RationalizingOne common design principle is to put the AI on the same terms as human players. As such, whichever option is chosen should be kept compatible with the handling of player weapons. This makes implementing the interfaces much easier, and has most of the advantages of embodiment. Deciding FactorsAside from these primary concerns, the model chosen must exhibit the following properties:
As well as maintaining compatibility with the handling of human weapons, our decisions will need to promote these properties. AssumptionsThroughout Part III of this book, we'll assume that the weapon itself is mostly autonomous. This means that it can deal with reloading without the AI having to worry about it. It would be fairly trivial for the AI to check the state of the weapon and reload if necessary. This would probably be our approach if the game design allowed this. The AI can focus on aiming and firing it without concern. This makes it more convenient to focus on the difficult part of the task and not worry about the little details. Projectiles are assumed to be governed by well-determined rules, such as the physics applied to the game entities. This helps devising an AI that does not need to deal with arbitrary projectiles. Specifically, the development focuses on the ones that fly in a straight line (such as bullets or rockets). |