"To understand why the Ferguson model was in much need of having peer review of the model and code, look at how his past models have utterly failed. He has a horrible track record.

• 2001: Ferguson model on foot and mouth disease to cull animals to prevent spread. Cost 10 billion pounds, turned out to be "severly flawed"
• 2002: Ferguson predicts Mad Cow disease worse case would be 50,000, then 150,000 deaths. Total deaths were 177.
• 2005: Ferguson claims 200 million could die from bird flu,. Only 282 people died worldwide.
• 2009: Ferguson claimed swine flu had fatality of 0.3-1.5%, lead to 65,00 deaths in UK. Only 457 people died, death rate of 0.026.
• 2020: Ferguson models COVID-19, worst case is 2.2. million dead in US, 500,000 in UK. No one reviewed the code or model prior to governments using it to base decisions. So far model proven false.

A "do nothing" model is basically what Sweden has done. The projected deaths in the US was 0.75 of the population, and the UK was about 0.67% of the population dying. Taking 0.7% of the Swedish population gives 71,610 dead. Is that was happened? No. Sweden has 2,586 deaths.

Ferguson's model was off by an order of 27 when looking at Sweden. That's a gross overestimation of deaths from a 'no nothing' model of governments. Sweden has made recommendations, and people have chosen to isolate or distance at their leisure. But no lockdown measures were implemented. No forced closing of businesses.

This whole Ferguson model was wrong from the get go, it was never peer reviewed, when asked, the code was not open sourced, but kept hidden, and Ferguson ran to Imperial College funded Bill Gates' company Microsoft to redo the code instead of letting other experts review it."
@krnel/neil-ferguson-bad-models-and-connections-to-bill-gates