This repository contains the source code used for the experiments described in the paper "Shallow decision-making analysis in General Video Game Playing" by Ivan Bravi, Jialin Liu, Diego Perez-Liebana and Simon Lucas.
The GVGAI code that implements the Shadowing Agent is coming soon to the repo! It will also contain the sample agents described in the paper except the Genetic Programming MCTS agent.
The plots are relative to the five games described in the paper: : Aliens, Brainman, Camelrace, Racebet2, Zenpuzzle; On the y axis you can find the main agents (the ones actually playing the game) while on the x axis there are the shadow agents. The label of the main agent also indicates between squared brackets the win percentage of that agent.
- Aliens: a game loosely modelled on the Atari 2600’s Space Invaders, the agent on the bottom of the screen has to shoot the incoming alien spaceships from above avoiding their blasts;
- Brainman: the objective of the game is for the player to reach the exit, the player can collect diamonds to get points and push keys into doors to open them;
- Camelrace: the player, controlling a camel, has to reach the finish line before the other camels whose behaviour is part of the design of the game;
- Racebet2: in the game there are few camels racing toward the finish line, each has a unique colour, in order to win the game the agent has to position the avatar on the camel with a specific colour;
- Zenpuzzle: the level has two different types of floor tiles, one that can be always stepped on and a special type that can be stepped on no more than once. The agent has to step on all the special tiles in order to win the game;
The paper provides a single example for the Pure Agreement plot, in the folder plots/PureAgreement you can find the plots for all the games described in the paper. The brighter the colour the higher the agreement while the darker the lower.
Similarly to the previous kind of plots, the paper contains just an example of Decision Similarity plot, all the plots can be found in the folder plots/DecisionSimilarity. In the plots the darker the colour the higher the probability of selecting actions is similar, instead, the darker the more different. The measure uses the symmetric KL-Divergence metric.