to 12 is also equal: zero.
