The correct approach as a villager is to take the sensor as bayesean evidence. What is your prior that there is a wolf nearby right now? What is the probability that the boy would cry 5% wolf when there is or is not a wolf (hint: it’s probably not 5%/95%, and doesn’t even need to add up to 1)?
In villages where wolves are common, it probably shouldn’t change your estimate at all. In villages where wolves have never been a problem, it may move your estimate even higher than 5% (because the kid can be wrong in either direction, and someone even bothering to point it out is pretty unusual at all).
The correct approach as a villager is to take the sensor as bayesean evidence. What is your prior that there is a wolf nearby right now? What is the probability that the boy would cry 5% wolf when there is or is not a wolf (hint: it’s probably not 5%/95%, and doesn’t even need to add up to 1)?
In villages where wolves are common, it probably shouldn’t change your estimate at all. In villages where wolves have never been a problem, it may move your estimate even higher than 5% (because the kid can be wrong in either direction, and someone even bothering to point it out is pretty unusual at all).