MbShea@PWP@English@UMCP

Audience:
This document will be written for individuals with some level of statistical or data science experience. Expertise is not required, but some experience in these fields would be ideal.

Context:
The sports betting market is huge, and very lucrative. Sports betting companies, such as those based in Vegas, bring in millions in revenue off of sports, without any expenditure on the sports themselves. There are many different ways to bet, and one of the most popular is betting against the spread, the predicted score difference by Vegas. However, Vegas changes the spread and the payout rates as more people bet. Because of this, one could take advantage of statistical and machine learning models to predict ideal betting situations and make long term profit.

Purpose:
The purpose of the document is both to inform about the situation and to teach potential methods that can be used in this project, and other predictive projects. The document will review how the sports betting world works and where the ideal points of interest for predictive betting are. Then, the document will explain and show how to use data gathered from online sources. The document will then fit various models to the data, using both machine learning and classic statistical methods to make predictions.

Document Type:
The document will be a Jupyter Notebook, in the form of a mix of commented code and text. This will be ideal in allowing both explanation and display of processes methods used.

Design/format:
The document will begin with a background on sports betting, highlighting the different forms of betting, and which type we will be targeting. From there, we will walk the reader through the process of gathering data and determining what data to use. Next, the project will show how to fit statistical models to the data, and how to find machine learning methods to use. The document will apply these methods and show how to test how well the data fits use cross validation. The document will then end with suggestions on how to use these models, or how to use the entire process for other pursuits of prediction.

Citation style:
There will be some formal APA citations for the background information. For the code, imported libraries will show where the methods were taken from, and comments will serve as natural language citations.

Looks fine, ME, and I am pleased you will write a "bilingual" document. I might a cover note to me, to make sure we meet written prose requirements for the program.

I will need to be part of an email with Dr D, though. So let's make sure we do that.

Mb