Back to projects
LeadPersonal2024

Dota 2 Draft Assistant

Java Desktop App with ML & Steam API

Technologies

JavaJavaFXPostgreSQLSteam APIMachine LearningOAuth

A desktop application for Dota 2 players that provides data-driven draft recommendations using machine learning and real-time player statistics.

Technical Implementation

- **Steam API Integration:** OAuth authentication with Steam, player match history retrieval, rate limiting handling

  • **PostgreSQL Database:** Hero statistics, match data, and synergy calculations stored and indexed for fast retrieval
  • **ML Model:** Win probability prediction based on team compositions using historical match data
  • **JavaFX UI:** Responsive desktop interface with real-time updates during draft phase

Features

- Player lookup via Steam ID - Hero synergy and counter-pick analysis - Real-time win probability visualization - Historical performance tracking

Key Highlights

  • Integrated Steam API with OAuth authentication
  • Built ML model for win probability prediction
  • Designed PostgreSQL schema for fast analytics queries

Other Projects