[marktwain]

As a result, Imperial’s model is vulnerable to producing wildly different and conflicting outputs based on the same initial set of parameters. Run it on different computers and you would likely get different results. In other words, it is non-deterministic.

How is such a model used? It is simple. The "expert" runs the model multiple times. They may tweak inputs a bit if the results don't "look right".

When the results "look right" (confirm the "expert's opinion"), they are accepted and treated as gospel.

It is the way it works.

Models used for political purposes are not independent indicators of the best policy decisions.

They are props used to support the particular "expert's" desired result.

[ProtectOurFreedom]

As a result, Imperial’s model is vulnerable to producing wildly different and conflicting outputs based on the same initial set of parameters. Run it on different computers and you would likely get different results. In other words, it is non-deterministic.

Honestly, that sounds like a Monte Carlo Simulation where you instruct the computer to vary each input according to a probability distribution you specify. Millions of runs, each with different results, then converge on a most probable answer.

Github returns 10,959 repository results for "Monte Carlo." There are 1,712 hits in C++ (others who have read the code say it is in C++ and not in FORTRAN). There are four hits for "Monte Carlo Pandemic."