Skip to content

A micro trading tool to predict Magic The Gathering cards prices

License

Notifications You must be signed in to change notification settings

thomasperrot/MTGTrader

Repository files navigation

GitHub forks GitHub stars GitHub license

MTGTrader

A micro trading tool to predict Magic The Gathering cards prices. 💥

Overview

  • This project crawls data from different sources and uses machine learning model to predict Magic The Gathering cards price.
  • Collected data are cards prices and last major tournaments. Price data are collected from magiccardmarket.eu, and tournament data from mtgtop8.com.
  • All data in the application can be accessed through a REST API

Environment, Architecture

This project contains two parts:

  • a Django application that contains all the logic.
  • a VueJS application to display results.

To start the hole application:

$ ./script/install_app.sh
$ ./script/restart_app.sh

Entry points

The Django application allows the user to:

  • view data in a REST API
  • manage database and schedule tasks in the Django admin interface (with django-celery-beat plugin)
  • manage queues and exchanges in RabbitMQ management interface
  • visualize monitoring dashboards about tasks (with Flower, a celery plugin). Picture is below.

Flower dashboard

Credits

This project relies on a research paper "Prediction of Price Increase for Magic: The Gathering Cards", published by Matthew Pawlicki, Joseph Polin and Jesse Zhang, from Stanford University. This paper is available in the doc folder.