Designing the WH40k Tournament Analyzer
How I'm approaching the architecture of a Warhammer 40k competitive meta analysis tool, from data ingestion to statistical modeling.
One of the projects I’ve been working on is a Warhammer 40k Tournament Analyzer. The goal is simple: use real tournament data to understand what’s actually working in the competitive meta, rather than relying on gut feelings or forum opinions.
The Problem
Warhammer 40k is a game with enormous complexity. Hundreds of units, countless list builds, multiple mission types, and a meta that shifts with every balance patch. How do you figure out what’s “good” when the answer changes depending on the mission, the matchup, and the player?
Data Source: Best Coast Pairing
The foundation is data from Best Coast Pairing, a platform that records online tournaments. From it, I can extract:
- Win/loss records
- Points scored
- Missions played
- Opponents faced
- Full list composition
- Results per mission
Mission-Based Role Classification
One of the key design decisions is classifying performance by mission type:
- Take and Hold - Control objectives
- Purge the Foe - Kill-based scoring
- Disruption - Interference and denial
- Reconnaissance - Scouting and intel
- Priority Assets - Objective control
Each army performs differently depending on the mission. A list that dominates in Purge the Foe might struggle in Reconnaissance. Tracking this separately gives a much richer picture than a single win rate.
The Architecture Challenge
The hardest part is data aggregation. In a single tournament, an army might play 5 rounds across 3 different mission types. How do you aggregate that data meaningfully?
Some questions I’m still working through:
- Should win rate be weighted by opponent strength?
- How do you measure consistency vs. ceiling?
- What does “optimal” even mean in this context?
These are the kinds of problems I enjoy solving. More updates coming as the project progresses.