I remember that in early spring when COVID pandemic started everyone was fascinated with all the mathematical modeling tools that predict the spread of the virus. It seemed that finally, some math is getting into the mainstream. Now the situation drastically changed – all the tools that were used (at least in Poland) failed to predict the second wave. Aposteriori it is rather clear that these tools are based on flawed assumptions. Most of the prediction tools use mean-field approximation, which makes the cascading process deterministic. In such an approach, uncertainty can only appear as a result of our lack