Originally Posted by
hhdl
Of note, it's a well known statistical phenomenon that the average observed load factor by a passenger will be higher than the average load factor (using load factor to mean what a passenger could observe: seat occupancy, ignoring revenue etc.) reported by an airline.
To make an extreme example. Imagine 10 flights with 100 seats apiece. 4 flights go out completely empty and 6 go out 100%.
Airline's perspective: 600 seats sold/1000 seats available = 60% LF
Survey passengers: 600 out of 600 will report 100%
Even ignoring that you're more likely to get a volunteered response if the LF is abnormal, the sample in a passenger survey weights flights with more pax, which correlates with LF.
Right, you made my point. In your example, I am much much more likely (certain?) to be traveling on one of the full flights. Therefore, the "anecdotal 100% " is more more realistic prediction of what to expect when I am on the plane than the "real 60%".