UPE

An innovative prediction engine for ultimate frisbee

Project Overview

The UPE - the Ultimate Prediction Engine is a web-based application built around a prediction engine for the AUDL (American Ultimate Disc League). Predictions take into account historical data of teams, players, and games along with historical and forecasted weather data. Driven by the weather-dependent nature of the sport and the lack of any interactive platforms for the league, we wanted to create a platform for fans of the sport to get clear and accurate predictions for both upcoming and fantasy matchups.

The project was built with React.js, Node.js, Flask, Scikit-learn, MySQL, and Python. We scraped around 75,000 data entries from two sources: the AUDL’s backend and the Visual Crossing Weather API. All of this was then deployed to an AWS EC2 instance.

Project Materials

The Github repository and all code for the project can be found here.

Writing Assignments

Project Proposal

Available Here

Writing 1: Executive Summary Elevator Pitch

This writing assignment was completed individually and differs for each person.

Writing 1 is included as one of the first parts of writing 4.

Writing 2: Technical Summary

This writing assignment was completed individually and differs for each person.

Writing 2 is included as one of the first parts of writing 4.

Writing 3: Project Description

Available Here

Writing 4: Technical Design Document

Available Here

Presentations

Presentation 1: Project Elevator Pitch

Available Here

Presentation 2: Project Design

Available Here

Presentation 3: Demo 1 - Alpha Prototype

Available Here

Presentation 4: Commercial Presentation

Available Here

Presentation 5: Final Presentation

Available Here

Videos

Promotional Video

Available here.

Project Video

Available here.

Poster

Available Here

Indiviudal Slides Available Here