Fb Betting Graph

(credit to @notadamking, this repo and this article)

This repository contains a Jupyter Notebook (source/OHCL FB Bet Graph.ipynb) with a custom class FootballBettingGraph that visualises bets (balance, profit, volumes, individual bets).

The idea is similar to what Stock Traders would use to visualise market movements, applied to Football Betting. This can be used to visualise the training process of a model and how it makes decisions when to bet. The class can also be adopted to plot bets real-time.

Includes some test data (source/test_data.xlsx) and an example on how to run the graph using that data.

Installing

Clone the repo, install dependencies, open OHCL FB Bet Graph.ipynb using Jupyter Notebook (or something else) and run it.

Requirements

See requirements.txt.

Improvements

• _render_balance can use plot_date to create a more dynamic visualisation of bets. However, this would tie the implementation down to using dates
• Profit column is redundant as it can be calculated from the opening and closing values
• in volumes, display more details about the type of bets (win? lose? something else?)