Soccer-Game-AI

Autonomous 2D top-down soccer simulation built in C++ with SDL2, inspired by the architecture from Programming Game AI by Example.

This project demonstrates coordinated multi-agent behavior, team-state logic, ball physics, and a complete playable match loop (kickoff, goals, score tracking, winner detection).

Match Preview

game_preview.mp4

What This Project Demonstrates

• Autonomous gameplay with no player input required.
• Team-level and player-level decision flow using finite state machines.
• Real-time role transitions (chase, support, receive, return home, defend/attack switching).
• Ball possession and scoring lifecycle (goal detection, reset, kickoff preparation).
• Visual game presentation with a rendered pitch, team entities, keeper behavior, and top-center HUD scoreboard.

Core Systems

AI and Decision Model

• Team FSM: PrepareForKickOff, Attacking, Defending.
• Player FSM: Wait, ReturnToHomeRegion, ChaseBall, Dribble, ReceiveBall, SupportAttacker.
• Role selection: Closest-player-to-ball updates each frame to drive who pressures the ball.

Physics and Movement

• Vector-based movement with steering behaviors (seek, arrive, pursuit).
• Ball movement with friction and velocity decay.
• Post-update overlap resolution to prevent players and keepers from visually passing through each other.

Match Flow

• Real-time score tracking (Red vs Blue).
• Automatic reset after non-winning goals.
• Game freeze when a team reaches winning score.
• Winner highlighting in the HUD.

Visual and Audio Layer

• Procedural pitch rendering (gradient grass, stripes, center line/circle, boxes, goal visuals).
• Stylized procedural players/keepers/ball for clear tactical readability.
• Optional looping crowd ambience from WAV file.

Project Structure

include/   Headers (entities, states, math, FSM, messaging)
src/       C++ implementation (game loop, pitch, teams, players, ball, rendering)
imgs/      Optional image assets
screenshots/

Build and Run

Requirements

• g++ with C++17 support
• SDL2 development libraries
• SDL2_image development libraries

Commands

make
make run

Current Scope

This repository focuses on simulation architecture, autonomous flow, and readability of game AI behavior. It is intentionally lightweight and suitable as a foundation for:

• richer tactical logic,
• stronger passing/shooting heuristics,
• animation and polish,
• and deeper match analytics.