4 ms·
I attended a talk by an economist, and he gave the following explanation of how things worked. The assumption is: the value of a seat goes to 0 at the instant a
by ajays 16y ago
I attended a talk by an economist, and he gave the following explanation of how things worked. The assumption is: the value of a seat goes to 0 at the instant a plane takes off.
For a given number of seats available, you can plot a price (Y axis) -vs- time left (X axis) curve. This curve will gently slope down and hit 0 when t=0. Now you can plot a whole series of such curves, one for each N (where N = number of seats available). As N goes lower, the curves are higher.
If you're at a certain point in the curve for a particular N, if a seat gets sold, then you can jump up to the curve for N-1 . Obviously, the airlines' systems aren't fast enough to do this in realtime; so they do this adjustment in batches (every Monday and Wednesday?).
Anyways: this was what he claimed was going on. I'm not sure how correct the above is.
- prodigal_erik 16y agoThis gets into iterated game theory: even though the value of an empty seat is basically zero at takeoff (marginal fuel and cabin crew labor don't add up to a lot), the airline has a long-term incentive not to reduce the price so that the market for that one flight clears, because in the future it doesn't want people waiting for cheap last-minute tickets they expect will again be available.