Time to talk about how amazingly good my 12 June forecast
was, plus how awesome I am in general.
It was just a very rough and ready spreadsheet model but it was superior to any other forecast produced for the UK. When others were predicting a relentless rise to over 100,000 cases a day, or over 200,000, I was pretty much a lone voice for a lower level and an end to exponential rise.
For context, cases on 12 June were at the 6,700 level and rising fast.
1) "Gives new cases rising to 60,000 a day" - I was close, cases rose to a peak of 54,700 on 17 July
2) "basically at a plateau for several months" - this is the key thing I got right and as far as I know everyone else missed. Most assumed a spike like waves one and two, rather than a long flat purgatory. What I missed was the sudden dive in the second half of July to the approx 30,000 level, but nobody had this on their radar. And we don't really understand why that dive happened. This meant the long plateau has been at the 30,000 to 40,000 level, rising recently to 40,000 to 50,000, rather than my forecast of 60,000
3) "60,000 cases in vaccinated-UK is a bad wave but very different to 60,000 in unvaccinated-UK. I think we get away with it." - when others were forecasting doom and despair, I was spot on that we had a good chance of getting away with it. There's been no disaster, no breaking of the NHS, no reversal of course back into lockdown. As I've said many times, what the govt did isn't what I'd choose and it's like speeding down a foggy motorway at night - you probably get away with it but it's not worth it.
4) "But it's 60,000 cases a day for a long long time, rather than the temporary peaks seen in waves 1 and 2." - again I predicted the key feature of the last few months, the long haul at the high case load.
5) NHS: "Temporary for a month is a very different problem to persisting for several months." - I identified the key issue for the NHS in another post, that it wasn't going to be a short lived spike with emergency measures and exhausted staff getting through the crisis then going home to rest. Instead it's been what we've seen, relentless stress on hospitals that never lets up, month after month with no prospect of any respite on the horizon.
6) "at a guess this wave would only kill 10,000 people which is nothing to Johnson." - deaths since 12 June have been 13,200. Twitter's doom gang has been wildly off with their death forecasts, even official SAGE forecasters were predicting 45,000 across the summer. I kept it in proper perspective.
The professional modelling of Covid has been pretty much a disaster. We've talked a lot about how bad it was in March 2020. It was spot on in autumn 2020 and for the disaster of "saving Christmas". But it went back to being bad in June 2021. Imperial, Warwick, LSHTM - they were all wildly off. Reality wasn't just below their central expectation, it kept way down below their likely range lower limits. It's embarrassing. There's no way a bored amateur with a couple of hours to spare on a Saturday morning and a blank spreadsheet should be beating big teams of professionals. Have they got the guts to do a post-mortem on why they got it so wrong?