Airplane Boarding

Gate lice

We simulated gate lice. Boarding got faster.

Gate lice are the passengers who crowd the gate and board before their group is called. We boarded the same simulated flight twice, once by the book and once with queue jumpers, and the flight with queue jumpers was full first.

A fewNearly half the plane
By the bookEveryone waits for their group to be called 0:00
With queue jumpersThe same passengers, bags and seats 0:00
  • Waited for their group
  • Boarded before their group
  • Stopped at their row, blocking the aisle

Average of 40 simulated flights with these settings: full 1:45 sooner with queue jumpers (18:05 by the book, 16:20 with them). The flight with queue jumpers finished first in all 40.

What the simulator found

Simulator result: On an easyJet Airbus A320 with 175 passengers, typical carry-on bags and one forward door, airline group boarding took 18:05 on average when every passenger waited for their group. When about 29 of them boarded one or two groups early, the same flights averaged 16:20. That is 9.7% faster, and the flight with queue jumpers finished first in all 40 paired flights.

Simulator result: More queue jumpers made it faster still: with about 58 of 175 boarding early, the average fell to 15:03. Plain back-to-front zones gained the most: 20:03 by the book, 16:52 with about 34 queue jumpers. The Steffen method went the other way.

  • Back-to-front zones
  • Airline groups
  • Windows first
  • Steffen method
  • One shuffled line, no groups

Mean boarding time in minutes, Airbus A320, 175 passengers, 40 simulated flights per point. The horizontal axis is the setting: the share of travel parties told to board early. Because whoever is called first cannot jump anyone, slightly fewer passengers than that end up boarding early.

The same numbers as a table
Boarding methodNobody early10%20%30%40%50%
Back-to-front zones20:0318:2616:5215:5115:1915:16
Airline groups18:0517:0816:2015:4315:0314:49
Windows first13:0613:2013:3013:4813:5714:09
Steffen method9:3710:2611:1312:0712:5613:36

It was not only this aircraft

Simulator result: Airline group boarding with the same queue-jumper setting, on the five aircraft tested here. The gain was between 6.8% and 10%.

AircraftPassengersQueue jumpersBy the bookWith queue jumpersChange
easyJet Airbus A3201752918:0516:209.7% faster (±1)
Ryanair Boeing 737-8001803018:3316:4110% faster (±1.4)
American Airbus A321neo1843520:3019:076.8% faster (±1.1)
Delta Airbus A220-3001212414:3113:287.2% faster (±1.4)
United Boeing 787-92314613:2612:247.6% faster (±1.4)

Mean of 40 paired flights per aircraft. The figure in brackets is the 95% confidence interval of the mean change, in percentage points.

Why queue jumpers speed up group boarding

Boarding is slow because of bags. Model assumption: In this model a passenger lifting a bag into the bin blocks the aisle at their row, and the row behind it, until they sit down. Everyone behind them stands and waits.

Group boarding makes that worse than it needs to be. When the gate calls the rear of the cabin, every passenger in the aisle is headed for the same ten rows. Only a few of them can reach a bin at once, and the rest of the aisle stands empty in front of them.

A queue jumper is usually headed somewhere else. While the called group waits its turn at the back, the queue jumper stops at a row nobody else is using and stows a bag at the same time. Simulator result: With airline groups on the A320, the average number of passengers stowing a bag at any moment rose from 1.19 to 1.32, and the time each passenger spent standing still in the aisle fell from 71 to 58 seconds.

Queue jumpers, in other words, shuffle the line. Simulator result: One fully shuffled line with no groups at all boarded this A320 in 14:50, well ahead of airline groups at 18:05.

Research finding: Published work points the same way. Jason Steffen's 2008 simulation study and the 2012 experiment with 72 people in a mock cabin both found that boarding in row blocks was slower than letting passengers with assigned seats board in random order.

Who pays for it

The flight leaves the gate sooner, but the time saved is not shared out evenly.

Simulator result: On the A320 with airline groups, the queue jumpers sat down 5:05 earlier on average than they would have by waiting for their group. The passengers who did wait came out about even as a group, sitting down 0:08 earlier on average. Individually it was a coin toss: 51% of them sat down later than they would have on the by-the-book flight.

Model assumption: The overhead bins never fill up in these runs. On a real full flight they do, and a passenger who boards early takes bin space from somebody who was entitled to board first. That is a large part of why people jump the queue, and why the people behind them mind.

Where queue jumpers do slow boarding down

Simulator result: With the Steffen method, the same queue-jumper setting made the A320 slower: 9:37 by the book, 11:13 with queue jumpers, 16.6% more. It was slower in all 40 paired flights, and 92% of the passengers who waited their turn sat down later.

The Steffen method works by spacing passengers two rows apart so that many of them can stow at once. Each person out of place lands in one of those gaps and closes it. The more carefully a boarding order is tuned, the more it loses when passengers ignore it.

Simulator result: Windows first, aisles last sat in between. On the five aircraft tested the change was small in both directions, from 1.9% faster to 4.2% slower.

So the result is less a compliment to queue jumpers than a measure of group boarding: in this model it is slow enough that disorder improves it.

What the simulation leaves out

These are simulated flights, not measurements at an airport. Model assumption: The model is kind to queue jumpers in several ways:

  • Queue jumpers are let through. No gate agent sends anyone back, nobody argues, and a rejected boarding pass costs no time.
  • The crowd in front of the gate does not exist here. A knot of people blocking the lane can slow down the passengers who were called, before anyone reaches the aircraft door.
  • A queue jumper moves one or two groups ahead and merges into that group's line at a random place. Nobody walks straight to the front of the whole line, and nobody in the assistance group is displaced.
  • The aircraft boards through one forward door from a jet bridge, and every passenger turns up.

The absolute boarding times of this model have not been validated against airline data. The comparison between two runs of the same flight is the result, not the minutes.

Simulator result: The direction of the result held under the other assumptions we tried, on the A320 with airline groups and the same queue-jumper setting:

Changed assumptionBy the bookWith queue jumpersChange
Heavy bags and bins that fill up34:1930:4310.5% faster
No families or travel parties19:4317:3410.9% faster
Each group called 60 seconds apart18:3516:509.4% faster
Passengers drift both ways: some early, some late18:0517:174.4% faster

Airlines also have reasons for a boarding order that have nothing to do with speed: an early boarding group is something they sell. In October 2024 American Airlines began testing gate software that sounds an alert when a boarding pass is scanned before its group has been called, and a month later said it was expanding the system to more than 100 airports (Associated Press, Fortune).

Questions

What are gate lice?

Gate lice is airport slang for passengers who crowd around the gate and try to board before their boarding group is called. On this page a queue jumper is a passenger who boards one or two groups earlier than the group printed on their boarding pass.

Do gate lice slow down boarding?

Not in this simulation, when the airline boards by group. On a simulated Airbus A320 with 175 passengers, airline group boarding averaged 18:05 with everyone in turn and 16:20 with about 29 queue jumpers, which is 9.7% faster. The simulation does not include the crowd at the gate itself or arguments with gate agents, which can cost time before anyone reaches the aircraft.

Why would people boarding out of order make it faster?

Group boarding sends everyone in a group to the same part of the cabin, where they wait behind each other to put bags in the bins. A queue jumper is usually headed for a different part of the cabin, so they can stow a bag at the same time as somebody else. The line becomes more mixed, and published research has found that a mixed line boards faster than row blocks.

Does that mean I should board before my group?

No. In the simulation the time saved goes mostly to the queue jumpers themselves, who sit down about 5:05 earlier, while 51% of the passengers who waited sit down later than they otherwise would have. On a real flight the queue jumper also takes overhead bin space from passengers in earlier groups, which this model leaves out.

Is there a boarding method where queue jumpers do slow things down?

Yes. With the Steffen method, a carefully ordered sequence from research, the same share of queue jumpers made the simulated A320 board 16.6% slower (9:37 against 11:13). The more precise the boarding order, the more it loses when people ignore it.

How these numbers were made

Each comparison is a pair: the same seeded flight, with the same passengers, seats, bags and walking speeds, boarded once with everyone in turn and once with queue jumpers. Every figure on this page is the mean of 40 such pairs, seeds 20261100 to 20261139. The cabin above plays one pair at a time, so a single flight lands near the average, not on it.

Settings: typical carry-on bags, mixed walking speeds, a quarter of eligible seat groups travelling as a party, 3% of passengers needing assistance, one forward door, load factor as listed per aircraft. Airline groups here means assistance first, then families, a few premium rows, and then the rear, middle and front of the cabin.

The engine is the one behind the boarding simulator on this site, and the guide describes how it moves passengers. The figures on this page are in one data file, generated on 2026-11-10.

This is an educational simulation. Results depend on the assumptions and passenger behaviour selected. It is not an airline-certified operational tool.

Open the full simulator