PityLine

Game developers often struggle with balancing drop rates, leading to player frustration during long dry streaks. This skill simulates loot rolls to demonstrate exactly how a pity counter stabilizes outcomes without changing the underlying math of the drop table.

What it does

• Loot Simulation generates twenty consecutive kills based on an existing drop table and a specific seed.
• Pity Comparison runs the same simulation side-by-side with a luck protection counter.
• Rare Item Targeting automatically identifies the rarest item or uses a user-defined target for the pity mechanic.
• Forced Drop Logic replaces dry kills with the rare item once the defined threshold is met.
• Data Export writes the simulation results to pity-log.md and structured JSON for further balancing analysis.

How it works

1. Provide an existing drop table with item names and weights.
2. Define an optional pity threshold, target rare, and seed.
3. The skill executes a Python-based roll script to generate two sets of results.
4. Review the comparison report showing raw vs. pitied rare frequencies and dry streaks.

Frameworks & tools

The skill utilizes a python3 simulation script (scripts/roll.py) to handle weighted RNG and writes state to Markdown and JSON files.

Why this beats prompting it yourself

Manual prompting often results in "hallucinated" randomness where the LLM simply picks items it likes. This skill uses a deterministic script to ensure the RNG is consistent across both test groups, providing valid data for game balance.

Use cases

• Testing the feel of a 1-in-50 drop with an 8-kill pity floor.
• Generating documentation for players to explain how luck protection works.
• Exporting raw loot data to JSON for integration into game engine testing.
• Visualizing dry streaks to decide on fair mercy mechanics.

Known limitations

Requires a table with at least two unique item names. Does not support affix generation, item renaming, or creating entirely new tables from scratch.