Simulations can be described as "getting real-world experience without real-world consequences."

As a high school educator, I have envisioned opportunities to move textbook learning into a simulation. Students make decisions based on their STEM skills. And students experience the outcomes of those decisions. If studentsfail, they try to figure out what went wrong. And that is a great learning experience.

But simulations are expensive to develop. And there is a cultural resistance to using them. If students can pass a traditional exam, why should they put themselves to any real-world test?

During the outbreak of COVID 19, the media was trying to explain "flattening the curve" of the pandemic. I thought a pandemic simulation could show how "flattening the curve" works. People could roleplay politicians and epidemiologists making the difficult decisions of forcing interventions onto a reluctant population.

Here is the realism of this simulation. Deploy interventions and the effects of the pandemic are reduced. Lift interventions and the pandemic spreads. COVID 19 taught us this lesson many times. Yet interventions also reduce the psychological health of the people. If there are too many interventions for too long, politicians start losing popularity. It is such a balancing act. I often felt sorry for the politicians trying to find this balance.

I am hoping that this simulation teaches more world citizens to accept pandemic interventions as a small sacrifice to help fellow citizens, not as a loss of freedom. There will be more pandemics.

I had three objectives when designing this simulation. First, I had to build it on the inexpensive side, less than $10,000. Second, it had to be simple enough for a high school STEM student to use and experience. Third, it had to be somewhat realistic. It took me eight months to find a design with these three objectives. I cut back many of my original design to keep the development cost and the learning curve low-while retaining enough elements of realism for a STEM challenge

In January 2021, I called in my favorite website developer, Subha Brota Nath from Kolkata India, into the project.

This simulation uses Monte Carlo techniques to show the spread of the pandemic. Every citizen in the 1000-person fictitious village becomes a record in a database. Every day, a random number is generated for each record. That random number determines if a healthy record becomes infected or if an infected record moves to the next stage of the illness.

The nature of Monte Carlo simulations is that no two runs are the same-even with the same starting inputs. As such, there will be a little luck involved with reaching a high score in this simulation. But any high-score luck will come only after finding the right combination of interventions. That right combination will come with experimentation and analysis. In other words, STEM principles will be required to get a high score.

As mentioned in the previous section, I did not want luck to be a big factor in attaining a high score. But as the mathematical model was taking shape, luck was becoming far too important.

I really wanted to start each run with only one asymptomatic citizen as the storyline suggests. However, a one-infection start has a low infectivity. The next infection could happen on the 2nd day of the run or the 20th day. An earlier second infection means a bigger spread than a later second infection. In other words, there was a wide variance of pandemic outcomes, depending on when this second infection happened. And a wide variance means luck plays a bigger role in getting a high score.

To reduce the luck, I designed a constant early start. The player has no control over the first 25 days. Every run for every virus starts the same. By Day 25, each run has these infected citizens: eight asymptomatic, four symptomatic, one sick, and one very sick. This result on Day 25 produced much less variance than one asymptomatic citizen on Day 1. So, most of the luck has been removed. Learning from previous runs is what will bring success in Pandemic Simulator.

It is on Day 26 when the Monte Carlo math takes over.

It is also on Day 26 where interventions can be first deployed. While you can set your interventions to deploy before Day 26, they won't take effect until that day.

The Monte Carlo model runs through its math one day at a time. This data is kept in a daily database.

To keep the report on the simple side, the software takes data from every seventh day to create a weekly report.

We have noticed that sometimes this weekly data seems a little illogical. This is because the extraction of every seventh day sometimes leaves out important data from the first-to-sixth days. The derivative graph is more problematic in this regard.

During our testing, when we saw such anomalies, we inspected the daily database for possible errors. We didn't find any. The anomalies were caused by the truncation of the data.

Be assured that the Weekly Reports are still at least 98% accurate and will help you see the trends to improve your score for the next run.

If you want to see the runs with no data gaps, you can use the daily database to generate your own analytical spreadsheets.

Exponential growth or decline is a common mathematical function for pandemics. If you see a straight line on a log graph, the pandemic and interventions are following this kind of mathematical relationship. Maybe finding the rate of growth or decline will be useful in your analysis.

Useful insights can be learned from the Derivative graph, but with values that are around 0%. To see that 0% better, we limited derivative values to a maximum of 100%. If you see a value of 100% on the graph, it could be 100%, 200%, or even 300%. Focus on the areas closer to 0%. Why is it slightly positive? Or slightly negative?

In the tables and graphs, you will encounter the terms "healthy" and "recovered". Technically speaking for this software, "Healthy" is the number of citizens who have never been infected with the virus. "Recovered" is the number of citizens who have been infected but are no longer carrying and spreading the virus. In essence, the number of true healthy citizens is "Healthy" plus "Recovered."

The log and derivative graphs use the word "infected."

For the first six viruses, Infected = Asymptomatic + Symptomatic + Sick + Very Sick + Hospitalized.

For the last six viruses, Infected = Symptomatic + Sick + Very Sick + Hospitalized.

Epidemiologists use more complex pandemic simulators to predict the spread of a virus. These simulators require powerful computers and long run times because there is more math and more variables to manage. Such simulators require experts to properly set up the run parameters and interpret the results.

This simulator should not be regarded as a substitute for the professional simulators.

But this simulator still allows players to experience working with pandemic data. Comparing two sets of data and basing the next decision on that comparison is developing critical thinking skills, which can be taken to many other STEM fields of study.

Each virus will have its own infectivity and lethality. Intervention effectiveness will also change with each virus. Do you deploy heavy interventions to bring infections quickly to a low level-and annoy the citizens too much? Or do you take a softer approach just to hold the pandemic under control? The choice is yours.

Players aspiring for high scores will require at least five runs to understand what interventions work better. When a good combination of interventions is found, the player should do some fine-tuning to find the edge in finding a higher score. Plan on 10 runs to master each virus!

When you get a high score, move to another virus. It will require a different combination of interventions.

If Dave Volek's Pandemic Simulator acquires some popularity and profitability, here are the plans for expansion:

- Expanding simulated run time from 25 weeks to 60 weeks-or more.
- Two or more populations who have a different infectivity and lethality to the virus.
- Two or more concurrent viruses with different infectivity and lethality profiles.
- Vaccines, with specific efficacies to the virus.