Thread by @Chrisbilbo

NSW R_eff as of December 18th with daily cases and restrictions. Latest estimate: R_eff = 3.35 ± 0.25

Cases shown on both a linear scale (first image) and log scale (second image).

More info chrisbillington.net/COVID_NSW.html

#covid19nsw #covidsydney

R_eff in Greater Sydney vs rest of New South Wales:

Greater Sydney: R_eff = 2.37 ± 0.27
NSW excluding Greater Sydney: R_eff = 8.55 ± 0.78

(Cases shown on a log scale)

(note region-specific data is several days old, does not include today's cases)

#covid19nsw #covidsydney

R_eff in LGAs of concern* vs the rest of NSW (*all of Penrith included):

LGAs of concern: R_eff = 1.92 ± 0.22
Rest of NSW: R_eff = 5.57 ± 0.50

(Cases shown on a log scale)

(note LGA data is several days old, does not include today's cases)

#covid19nsw #covidsydney

The Hunter region: R_eff = 16.83 ± 0.00

Illawarra region: R_eff = 2.08 ± 0.70

Western New South Wales: R_eff = 5.12 ± 2.51

(Cases shown on a log scale)

(note region-specific data is several days old, does not include today's cases)

#covid19nsw #covidsydney

Expected numbers if the current trend continues:

day cases 68% range
Sun: 3146 2782—3531
Mon: 3892 3426—4387
Tue: 4815 4208—5458
Wed: 5939 5161—6781
Thu: 7306 6313—8403
Fri: 8976 7700—10393
Sat: 10983 9356—12806

Doubling time is 2.9 days.

Note that the above trendlines are simple exponential extrapolations. In reality the curve will bend once most have been infected even if nothing else changes. An SIR model that assumes 20% of cases are detected looks like the following (log scale and linear scale shown).

SIR models are pretty simplistic too though, and should be taken with a large grain of salt. The timing of the peak might be a reasonable estimate, but the height could be dramatically different.