> what you think the chance of a nuclear war are, for any given year. Statistics is only useful if you actually have sufficient numbers. Otherwise you can just

> what you think the chance of a nuclear war are, for any given year. Statistics is only useful if you actually have sufficient numbers. Otherwise you can just make any claim you want supported by any arbitrary model you can think of to support your preconceived idea - and let's face it, nobody will blindly create a model and then believe anything it says, instead they will tweak it until the result "sounds believable", no matter where they started.