[gus_massa]

In https://medium.com/@joschabach/flattening-the-curve-is-a-dea... the estimation is that the number of ventilators is only a 2% of the total of the population that will need them. So the curved need to be flattened even more. (I don't know if that is really the limiting factor of the attention.)

The numbers of the days in the x axis follow a strange pattern. What about using the multiples of 5 or 7?

If I change the number of "Random People Met Daily" to 1, the peak of the curve is outside the range of the graph. I guess the scale should be configurable.

The number in the "Close Circle Size" is used to simulate small clusters, or each day it is used to pick someone at random from the population? (Does each sim has it's "family"?)

[kasra85]

With regards to the need for flattening even more, I totally agree. But I think you are under the impression that the y axis in this model can simulate the real world. That's not the case. This is a "super simplified model" to show the effect of simple changes - relative to other scenarios, not necessary being an absolute measure.

The close circle size is the family/close friend size. In actual scientific models, this is a decimal value to represent the average of the community, but in this model, I kept it as a whole number to simplify my calculations. If I get a chance, I'll improve this further.

Anything that falls into more than 150 days is way beyond any sane model can predict. Cutting the "Random Contact" to 1 greatly drops the rate for at least 150 days - this is exactly the point of this model on showing how important it is to reduce social interactions.

[gus_massa]

I just noticed that the simulation does not have a field for the total size of the population. The initial state is 5 sick people in a population of 5000.

The initial grow is exponential until the logistic part of the curve kicks in. The current grow is something like 25% daily, so a x10 in the population is a delay of the peak of 3 days or something like that. It would be more days if we flatten the curve. (With the initial values in your simulation, the peak is at 60 days. For a big city with 1 million people, the delay is like 2 weeks.)

It's clear that this is a very simplified model, but they are useful to get a feeling of how the parameters affect the epidemy.