Read on Twitter
oh cool the graph that they showed for Spain was the ECDC one by date of report (which shows an increase consistently) not by date of symptoms (which is more or less completely flat - ignore the dropoff at the end)
not like anything important like "the UK's entire pandemic policy" is resting on this complete clownshow
we even wrote a piece last week about the Spanish data showing how to properly present it, accounting for lags etc, what it's showing. but no fine just pick up whatever data and make decisions on that, no point giving it 30 seconds' thought https://www.cebm.net/covid-19/does-the-resurgence-of-covid-19-in-spain-exhibit-exponential-growth/
I'm sorry but this is a total and utter joke. We're six months into this and the most *basic possible stuff* is being overlooked.
[INCHOATE SCREAMING]
Best of all - the French graph is subject to *exactly the same problem* but there isn't data by date of specimen/symptoms that allows us to get a useful picture. Just completely useless.
To be clear this is *exactly the same data* presented in two different ways - one (the way the government used it) completely useless and the other (the way we used it) actually informative. I cannot get my head around this sort of ridiculous, stupid error.
This utter farce is now written up https://www.cebm.net/covid-19/epidemic-curves-organised-by-reporting-date-and-specimen-date/
In short: the way the Spanish data was presented today was really, really unhelpful because it did not use the date of symptoms but the date of reporting. If you do it properly you get a flattening off over time not apparent in the way they presented it.
Contra what people were telling me earlier, this is not just due to a lag in reporting. We applied a 14 day lag in the original piece we did and it was still there on September 3 https://www.cebm.net/covid-19/does-the-resurgence-of-covid-19-in-spain-exhibit-exponential-growth/
