Gunner FPS

A First-Person-Shooter (FPS) game.

The project information is divided into the following sections.

What is it?

• A scalable Unity C# simulation framework capable of modelling 10,000 synthetic players with fixed real-skill distributions, realistic multiplayer dynamics (probabilistic duels, assists, clutches) and player behaviours (smurfing).
• It allows reproducible large-scale benchmarking of different matchmaking systems.
• I have implemented the entirety of the project, using Unity Engine, C#, and Python for data analysis and visualisation.

Motivation

• This project is the final MSc coursework project, and is motivated by the lack of matchmaking algorithm benchmarking softwares.
• The project implements and assesses multiple matchmaking algorithms (Elo, Glicko, TrueSkill, a custom SmartMatch System) in a repeatable manner, providing data-driven insights into their trade-offs.

Work Done

• Implemented synthetic player behaviour, realistic match simulation, and the matchmaking algorithms (Elo, Glicko, TrueSkill, and a novel SmartMatch).
• Performed comparative analysis of fairness, rating convergence, smurf detection, and player experience across algorithms, producing data-driven insights using Python.
• Created methods for results visualisation and reporting, including graphs and statistical summaries.
• Managed UI and UX for the simulation control and data presentation.
• Created a robust and extensible software capable of simulating various matchmaking systems under identical conditions, resulting into reproducible and verifiable results.

Code Example

The coroutine (SimulateMatch) simulates a competitive match between two teams, round by round, using an Elo-based probability model for engagements. It also updates per-player statistics such as kills, assists, clutches, and adjusts Elo ratings at the end of the match.


double p1WinProb = 1.0 / (1.0 + Math.Pow(10, (p2.playerData.RealSkill - p1.playerData.RealSkill) / 400.0));
bool p1Wins = rng.NextDouble() < p1WinProb; 
if (p1Wins) 
{ 
    p1.playerData.Kills++; 
    p2.playerData.Deaths++; 
    aliveTeam2.Remove(p2); 
} 
else 
{ 
    p2.playerData.Kills++; 
    p1.playerData.Deaths++; 
    aliveTeam1.Remove(p1); 
} 

Full Method (GitHub)

This snippet shows how each duel outcome is probabilistically determined based on player skill ratings. Over many rounds, this produces realistic match results and allows for post-game Elo adjustments.

Photos

Main Screen:

Simulation Screen:

Exported CSV Data:

Verifications for Matchmaking Systems:

More details can be provided upon request.

Key Points

Gunner is a First Person Shooter game based on the training range mode of the famous FPS game called VALORANT. It is implementing some basic features of that game.

It has 6 different weapons such as assault rifles, sniper rifles, and Glock. The player can choose from these options via a "Buy" Menu and then can proceed to play any mode he wishes using the selected weapon.

The weapon models are from the Unity Asset Store but the Player Controller, Gun Behaviour, and Animations are developed in-house.

The Game also features training modes such as target training, Free Bot Shoot training, and Timed Bot Shoot training. All are derived from the "Range" mode of VALORANT.

The First-person camera controller is made in-house. The player movement is derived from the Character Controller component by Unity Engine. It is also used to implement more features such as Jump.

More details on the components and working can be provided upon request.