The main problem with real-world implementations of knights and knaves-style puzzles is that there is a chance that the player will simply guess which door to go through and get it right without actually proving his intelligence. So here is a meta-puzzle:The king has hired (You) to design a knight and knaves-style puzzle to ensure that only those with sufficient intelligence may enter his kingdom. How do you best verify that an applicant actually used logic rather than guesswork to pick the right door if there's not enough money in the royal budget to have a logician on hand to listen to every individual applicant's explanation? You may use as many- no uh actually, limit 100 doors and 100 guards. Yeah.picrel is common and basic variant of this type of puzzle to serve as an example
>>16938621For your image, ask "if I had asked the other guard if he was guarding the safe door, would 'yes' be his answer?"An answer of "no" means the door is safe. "Yes" means it is not.As to your actual question: just increase the number of doors arbitrarily and increase the number of questions you're allowed to ask. Someone who can figure out the solution I provided above should be able to solve it with absolute certainty in at most log_2(x) questions rounded up (where x is the number of doors). Chance of guessing is 1/x.
>>16938621>>16938863>limit 100 doors and 100 guards.Sorry, missed this part. At that point the chance of guessing is locked at 1/100 unless you stretch the limits of what a knights and knaves puzzle is allowed to be. Can the correct door change positions? Can all doors be deadly until a particular condition is fulfilled?Can we do it like "le hardest logic puzzle" and have the words for "yes" and "no" be some undecyphered language? We can make this shit arbitrarily complicated in a fairly trivial manner.
>>16938867Yes, yes, and yes. Unless that makes it too easy.
>>16938867>Can we do it like "le hardest logic puzzle" make them solve the 4chan captcha
