Quote:
Originally Posted by GGG
The value of notifying people in each case is near zero and statistically indistinguishable with current data.
|
It is nearly a mathematical certainty that people on are asleep at the time of the alert will on average have less info than those who are awake. Consider the case where police are looking for a vehicle. The search radius expands linearly with the time since the kid was last seen. Of people who have been asleep for the latter half the time between the time elapsed between "last seen and "alert issues", only those in the inner circle (half the current radius) could possibly have information, and whatever information they might have would be older than someone in the outer circle. But a person who is awake at the time of the alert could have information regardless of their distance from the kid's last known location.
The chance for an individual to have info is near zero, but it's very likely even nearer to zero for people who are asleep when the alert goes out. The only way it could not be is if there's some counterintuitive offsetting mechanism that overpowers obvious effects like the one I described above.