the number of cases already and the transmission rate (this ignores that eventually you run out of people to infect; i.e I'm using exponential growth whereas reality is logistic, but logistic looks like exponential growth early on).

Let's say the formula for new cases on a given day t is dC(t) = c*r, where c is the total number of cases yesterday and r is the rate of transmission. In the case that over some number of days the number of cases has gone up by 10 times, then even with a transmission rate cut by 90%, you'd have the same number of new cases after the transmission rate cut.

So taking those numbers as an example
dC(t) = X*r
dC(t+some number of days) = 10X*.1r = X*r

In our case, I'm not sure we have cut the transmission rate by 90%, but we do have roughly 10x the cases in the past 10 days.