What we know about the coronavirus, called 2019 nCoV, changes hour by hour, making it hard to predict how far the virus might spread and how many people could become infected.
One Chinese government expert, Zhong Nanshan, says the epidemic may peak towards the second week of February.
But academics from the University of Hong Kong, think that's more likely to happen sometime in April or May.
There is so much uncertainty about this new virus, that it's almost impossible to make accurate predictions.
First of all, we don't know how many people have the virus right now.
Reported cases have surged in the last few days, but that could be down to increased surveillance and testing.
There could still be thousands who aren't coming forward, either because they have mild symptoms or because they feel safer waiting it out at home.
Next, we don't know how easily the virus spreads.
On average, it looks as though each infected person has the ability to infect two or three more.
This is known as the reproduction number, and it's significant.
Up until now, the number of infections is thought to have doubled every six days.
That gives us some really big numbers.
Mass quarantines, the extended public holiday, plus putting the public on high alert may slow the spread of disease from person to person, but it's hard to know how effective these measures will be because of another big unknown.
We don't yet know how long it takes for an infected person to show symptoms, nor whether that person can pass the disease on before showing any signs of illness.
This raises doubts over the value of temperature screening, which has been a key control so far.
Finally, we don't know how severe the virus is.
It seems to be spreading faster and wider than SARS, which infected 8,000 people over eight months.
This coronavirus has gone past that number in less than two months.
Around 20 % of patients progressed from fever and a dry cough to more severe symptoms such as pneumonia and organ failure with around 4% dying.
But at the start of an outbreak serious cases tend to appear first—like the tip of an iceberg.
So the rate of severe illness may drop as milder cases turn up.
In SARS, the death rate was about 10%.
But even if this new virus has a lower mortality rate, it could still kill more people because it causes more infections.
And while it's spreading, it might also mutate to become more or less virulent.
We are better prepared than we were in 2003 during the SARS epidemic, but this still looks like a monumental global health challenge.
And until we know more about the virus itself, its final impact will be hard to predict.