GameSense: Sports Prediction Platform

Overview

A platform designed to intake college football data (game results and scores, team performance and records, historical season data for testing, and current season data for predictions and rankings), rank teams based on that data, then simulate and predict the outcomes of current and future games with simple win/loss percentages based on the data provided. Tests will be against known historical outcomes.

Tech stack

The languages in use for this project are Java, Javascript, and Python. Prediction method(s) to be determined. Potential candidates include Elo, Glicko-2, TrueSkill, machine learning, custom models, and others.

Project/Development plan

The plan is to use an AGILE work method, utilizing tickets created by the product owner and assigned by the sprint master during their respective sprint. There will be four sprints in total, each with a new sprint master.

• Sprint 1: Front-end and back-end initialization, proof of concepts for final result
• Sprint 2: Framework implementation, generally implementing plans from Sprint 1
• Sprint 3: Finalizing UI and communication between it and the framework/application
• Sprint 4: Additional fixes and features, followed by a final poster board

Project Documentation

For more detail on how the project is structured and engineered, please take a glance at the following documentation directories:

• Architecture (docs/architecture/): System architecture, domain models, API contracts, database design, and technical specifications.
• Standards (docs/standards/): Naming conventions, coding rules, formatting, testing, and development guidelines.
• Decisions (docs/decisions/): Architectural Decision Records (ADRs) documenting significant technical decisions and their rationale.

[This project is a work in progress. More will be added as the project develops.]