Moussaïd, M. (2019). Crowds on the move. In R. Hertwig, T. Pleskac, T. Pachur, & the Center for Adaptive Rationality (Eds.), Taming uncertainty (pp. xx–xx). Boston, MA: MIT Press. doi:XXXXXXX


This interactive element illustrates the collective dynamics surrounding a simple binary-choice problem.

Imagine a crowd of people facing a decision between option A and option B, while not knowing for sure which option is better. Such situations are ubiquitous in daily life: a group of commuters choosing between two exit doors in a subway station, shoppers picking one of two checkout lines, people evacuating a building through one of two corridors, or pedestrians choosing between two restaurants. How do people make this decision and what are the collective consequences?

You'll be presented with an animation below where 21 pedestrians face a choice between two exits. Only 20 can exit in each round. The right door is the better option: It fits 12 pedestrians at once, while the left door only has capacity for 8. The pedestrians’ choice is explained by one of these three decision rules:

  • In the , individuals choose an exit randomly, irrespective of the decisions of the others.
  • In the , individuals first collect information about the current exit usage and choose the less congested option.
  • In the , people have a preference for the most popular option.

Click start to observe the first of three simulations. Guess the decision rule the pedestrians are following based on their behavior. If you have already tried out this interactive element you can skip the example simulations.

Decision rule efficiency

The random choice rule yields a smooth traffic flow as long as the inflow of individuals remains lower than the exits’ capacity. The minority mode offers the best results. It can even handle situations where the inflow is moderately higher than capacity. In the most-popular mode, however, the crowd gathers around a given exit, creating serious congestion while leaving the other exit underused.

Which of these three behaviors can be expected in real life? The answer partly depends on the degree of uncertainty of the situation.

The random mode assumes that individuals do not pay attention to the behavior of others. In practice, this is unlikely to happen unless the choices of others are hidden. The minority mode is typical of situations where little uncertainty surrounds the payoff of each option. When lining up at the supermarket, for example, consumers know that all the checkouts are roughly equivalent. They therefore tend to choose the less congested option. But in which situation would people prefer the most crowded exit? Typically when a high degree of uncertainty surrounds the payoff of each option. During an emergency evacuation, for example, people might be unsure about whether a given exit is still open or blocked by fire. In this case, the choices of others can be used as a cue indicating that one option is safer than the other.

Simulate your own scenario

Now you can configure your own simulation of pedestrians trying to exit through one of the two doors. Their decision rule is set by the parameter. In most-popular mode you can vary the , which is how likely the pedestrians are to imitate others. The parameter determines how many pedestrians appear per round. How many pedestrians exit is defined by the doors’ . By changing the doors’ (making one door bigger or smaller than the other), you can allow more or fewer pedestrians to pass through each door.

Individuals choose a door randomly, irrespective of the decision of the others.

The higher the imitation strength, the more imitators there will be.

Note: You can update the parameters of this simulation as it is working.
Capacity used: Left exit 0 % | Capacity used: Right exit 0 % | Simulation round 0
anxious-agent overlay image

Too many pedestrians in the queue! Inflow parameter has been reduced to 10 and pedestrians are now choosing the least congested exit (minority mode). Take a moment and watch how the doors' capacities change or restart the animation!

question-agent image

Which decision rule is this? Random, minority or most-popular?


The main idea conveyed by this interactive element is that simple decision rules, such as deciding randomly, following the minority, or imitating the most popular choice, can have unplanned collective consequences.

In the most-popular mode, for example, individuals tend to follow their neighbors. At the individual level, this can be an efficient strategy to deal with the uncertainty of the environment because one can benefit from the knowledge of others. However, if too many people adopt the same behavior, the best exit can become congested, creating a maladaptive collective pattern. The efficiency of a strategy is, therefore, a matter of circumstance.

You have earned a badge! Click me to read more.

Click on any reward badge in the footer to review take home messages of this or other elements.