[viraptor]

I couldn't find more historic data points, but extrapolating from 25 since May to 150 per year is a bad approach. There's a difference between: it's been always happening, and this is the most recent summary -vs- April 0, May 1, June 4, July 20.

[tomp]

What would be a better approach, inabsence of other data points?

[viraptor]

Don't do analysis which assumes you can extrapolate.