Crank up social distancing
Flatten the curve and ease the strain on the hospital
adjust social distancing force
person symbols
About this Simulation

Created by Jeffrey Ventrella

1. Purpose
This is an abstract simulation that illustrates how social distancing reduces transmission rate of a contagious agent in a population, such as the virus that causes coronavirus disease. This reduces the number of people demanding medical attention at any given time.

This simulation DOES NOT use real-world data as input, nor is it intended to generate data that is consistent with any particular population. The passage of time is abstract - it is not meant to represent days or weeks. The hospital is included to show that limitations in health care in a time of intense demand can result in death.

Results will vary from one simulation run to another, because of randomness. But in general, fewer deaths result when social distancing is set higher.

At higher social distance values, sometimes the initially-infected person never touches anyone before it recovers or dies, so you may need to restart again to see how the infection spreads.

Try starting over with social distance at about 60%...then after abut 3000 timesteps, bring it down to 0%. See what happens when social distancing is relaxed too soon.

2. Transmission Rates
Transmission rate is modeled via a simulation of balls (representing people) that wander around randomly, frequently bumping into each other. The simulation starts with all people uninfected, except for one. When the infected person touches an uninfected person, that person becomes infected. Infection spreads faster as more people become infected, until a significant percentage of the population is infected, after which time transmission rate begins to decrease.

3. Social Distancing
The more frequently people touch each other, the higher the transmission rate. The social distancing slider directly effects transmission rate: when it is increased from "none" to a higher value, the people exert a repelling force between each other, if they are within a certain proximity. This reduces the chances of them bumping into each other, which flattens the curve.

If the slider is increased to the maximum value, the repelling forces between people becomes large enough to eliminate any chance of people touching.

4. Sickness, Recovery, and Death
A person can be in one of 4 states:
1. uninfected
2. infected (symptomatic or asymptomatic carrier)
3. recovered (immune)
4. dead

An infected person stays infected ("sick") for a fixed duration of time. Sickness increases linearly to the halfway point, and then it decreases linearly. People with low immune system strength will get more sick than average, and are more susceptible to dying (dying happens when sickness exceeds a critical maximum sickness value).

5. Immunity
Most people recover after the sickness period and then become immune (no longer a carrier, and no longer able to infect others). The population acquires herd immunity when the number of recovered/immune people is high enough to provide a shield to transmission.

6. The Hospital
The hospital constitutes a bottleneck to health care when the demand for care exceeds the capacity of the hospital. When there is a sudden surge of people needing care, the hospital reaches maximum capacity and can not attend to all the critically-sick people, so some of them may die in neglect.

The hospital always grabs the sickest people.

Hospital capacity is limited by three factors: (a) the length of time required to bring a person back to health (by continually decreasing the person's sickness value by a fixed amount); (b) the time it takes to bring people in and to take them out (only one person at a time); and (c) the maximum occupancy (16).

The hospital prioritizes removing dead people and bringing in sick people over releasing recovered people, which is why recovered people may not be released immediately. The exact details of these limitations are not important: the point is that the hospital has limits to how quickly it can provide care, and how well (like real world health-care systems).

Created by Jeffrey Ventrella:

results from adjusting social distance

Simulation parameters

Number of people = 200

Time marches forward in integral time steps.

The velocity of each person is changed via small randomized forces at each time step. Velocity is dampened for smoother motions.

Social distancing force is applied between any two people whose distance is below a fixed radius. This repelling force is stronger when the people are closer. The force ranges from 0 to a strong-enough force to keep them apart. The details of these physical forces are not as important as the end result: one parameter that changes the frequency of people randomly bumping into each other.

Immune system strength ranges from 0 to 1, and is distributed evenly among people. It remains constant throughout the simulation.

Sickness duration = 2000.

Sickness starts at 0 and reaches max sickness after 1000 time steps and then decreases, reaching 0 after 2000 time steps.

Max sickness = 1 - immune system strength.

Critical sickness threshold = 0.6 (qualifying the person for a potential hospital visit)

If sickness becomes greater than 0.9, the person dies.

Hospital healing rate is 0.0003 per time step. This is how much sickness is decreased per time step while the person is in the hospital. If the person had entered the hospital in a very sick state (close to 0.9) it may not be possible to bring the value down enough to save the person.

Maximum hospital occupancy = 16 people.

Duration of a single hospital patient transport = 50 time steps.

Other visualizations, and references

The online article, Why outbreaks like coronavirus spread exponentially, and how to 'flatten the curve' appeared in the Washington Post in mid-March, 2020. It was written by Harry Stevens. It features a simulated illustration of 200 bouncing people, showing how various parameters change the nature of the curve.

The use of statistical mechanics as a graphical tool for making complex information intuitive, as Harry has done, was the inspiration to design the simulation on this web site.

An article in the New York Times offers a similar interactive exploration of flattening the curve.

Other related visualizations: