Deep Neural Network Model for Detection of Cheating in Competitive Chess

Overview

Modern computer chess programs capable of easily beating any human player are causing a crisis in the world of competitive chess. These programs are simple to use and can be run on personal computers and even smartphones, allowing users to immediately find the best move from any position in a chess game. This has led to a significant increase in cheating in both online and over-the-board chess matches, resulting in player bans, multi-million-dollar lawsuits, and even criminal charges, but it is hard to prove that a player used assistance from a computer to cheat unless the player is caught in the act. Currently the best mathematical evidence for cheating comes from using statistics to determine the chance that a player could perform above their skill level for many games in a row, but this is not hard proof because 1) a player may play extremely well from time to time without cheating, and 2) a smart cheater might deliberately lose a game occasionally to fool the statistics. For this project I built and trained a neural network to find patterns that cheaters use. I analyzed hundreds of real grandmaster games and used the Stockfish chess engine to generate a large number of games used to train a neural network built with the TensorFlow python library. Moves were chosen to match the statistics of real players and known cheaters. My trained network can detect cheating within a single game with high accuracy.

Deep Neural Network Model for Detection of Cheating in Competitive Chess

Overview

Source coverage

Awards (1)

Competition history

Resources

Related projects

A Novel and Convenient Go-Games Win-Rate Estimation and Misplay Detection Model Based on Convolutional Neural Networks

A Novel Approach to Detecting Academic Dishonesty Involving Artificial Intelligence

A Deep Neural Network Method for Steganographic Image Detection

Building a Deep Neural Network to Automate Protection Against Online Cyber Attacks

Deepfake Detection Using Deep Learning

CHESSTER- The Chess Playing Robot

Creating a Neural Network to Play the Game of Connect Four Using a Genetic Algorithm

Detection of AI-Generated Music Using Convolutional Neural Networks

Browse more like this