H1N1 in 2009 may fit well with the ‘antigenic shift’ paradigm of influenza, where there is a reassortment of genes from different types of flu (which might both infect the same animal at once) that create a new hybrid to which there is little pre-existing immunity and which can be markedly virulent. This is classically in contrast with year to year ‘antigenic drift’ of the flu featuring small accumulated mutations that have tended to be less virulent. Both of these fit well with the random probability idea to some extent, but are very different mechanisms. I don’t know if people have tried to induce antigenic shift in a lab; probably not the safest idea.