It's just statistics. You look at how often we see GRBs, how long they last (on average), how intense they are (on average) then do a big time integral and figure out the odds of a big one hitting earth over such and such a period of time. There is no better than a coin-flip, the coin flip is the odds of it having happened (by the author's approximations). If I can help explain any further please let me know.