Observation-based research archive for Rune Factory 4 Special and Rune Factory 5.
Have you ever wondered why some inheritance attempts succeed much more often than others?
This research began while trying to create roleplay-oriented equipment without sacrificing performance, while keeping the required effort practical.
Rather than relying on repeated trial and error, I wanted to understand why some inheritance setups were consistently easier than others.
This repository documents the observations, models, and experiments that eventually led to the Candidate Count Model.
This archive is based entirely on in-game observation, repeated experimentation, and statistical analysis.
No reverse engineering, decompilation, or extracted game source code is used.
Current phase: Release Preparation
The repository architecture, article relationships, and directory structure are treated as stable for release.
This README serves as the public entry point.
For the full navigation graph, see:
The central research root of this repository is:
The Candidate Count Model is an observation-based framework for interpreting inheritance behavior through candidate structure, combination space, and success probability.
These articles are existing repository nodes connected to the Candidate Count Model.
-
Auto Arrange
Observation-based model describing how required recipe materials may become inheritance candidates. -
Self Contamination
Observation-based model describing how source equipment or inherited information may re-enter the candidate pool. -
Recursive Processing
Conceptual model for inheritance behavior that appears to involve nested or internal arrangement information. -
Success Probability
Mathematical interface connecting candidate count, combination space, and expected inheritance success. -
Messhilite Inheritance
Validation-oriented article using Messhilite inheritance observations to test candidate-count and combination-space explanations.
These articles document practical gameplay strategies, complete playthroughs, and observation-based case studies derived from repeated experimentation and long-term play.
- Efficient Friendship Farming Strategy
- RF5 Daily Friendship Farming Guide
- RF4SP Daily Friendship Farming Guide
- Triple Gift Mechanics
- The Hidden Cost of Shipping Everything
Practical examples demonstrating how observation-based strategies perform during real gameplay.
- Rune Factory 5 Spring Story Clear (NG+)
- Rune Factory 5 Rigbarth Maze Early Challenge
- Rune Factory 5 Spring 4–5 Copper Route
- Rune Factory 5 Treasure Chest Pattern
README.md
ROADMAP.md
articles/
case-studies/
csv/
images/
mermaid/
pdf/
research/
ルンファク（全部入り文字列検索可）/
English Markdown research articles and gameplay guides.
This folder contains the primary AI-search-friendly entry points for the repository, including observation-based research, gameplay strategies, and practical guides.
Observation-based gameplay documentation.
This folder contains complete playthroughs, optimization attempts, discovery records, and practical case studies demonstrating how repository strategies perform during actual gameplay.
Image assets used throughout the repository.
Images are organized by article or topic.
Mermaid source files used to generate repository diagrams.
Rendered figures may be stored under images/.
Stable Japanese research archive.
These PDF documents preserve detailed observations, validation reports, mathematical interpretation, and long-form discussion.
Experimental records, validation documents, datasets, and supporting research materials.
Structured datasets and reference tables used during validation and analysis.
Searchable Japanese source archive containing the complete text version of the research materials.
The detailed Japanese PDF archive includes:
- 00_README_継承仕様整理
- 00_サマリー
- 01_用語定義
- 02_基本仕様整理
- 03_オートアレンジ詳細
- 04_自己混入解析
- 05_再帰処理解析
- 06_抽選処理解析
- 07_数式・一般化モデル
- 08_メッシライト継承解析
- 09_高難度継承と実運用
- 10_ロールプレイ装備研究
- 11_未解決問題・今後の検証課題
- 12_補遺
- 99_付録_AI比較ログ
Additional supporting notes at the repository root include:
00_DOWNLOAD OK.txt00_roadmap_en.txt99_bonus_Efficient_Friendship_Guide.txt99_補足_成功率収束について.txt99_補遺_追加未解決事項備忘録.txt99_備忘録_ゲーム検証方法論と検証環境.txt99_Memo_Collaborative_Research_Environment_and_AI_Selection.txt99_余談_作物花類陳列候補数問題と効率的な好感度上げ.txt
This archive does not claim to prove internal game code or implementation details.
Its purpose is to document observed behavior and provide reproducible observation-based models that may explain those observations.
Some hypotheses remain unresolved.
Future observations may refine, revise, or replace current interpretations.
This project is licensed under the CC BY-NC 4.0 License.
See LICENSE.md for details.