|
|
|
The KillerAgent is implemented using a set of simple heuristics. The
agent simply keeps track of direction, failed moves and successful moves. If
the number of consecutive failed moves exceed a predefined threshold then a
direction switch is performed. The same simple idea is used if too many moves
have been successful. However, the threshold is higher in this case. When the
agent detects food in its sensor it will immediately navigate to and eat the
food. If the agent senses other agents within its sensor range it will
prioritize an attack over any other action. If the agent itself is attacked it
will flee. This agent does not utilize the multiply feature or any teamwork
strategies.
|
|
|
|
|
| |
|
Author: Linus J. Luotsinen
|
|
