Quote Originally Posted by CCFCC3PO View Post
True, and small changes in the model will have huge implications in later projections. There's lots of models, I'm wondering why they chose to go with this one.
Because it came out yesterday, I posted about it in the main thread and spotted the same issue (ICU bed estimates).

As for judging that model by our daily figures, what we see are registered deaths with family consent who were hospitalised at the time of death. There could be a time lag and we may not be seeing anything like accurate numbers for total deaths.

The ONS report is going to be next to useless because we werent/aren't testing in care homes. There was a sharp increase in overall deaths recorded for the week ending 27th march. The ONS only attributed 20/~500 of these to covid because their death certificate said covid was the cause of death. The actual number caused by covid but undiagnosed could be way higher and really it is pretty impossible to know (because we arent testing widely).