Faculty Advisor: Dr. Andrew Penland
Authors: Daniel Hammer, Nicholas O'Kelley, Andrew Shelton
Our research focus is on the applications of evolutionary algorithms in game theory as applied Two Player Games in search of finding optimal winning strategies for each of games. Once we have substantial data, we plan to continually analyze the winning trend over the period of games to see if there are any game agnostic strategies.
A more detailed write up of our current progress can be found in the
simple_games
directory in the Jupyter Notebook called Overview.ipynb
.
Pass meetings notes / minutes can be found in the meeting_minutes/
directory.
A readme is supplied to organize the different dates.
We didn't always keep digital notes, so as they are converted those will be
added. New notes are added in markdown files or jupyter notebooks in the
notes/
directory.
The first kind of two-player games we will consider work as follows:
Chess violates Rule 3: I can not move your pieces, so I don't have the same moves as you.
Two-player graph coloring violates Rule 4. If the second player has no legal moves available, they still win (as long as the graph is not colored).
Nevertheless, this class of games makes a good starting point for our analysis.
Returning to the Graph coloring problem, we began to implement the game following our current game structure and rule definitions. There was one issue: CGC failed to meet all the rules. In our meeting on February 19th 2021, we decided to have our prior rule definitions be the summary of our focus for the Fall semester (2020) and the Spring (2021) focus on a slightly modified rule structure:
This new rule set allows for the Competitive Graph Coloring Game to be added for analysis since there is a win condition for a stalemate.
The big ideas for this section include:
Have a feature that you want to see added? Then fork this repository, craft the feature, and then open a pull request and we will be in contact with feedback!
Keep in mind: