SleepSiphon

SleepSiphon
Author(s)	Origin
Extends	AdvancedRobot
Targeting	Virtual Guns: HOT Circular Targeting Linear Targeting Mean Targeting (Circular)
Movement	Minimum Risk Movement, Avoids: HOT Circular Targeting Linear Targeting
Current Version	1.7b
Code License	ZLib
Download

Background Information

What's special about it?: It is my first Robocode robot. Won first place in my school's competition.

Great, I want to try it. Where can I download it?: You can download SleepSiphon here!

How does it move?: It uses a version of Minimum Risk Movement. Generates destination points in a circle around itself. Calculates danger values for points in a straight line from its current location to that destination point based on a danger function. (Enemy Virtual Bullets data, distance from other enemies, distance from walls)

How does it fire?: It has four Virtual Guns: HOT, Linear Targeting, Circular Targeting, and Mean Targeting (Circular). Chooses the best gun for each enemy.; Falls back to a (naive and inefficient) Displacement Vector-based KNN (not full Dynamic Clustering) gun if the virtual hit rate against an enemy is below a threshold.

How does it dodge bullets?: Enemy Virtual Bullets are a factor in its path-point danger function.

How does the melee strategy differ from One-on-one strategy?: It doesn't, currently.

How does it select a target to attack/avoid in melee?: Attacks whoever is closest.

What does it save between rounds and matches?: Between rounds, it stores information for the Displacement Vector gun. Nothing between matches.

Where did you get the name?: This bot was an obsession of mine for a good 2 weeks... My sleep schedule suffered during that time...

Can I use your code?: Well, I won't stop you, but you really shouldn't. This is probably some of the most disorganized code you will ever read.; Any use of my code must follow the ZLib License.; * SleepSiphon Github Repository

What's next for your robot?: Slight modifications to make it more effective in RoboRumble. Because this bot started as a high school project, I built it with the assumption that the other students' bots would follow set patterns & not learn. Obviously, this is not the case in RoboRumble, so some slight changes and optimizations can be made to my prediction. After that, however, I will be moving on to new robots.

* Inspiration for debugging graphics and "path risks" from Neuromancer

* "Never closest" movement, inspired by Shadow.

* I learned a lot about avoiding aggressive bots from playing against Bastion. (not on RoboWiki)