New REACT paper is out https://www.imperial.ac.uk/media/imperial-college/institute-of-global-health-innovation/REACT1_12345_Interim-(1).pdf">https://www.imperial.ac.uk/media/imp... and is being sold as "rate of growth is slowing", and I think this is wrong and dangerous. See for example report from @JamesTGallagher here https://www.bbc.co.uk/news/health-54366478">https://www.bbc.co.uk/news/heal... (who is only reporting the study, so not blaming him).

This is the study that said before "R=1.7" and everyone panicked. They are now saying "R=1.1". I suspect they were both wrong, and it was actually more like R=1.4 each time. This is something @jamesannan has been pointing out for a while.

The R=1.7 and R=1.1 are the slopes within the little regions on graph B. But you can see how noisy the data is, and the lines don& #39;t join up in a sensible way. Whereas graph A gives a more coherent picture across time periods and has no slowing down at all. So I& #39;m Team Graph A.

Looks like they had about 40 positives out of 9000 tests on each of the last 9 days. But with that sample size, you can pretend it& #39;s a Poisson(40). So daily counts would be of the order of 40 +/- 13. The first two days were on the high side, hence low R estimate.