The neverending story.

People are always whining about how short many computer games are – nowadays you can sit down for a weekend and beat most games hitting store shelves. An attempt by scientists to prevent dead-ends in games may have promise, however.

Though the afore-linked article is not clear on the specifics, it sounds as though the computer algorithms used disallow final conclusions, such that achieving a goal will instead of ramifications on future goals. In other words, as a character achieves a goal (a traditional end-game scenario, perhaps), new goals will be generated taking the manner in which the goal was completed into account.

I anticipate major shortcomings, however, if such an implementation were to see its way in a production game. For one, certain actions may very well lead to devastation in the game world. Referring to one of the article’s examples, if a character kills a key ally or destroys a key item in the game, it may set the game world on a slippery slope, such that one can predict the likelihood of certain scenarios. For instance, if a key item were necessary in securing a particular goal (destroying the Ring of Power ala Lord of the Rings, for example), and the character let the Ring fall into Sauron’s hands, though the game may go on, is playing further really worthwhile, since you can no longer win?

It sounds to me like this system would instead be better implemented in games benefitting from not having an end-goal. Take the now-popular Animal Crossing for example – by generating new characters and multiple ways to solve a puzzle, the game world remains less stale and playability goes up. Similarly, games like the oldschool Privateer, that have a strong emphasis on missions outside of the main storyline, are ripe for this kind of algorithmic study.

In the end, though, the whole idea of randomizing missions and the worldmap, extending them as time goes on, and having players effect future events, is nothing new. Perhaps these scientists in particular will bring in some great innovation, but I’d hardly call their research ground-breaking.

Tagged as: mendax.org