@kmic The very article you linked serves as a great example of the contrary - inoculation did work and it was observable with even early statistical methods.