About This Analysis

This project is a data-driven analysis of gameplay from my game Wormhole's All the Way Down, designed to explore how telemetry can be used to evaluate and balance game systems in real time.

Purpose

The goal of this analysis is to answer a core game design question:

“Is the game challenging, fair, and engaging?”

Rather than relying purely on intuition, this project uses structured gameplay data to identify patterns in difficulty, player performance, and system behavior. The insights gathered here help guide balancing decisions such as enemy tuning, spawn rates, healing effectiveness, and pacing.

Metrics & Signals Analyzed

Each play session is broken down into both summary metrics and event-level data, allowing for both high-level trends and detailed behavioral analysis.

Key metrics include:

• Win Rate & Wave Progression
  Used to evaluate overall difficulty and identify where players struggle.
• Damage Taken vs Healing
  Helps assess survivability and whether sustain mechanics are balanced.
• Time-to-Kill (TTK)
  Measures how long enemies take to defeat, indicating whether enemies feel too weak or too tanky.
• Enemy Spawn vs Kill Rates
  Highlights whether players are being overwhelmed or maintaining control.
• Boss Entry State (HP & Lives)
  Provides insight into how punishing earlier phases are before the final challenge.
• Damage Spikes & Event Pressure
  Correlates sudden increases in player damage with special mechanics (e.g., enemy abilities), helping identify unfair or overly punishing moments.
• Pressure Score (Custom Metric)
  A composite metric combining:
  • Damage Rate
  • Enemy Density
  • Healing Rate
  This is used to quantify how “intense” the game feels over time and ensure a healthy balance between tension and recovery.

Playtesting Context

All data currently displayed was collected from solo playtesting (the developer).

Because I have full knowledge of:

• enemy behaviors
• spawn timing
• special mechanics

…this introduces a natural skill and information bias.

To compensate, the analysis is interpreted as an upper-bound (optimal play) rather than an average player experience. Target balance ranges are adjusted accordingly—for example:

• A higher expected win rate (≈80–95%)
• Stronger boss entry conditions
• Lower baseline damage intake

Data Collection Method

Gameplay data is captured using a custom-built JavaScript session recorder integrated directly into the game.

For each play session, the system records:

• Config Snapshot
  A full snapshot of game parameters at runtime (enemy stats, spawn rates, wave configurations, etc.)
  → Ensures all analysis is tied to the exact version of the game being tested.
• Event Stream
  Timestamped gameplay events such as:
  • Enemy spawns & kills
  • Player damage & healing
  • Wave transitions
• Session Summary
  Aggregated metrics computed at the end of each run for fast querying and visualization.

All data is stored in a database and visualized in real time through this dashboard, allowing for continuous iteration and balance tuning.

Why This Matters

This project demonstrates how game design decisions can be informed by data, not just intuition. By combining telemetry, visualization, and lightweight analysis, it becomes possible to:

• Identify balance issues early
• Validate design changes objectively
• Create a smoother and more engaging player experience

It also serves as a practical example of integrating:

• Frontend visualization
• Backend APIs
• Database-driven analytics
• Real-time data pipelines

into a single cohesive system.

This analysis is ongoing and evolves as more gameplay data is collected.