Ouch. If this were ADoM I would say to start kicking the walls... but this is Nethack, and without the right scroll, spell, or piece of equipment, there's no guarantee you'll ever find that door. I feel for you, Dudley... ouch.
This brings a question to mind: if you're standing next to a hidden door, and there's nothing else notable adjacent to you (say, a trap), what is the exact probability of finding the door by [s]earching once? Does this probability ever change?
Got it... assuming 0 luck and no lenses or excalibur, the chance of finding an adjacent door is 14%. To ensure a good (>95%) probability of finding a door, one would need to search a location 20 times (although, to ensure this probability, Dudley would only need to search every third location, since all adjacent squares are checked simultaneously.
To ensure that the entire perimeter of that given area is checked, Duds would need to n20s at 19 squares (this checks some perimeter squares twice, of course). The probability of finding the hidden door, assuming there is only one (already an unlikely assumption), can be expressed as a binomial distribution.
At most, Dudders would find the door on the very last of the 19 squares (assuming he finds it in the 95% probability range). This means it would have taken at most 380 turns just to find a door out of the area. Although, I've made several assumptions about certain probabilities and conditions, so this may not be a likely estimate of the probability.
If I were even more bored and if it weren't 3:18 AM here, I might calculate the probability distribution for the number of turns without all those assumptions... but I need some sleep now.
Also, I hate trapdoors.
#6
Anonymous
Tue, 8 Sep 2009 07:33:14
Also, doors are more likely to appear in vertical walls than in horizontal ones. Experienced NetHackers (or NetHackers who studied the level generation code) are better at guessing where the door will be.
Sorceror: Knowing Dudley's luck, that did happen :p
NobodySpecial: Wasn't that particular one a woozle (and wizzle once Piglet joined in)? (Winnie-the-Pooh story for those who don't know. I used to read those all the time.)
Ah, Dudley gets familiar with what I call "The Searching Game."
Seriously though, Rogue had the probability of doors being secret be related to dungeon depth. Nethack appears to have, shall we say, done away with that particular idea, much to my chagrin and the chagrin of characters of mine who have died by kicking the walls in sheer frustration.
14 Comments
/me heads off to wikihack
To ensure that the entire perimeter of that given area is checked, Duds would need to n20s at 19 squares (this checks some perimeter squares twice, of course). The probability of finding the hidden door, assuming there is only one (already an unlikely assumption), can be expressed as a binomial distribution.
At most, Dudders would find the door on the very last of the 19 squares (assuming he finds it in the 95% probability range). This means it would have taken at most 380 turns just to find a door out of the area. Although, I've made several assumptions about certain probabilities and conditions, so this may not be a likely estimate of the probability.
If I were even more bored and if it weren't 3:18 AM here, I might calculate the probability distribution for the number of turns without all those assumptions... but I need some sleep now.
Also, I hate trapdoors.
NobodySpecial: Wasn't that particular one a woozle (and wizzle once Piglet joined in)?
(Winnie-the-Pooh story for those who don't know. I used to read those all the time.)
Seriously though, Rogue had the probability of doors being secret be related to dungeon depth. Nethack appears to have, shall we say, done away with that particular idea, much to my chagrin and the chagrin of characters of mine who have died by kicking the walls in sheer frustration.
Yes, that sounds right. I guess I just lost Pooh points. >_<