Well, the general reasoning you applied is in fact the basis of several 'plot-based' approaches; a sequence of 'events' that change the narrative world in a cer

Well, the general reasoning you applied is in fact the basis of several 'plot-based' approaches; a sequence of 'events' that change the narrative world in a certain way is chosen based on (any number of) criteria I worked particularly on a system that involved an automated planning paradigm - so there were 'goals' that the story wanted to be met (and could be determined/influenced by an author-user) and a catalogue of abstract definitions of events as planning operators. It worked somewhat like you said, except (for practical purposes) mostly backwards (as in backwards-planning) I unfortunately lost contact with the research area a bit, and am not familiar with how it progressed - but it certainly is interesting in a distinct way; it really lets you explore a particular creative side that many things in computer science neglect