Danial-Hussain / Tennis-Ranking

A tennis season simulation with a suite of ranking algorithms.

Tennis Ranking Simulation

A tennis season simulation for evaluating various ranking algorithms.

Usage

How to configure requirements and run the simulation.

Requirements

Packages used in this project can be referenced from the Project.toml file.

Run Simulation

include("../src/simulation.jl")
using .TennisRanking

simulation(
    player_count = 700,
    seasons = 1,
    number_of_replications = 1,
    burn_in_seasons = 1,
    ranking_schemes = Vector{String}(["ATP"]),
    metrics = Vector{String}(["Spearman Correlation"]),
    verbose = true,
    strength_method = "BT"
)

Options

Arguments
=========
- `player_count`:
    the number of players to create.
- `seasons`:
    the number of seasons to simulate.
- `number_of_replications`:
    the number of times to repeat the simulation.
- `burn_in_seasons`:
    the number of burn in seasons to use.
- `ranking_schemes`:
    which ranking algorithms to test.
    options:
        1) "ATP"
        2) "Elo"
        3) "Dwz"
- `metrics`:
    the metrics to be evaluated during the simulation.
    options:
        1) Spearman Correlation
        2) Kendall Tau Correlation
- `verbose`:
    whether to print each stage of the simulation.
    options:
        1) true
        2) false
- `strength_method`:
    method to determine the probability team 1 beats team 2.
    options:
        1) "BT" - Bradley Terry Model
        2) "BT_Exp" - Bradley Terry Exponential Model