Some bad news and some good news.
Bad news first: I've tried to calculate the R0 (R-nought) of the cruise ship. R0 is the number of people that an infected person
can infect infects until they have recovered to a point where they are no longer infectious. It's difficult to estimate R0 accurately given the data we have, so when I looked at the result I got, I thought I'd made a mistake. Also I haven't done any sensitivity analysis on my models yet. However, I then found this,
not peer reviewed article, and it seems we have used basically the same method coming to basically the same results. (Nasty equations ahead):
Background Since the first cluster of cases was identified in Wuhan City, China, in December, 2019, 2019–nCoV has rapidly spread across China as well as caused multiple introductions in 25 countries as of February, 2020. Despite the scarcity of publicly available data, scientists around the...
www.medrxiv.org
TL/DR: The article mentions a R0 of 7.05 (95%CrI: 6.11–8.18). That's close to being as contagious as measles (R0 = ~12-18), and much higher than WHOs estimate of ~2. I also agree with the authors that even with the current lockdown of Wuhan and Hubei, R0 is still around 3. It does seem to have dropped to ~1.5 the last couple of days, but that is after the large number of "new" cases added at February 12th. All in all it seems that this virus is so contagious that it's going to be extremely hard to contain, and secondly that it will spread fast.
GOOD NEWS: The article also reaches the conclusion, that the CFR estimates at ~2% are too high. This is due to a large number of unreported cases in China. It seems to me, that most of the cases reported, and there are many, are more or less severe cases, whereas the milder cases are not being registered in the official numbers. WHO kind of insists on a lab test to include it in their numbers, even though they have started to mention clinical tests in their daily reports. Instead of CRF the authors calculate IFR, which is based on an estimate of infected people (~1 million), and they get a IFR of somewhere between 0.05%-0.30% depending on time compensation (time from infection to death). My model shows an IFR of 0.4%, but I have only estimated ~0.5 million infected, meaning that if I use 1 million instead I get 0.2%.
Summing up, it seems that the virus spreads faster than previously estimated, but that it's less deadly than we first feared. I still don't see a way out of a pandemic, and probably sooner than we think, but it will hopefully be less lethal than we thought.
I'm neither a doctor, nor an epidemiologist. I know how to model, and read as much as I could find about modelling an epidemic (SIR), so this is just sort of complicated back of the envelope calculations, and I might very well be wrong.