Which boarding method gets everyone seated fastest?
Rear window seats board first and front aisle seats last, with diagonal bands in between. It is outside-in and back-to-front at the same time, and a version of it was used by America West and later US Airways in the 2000s.
Score: 100 = as fast as the faster of the Steffen and Cincin orderings on this flight and seed · 50 = one seeded random queue · below 50 = slower than that queue.
The reverse pyramid is the compromise between WILMA and back-to-front. Each group is a diagonal slice of the cabin: rear windows, then middle-rear windows and rear middles, and so on to the front aisles. Passengers in the same group are spread along a good length of the aisle, so fewer of them stow bags in each other's way, and windows still fill before aisles so almost nobody climbs over a seated neighbour.
In the simulation it usually lands close behind WILMA and well ahead of back-to-front. Race the three together with the heatmap on and you can see the difference in where the congestion forms: back-to-front builds a hot spot that marches forward, the pyramid spreads it thin.
Boarding a full aircraft takes on the order of tens of minutes, and the order in which passengers are called changes that more than anything the crew can do once people are aboard. This is an interactive airplane boarding simulator: pick a cabin, pick a boarding method, and watch every passenger walk in, stow a bag, squeeze past a seated neighbour and sit down, one simulated second at a time.
You can race up to eleven boarding methods side by side on identical passengers, including back-to-front zones, front-to-back, rotating blocks, WILMA (window, middle, aisle), reverse pyramid, the Steffen method, open seating, a realistic airline group order, and a custom rule you build yourself. Thirteen real cabin layouts are included, from a 72-seat ATR 72 turboprop to a 489-seat Airbus A380, along with single and multiple boarding doors, carry-on luggage load, families travelling together, priority boarding and passengers who ignore their group call.
The simulator also runs in reverse. Switch to Deplaning and everybody starts seated: compare row-by-row politeness with a free-for-all, aisle seats first, or alternate rows, and see that in this model the door onto the jet bridge, not the aisle, limits how fast a full cabin empties. My seat takes a seat number and tells you, for every strategy, when you would be called, how long you would stand in the aisle and when you would sit down. Sweep a setting varies occupancy, luggage, families or the share of passengers boarding out of turn across their whole range and charts where the ranking flips. Any race can be recorded as a video with the strategy names and clock captioned, and any result can be shared as a challenge link that dares the recipient to beat your time on the identical flight.
Every run is deterministic: the same seed and settings always produce the same passengers with the same bags and seats, so a difference in the result comes from the queue order alone. The full user guide explains every metric, the mathematics behind each method, and the assumptions the model makes.
This is an educational model rather than an airline operations tool. Published boarding research supports several of the broad patterns it reproduces; the exact times are outputs of this simulation.