@Y0uCann0tBeSer10us
I came across this interesting article at the weekend which discusses the SAGE modelling that has been widely reported and is being used by some to justify calling for tighter restrictions. Now, I knew that modelling was often wrong (usually being too pessimistic - e.g. this summer), but I had assumed that they were modelling all possible (or even likely) scenarios. It turns out that this latest modelling doesn't even consider the possibility that omicron is milder than delta (which is looking more likely every day), and if you do use include this possibility as a variable the picture suddenly looks a whole lot better and more restrictions become unnecessary. One of the SAGE modellers goes on to explain that they model what they're told to, which is only scenarios where action would need to be taken (i.e. not the positive ones where no action is needed). I also read somewhere else I think that they only considered hospitalisations and deaths, and assumed that others were modelling the effects of restrictions on various aspects (I'm not so sure they are, and have never seen this..). I find this totally astonishing, although it does go some way to understanding the disconnect between what the model says will happen and reality. What is worrying is that this one sided, overly pessimistic picture
is often used as evidence in the call for more restrictions.
In fairness, assessing "reasonable worst case" is an accepted way of modelling risk/ impact - which gives you an indication of what
could happen, and then you can put in measures to prevent or mitigate that. Although you can make assumptions as to how effective certain "inbuilt" features could be at reducing the impact, if there us uncertainty about that, you can't use that as an excuse not to put particular measures in place (precautionary principle). So I assume that as so little is known about the severity of omicron, they didn't have sufficient confidence that it would be milder, or well prevented by vaccines/ boosters so modelled it as if it wasn't milder, and reinfection/ vaccine escape was more likely.
Admittedly, in the kind of things I deal with (environmental risks) the mitigations aren't as damaging to society as lockdown etc... It's generally just a financial cost! So it does surprise me that they didn't have ANY consideration of a possible "best case" too. But from a modelling/ assessment point of view, it's not too unusual to just look at one scenario, when there is sufficient variability and uncertainty in that by itself! And, as for all viruses, the most effective way of stopping its spread is by preventing people from mixing/ meeting...so it's hardly surprising that that is what is being recommended, in a vacuum isolated from actual real life implications.
If we did have proof it was milder, it may have been modelled as being preferable to "let it rip" ( 
- hate that phrase) to burn out in the population, create immunity against omicron at least, and prevent the "kicking the can down the road" effect of lockdown. But I'm not a modeller and don't have the data, so I don't know...