A model is a simplification of reality. Since the COVID-19 pandemic began, researchers, academics, and others, have built models to try to understand and forecast how COVID-19 is impacting, and will impact, Virginia and the nation. While it is tempting to dive into the data, it is important to understand what models are, what they can do, what they cannot do, and why we use them. Remember, the goal of epidemic modeling is not to predict cases exactly, but to project possible outcomes given hypothetical scenarios.
What can models do?
- Help us understand complex epidemic dynamics.
- Make clear what we do and don't know about these dynamics.
- Identify ongoing trends in the data.
- Make approximate projections of future outcomes.
- Make approximate estimates of hospital burden and resource needs.
- Allow us to estimate the effects of interventions and variants.
What can't models do?
- Make precise predictions about future events. Models are tools, not crystal balls.
- Make choices among competing priorities.
- Make up for a lack of knowledge, data, or subject matter expertise.
Why use models?
- To make reality comprehensible.
- To see signals in the noise.
- To guide and test policy options.
- To project potential outcomes.
The goal of these efforts are to inform VDH and the public in general. Models give a rough idea of what is coming under different scenarios, can help the public health community optimize countermeasures, and can estimate resource shortcomings.
Remember that “all models are wrong, but some are useful.” – George Box.