In early 2020, various top of the line advanced models, as well as standard ones like SEIR, all predicted one massive wave of infections which, in the case of Sweden, would infect around 60% and leave around 50000 dead in a matter of two months. In reality, 5000 died in that time frame and less than 10% got the virus. I thus posed the question (knowing absolutely nothing of infectious disease modeling), how can the models be so fundamentally wrong? The common explanation that "people are self-isolating" and "the strategy is working" just did not add up, in my opinion, and so I kept studying the field, looking for natural reasons that could dampen the spread. I implemented and evaluated various algorithms by leading experts, started modifying them myself and eventually ended up writing 4 research papers on the topic. In this talk I will try to give a brief overview of my contributions and conclusions.