[conspire] Correction

[Nick Moffitt]

On 24Mar2020 12:53am (-0700), Rick Moen wrote:
> It's really rather academic whether the confirmed case number is 1900 or
> 2300 or 3800, etc.  The point is:

The confirmed cases metric is really hamstrung by our inconsistent testing regimen.  I don't trust those numbers, and agree they're only useful for comparing rate of change against the deaths figures (which are more accurately recorded, even if post-mortem testing is also questionable).

> Meanwhile, the best estimates I've been hearing of the doubling time is
> about three days (absent effective suppression techniques such as 
> consistent social distancing and the rest).
> 
> Exponential.  Growth.
> Three.  Days.
> 
> Concentrate on that.  And on the fact that reported confirmed case
> numbers are _inherently_ a drastic undercount.

The Financial Times here in London (I tend to cycle past their offices between my office and Borough Market to get lunch/coffee grounds) have finally opened up their data analysis page to the public without paywall:

	https://www.ft.com/coronavirus-latest

They changed the dashed lines to no longer show the "33% increase in deaths per day" to marking out doubling times in days, which is a more useful set of landmarks.

The stars on those graphs also indicate when a nation called an official lockdown/quarantine order.  The UK, like NZ, announced a few days in advance: we don't have anything taking legal effect until Friday.