Originally Posted by
WolfvanWeen
It most definitely depends on the moon phase. Depending on other star configurations it either happens in full moon or it doesn't. The more you become observant of these configurations, the more likely you will detect a pattern.
Which leads to another ( non astronomically based) question about availabilities which is of a somewhat statistical nature.
One would expect a somewhat random distribution of availabilities, though a bit forward sloped curve, sometimes interrupted by some Close-In dates being available because last minute unsold seat are put into system by the relevant carrier(s). I can understand those. Sometimes however as example, there may be No availabilities for say the entire calendar months September and October, but then suddenly there are three days of availability within one week in November, then No availability for entire calendar months of December and January but then suddenly on February 1 there are three, and on February 2 and 3 there are two on each day and then nothing for rest of calendar month of February followed by availabilities almost every day in calendar month of March ( Clustering). Likewise the probability of an longer dated availabilities in the middle of a calendar month should be near equal to availabilities during the first 3 to 4 or last 3 to 4 days of a calendar months. However,
if there are availabilities in a given calendar month I see a concentration in the first 3-4 day as well as the last 3-4 days of the calendar month.
Passenger booking patterns always follow a continuous upward sloping curve and while I am taking into account seasonal changes in schedules, there must be some non-automated i.e manual intervention to remove entire sections of periods of time, most of which also seem unrelated to major holidays. A statistician among us should be able to prove non-randomness as well as non-continuity. Does anyone recognize this pattern too and agree or disagree.