https://github.com/jef/2048-agent-based-solver

Monte Carlo Sim for 2048
https://github.com/jef/2048-agent-based-solver

Last synced: 10 months ago
JSON representation

Monte Carlo Sim for 2048

• Host: GitHub
• URL: https://github.com/jef/2048-agent-based-solver
• Owner: jef
• License: mit
• Created: 2015-11-03T21:18:35.000Z (over 10 years ago)
• Default Branch: master
• Last Pushed: 2020-03-05T06:51:45.000Z (over 6 years ago)
• Last Synced: 2025-02-24T08:45:40.131Z (over 1 year ago)
• Language: JavaScript
• Homepage: https://jef.buzz/2048-Agent-Based-Solver/
• Size: 293 KB
• Stars: 0
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE.txt

Awesome Lists containing this project

README

          
2048: An Agent Based Solver

=======================

The popular web game 2048 (https://github.com/gabrielecirulli/2048)

(designed by Gabriele Cirulli) is a game in which a single player plays against the randomness of the game.

Task at hand

------------

Creating an AI that can play the game well enough to reliably win.

 If the AI is winning then I will try to optimize it to get as high a score as possible.

Specific to the AI Agent

------------------------

It must select its move within one sixtieth of a second

(that is, it must make 60 moves a second or more). Since the game is random we will test the agent a

 number of times to collect a set of scores. All agents capable of winning the game will be submitted

 to a friendly in­class competition for highest score.