A practical ontology
Techniques, positions, concepts, and systems are modeled as named knowledge nodes, then connected by typed relationships that explain how jiu-jitsu actually chains under resistance.
Jiu-Jitsu Reference
A knowledge system for understanding your jiu-jitsu.
The project combines a curated technical ontology, a public web corpus of athletes and matches, and a mobile app for logging what you actually do under live resistance.
Ontology model
Closed Guard System
32Nodes
8Edge types
4Study prompts
Triangle / Armbar Dilemma
systemA named decision structure from closed guard.
starts_fromis_part_of
Posture Broken
conceptA prerequisite condition that makes attacks available.
requires
Triangle
technicalA loggable action that can roll up into a broader family.
is_variant_ofsets_up
Product intent
Relationships are curated when they can power retrieval, recommendation, study prompts, or mobile logging context.
Mobile journal
Recent signals
LandedKnee Cuttop game
MissedArm Dragstanding
ConcededMountvulnerability
Study queuenext
Inside Positionrequires
A living reference corpus
The web app curates athletes, matches, events, organizations, and canonical knowledge so the project can connect people, competition history, and technical structure over time.
A mobile knowledge journal
The mobile app is built for fast live logging: what you hit, missed, conceded, forgot, or need to revisit, without turning training into cardio or session-volume tracking.
Ontology first
The graph is built around practitioner questions.
Jiu-jitsu is not just a list of moves. The app names useful things, separates concepts from loggable actions, and uses typed edges to explain why one node matters to another: variants, systems, prerequisites, counters, starting contexts, endings, and follow-ups.
is_variant_of
is_part_of
starts_from
ends_in
sets_up
counters
requires
related_to
Mobile app
My Game
Active repertoire and decaying techniques
Recent vulnerabilities and partner-specific patterns
Canonical and personal knowledge nodes
Offline-ready log creation for poor gym Wi-Fi
Built for the gym
Log knowledge events, not workouts.
The mobile app tracks discrete live events against resisting partners: techniques, systems, positions, misses, concessions, and notes. That gives users a memory of their game and gives the graph enough signal to explain what is active, slipping, or missing.
Roadmap
The reference layer is the foundation for richer profiles and study.
Athlete profiles with real context
Profiles can grow beyond names and bios into source-linked records, matches, organizations, instructional footprints, and technical signatures.
Study cards from graph semantics
Relationship data will power focused prompts such as options from here, follow-ups when they defend, escapes, prerequisites, and same-domain discovery.
Curated expert signals
Future knowledge pages can surface reviewed posts and trusted source material for exact techniques without becoming a generic social feed.