This section details the technical structure of the discovery system and the content curation methodology. The design addresses two problems from the research: catalog overwhelm creates discovery paralysis, and users want to build genre expertise but lack structured guidance.

The system uses a Landmarks-to-Exploration architecture where users validate baseline literacy before accessing curated sub-genre regions. Content is organized through hub-and-spoke connection density that creates natural progression from accessible to advanced shows.

Decision Choice Rationale
Architecture Single expanded atlas Simpler mental model than tiered systems
Curation approach ML-first with editorial validation Reduces bias and scales efficiently
Launch scope 6 genres (not phased rollout) Complete coverage prevents “doesn’t work for me” reactions
Depth ~107 shows per genre (~640 total) Balances depth with curation feasibility

System Architecture: Landmarks + Exploration Layers

The Anime Atlas consists of two distinct layers that serve different validation and exploration functions.

Attribute Landmarks Layer Exploration Layer
Shows 7 per genre 100 per genre (4 regions × ~25)
Purpose Validate baseline genre literacy Structured discovery through thematic regions
Selection Culturally significant shows representing all regions Organized by connection density (gateway → terra incognita)
Time investment 75-90 min minimum (3 shows), 175 min for all 7 Ongoing — natural depth through hub-and-spoke
Design rationale Prevents gatekeeping while confirming genre context Single continuous atlas maintains simple mental model

Landmarks Layer Design

Landmarks shows must satisfy a critical constraint: collectively map to all 4 exploration regions to enable thematic routing. This “reverse logic requirement” emerged from testing region unlock mechanics ’…if Landmarks lacks representation for a region, that region becomes orphaned regardless of user’s validation path.

Landmarks Show Selection Criteria

Criterion Requirement Purpose
Cultural significance Top 5-10% MAL popularity OR defining moment in genre history Validates shows users likely already know
Region coverage Each of 4 regions has 1 Landmarks show with primary affinity Ensures thematic routing works for all user paths
Episode 1 accessibility Strong pilot episode with clear hook Supports 1-episode Chartable threshold
Temporal range Spans 2000s classics → 2020s hits Teaches genre evolution
Already-watched value HIGH for 3 shows, MEDIUM for remaining Power users auto-validate, don’t feel penalized

Action Genre Landmarks Example

Show Year Primary Region Landmarks Role Already-Watched Probability
Naruto 2002-2007 Battle Shounen 2000s battle pillar HIGH (millennials/Gen Z)
One Piece 1999-present Battle Shounen Adventure archetype HIGH (cultural touchstone)
Attack on Titan 2013-2023 Dark & Gritty Psychological horror anchor MEDIUM-HIGH (2010s defining)
FMA: Brotherhood 2009-2010 Dark & Gritty Philosophical depth MEDIUM-HIGH (universally acclaimed)
My Hero Academia 2016-present Battle Shounen Contemporary entry point MEDIUM (mainstream appeal)
Demon Slayer 2019-present Battle Shounen + Historical Emotional accessibility MEDIUM (recent phenomenon)
Mob Psycho 100 2016-2019 Over the Top Action-comedy hybrid, emotional depth MEDIUM (ONE creator crossover)

Validation Threshold: 3 shows watched (1 episode each) = minimum. 4-6 shows recommended for fuller context.

Rationale: 6 shows created an orphaned region (Over the Top had no primary Landmarks anchor); 7 ensures each region has a thematic entry point; still manageable time investment (140-175 minutes); “Already watched” auto-validation means power users complete instantly.


Landmarks → Region Unlock Logic

Landmarks shows contain thematic preference signals that route users to relevant regions without being prescriptive about individual show selection.

Landmarks Structural Connections