NPC Target Selection Logic for Natural Action Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing games where NPCs target objects based on distance result in unnatural behavior, as objects at similar distances are frequently selected, undermining the creature-like impression.
Innovation Solution
A game program that determines candidate objects within close and far ranges, varying action targets over time, and preferentially selecting objects based on elapsed time, proximity, and visibility, while limiting repeated actions on the same object.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the target of action is determined according to the magnitudes of the distances between the NPC and the other objects, then the selection process is simple and efficient, but objects at similar distances are successively selected as targets, resulting in unnatural behavior
Solution Approach 1:
The patent changes the selection parameters from purely distance-based to a multi-parameter system including distance, attention level, and elapsed time. This allows the NPC to select targets more naturally by considering multiple factors rather than just distance magnitude.
Solution Approach 2:
The patent introduces dynamic elements by making the attention level change over time and using elapsed time as a selection criterion. This dynamic approach prevents static distance-based selection from repeatedly choosing the same objects, thereby improving naturalness.
2Stability of the object's composition
If the NPC performs action on a first object and decreases its attention level to prevent frequent targeting, then the distribution of actions becomes more even, but the target selection still concentrates on objects within certain distance ranges
Solution Approach 1:
The patent implements feedback mechanisms where the attention level of objects is adjusted based on previous targeting, and this attention level feeds back into the selection process. Combined with elapsed time tracking, this feedback system ensures both even distribution and natural behavior by preventing both over-targeting and range concentration.
Data Source
AI summary
When there are a first number of first candidate objects in a first range close to a first object and there are a second number of second candidate objects in a second range far from the first object, the first object is caused to perform an action in one of (1) a first manner in which, before the action is performed with the second candidate object as a target, the action is performed with the first candidate object as a target a number of times that is equal to or less than the first number and (2) a second manner in which, before the action is performed with the first candidate object as a target, the action is performed with the second candidate object as a target a number of times that is equal to or less than the second number.


