A Genre System Built for Discovery
To make board games intuitively discoverable on Parlourite, we organized the field into 21 genres across 5 families. A record of borrowing established vocabulary and recalibrating the granularity for platform-level discovery.
Parlourite is a board game platform, which means the game you want to play has to be findable. “A mid-weight area majority game.” “A cooperative puzzle.” Entry points like these should lead to candidates without friction, and for that to work the platform needs an internal taxonomy with the right shape. We settled on 21 genres grouped into 5 families. Names follow established usage; granularity has been tuned for discovery on the platform.
The reference, and our use of it
Mechanism classification for board games has a long history. The Mechanism list on BoardGameGeek is a well-known reference, with roughly 200 entries and considerable depth. For design discussion and critical writing, that granularity and breadth are genuinely valuable.
Using it directly as the filter facet of a discovery UI is a different problem. Overlap between items and uneven granularity are natural in a reference work but awkward in a sidebar: too many choices, too much apparent duplication. The unit a player wants when drilling into “things similar to this” is coarser and cleaner. So we recalibrated the granularity separately, specifically for discovery.
Design principles
Two principles, kept short.
First, genre and expression form are separate. Roll-and-Write and Legacy / Campaign don’t require any specific core mechanism. They ride on top of push-your-luck, placement, cooperative, almost anything. Listing them as genres would double-count. We place them on an orthogonal axis.
Second, out-of-scope domains are declared explicitly. Dexterity games where physical skill is the core, tabletop RPGs where a GM’s judgment decides outcomes, storytelling games where subjective evaluation drives victory, word games where natural-language dictionary adjudication is essential. All compelling in their own right, none within Parlourite’s scope.
Five families
The 21 genres fall into 5 families, grouped by what each treats as the center of game state.
| Family | Center of state | Genres |
|---|---|---|
| Abstract | Position on the board | 5 |
| Spatial Placement | Shapes brought from outside the board | 2 |
| Card & Hand | Hidden-information holdings | 5 |
| Decision & Economy | Decision procedures themselves | 6 |
| Social & Solo | Relationships among players | 3 |
Abstract covers placing, moving, and resolving pieces on a board by line, connection, territory, capture, or progress. Chess, Go, and Backgammon all belong here. Spatial Placement brings tiles or pieces from outside the board and generally does not move them once placed. Card & Hand splits on how cards are played or acquired, from trick-taking to modern deck building. Decision & Economy groups auction, worker placement, area majority, engine building, push-your-luck, and action programming by the kind of decision at the core. Social & Solo distinguishes adversarial, cooperative, and solitary player relationships.
The distribution 5-2-5-6-3 is skewed on purpose.
Why spatial placement stops at two
The smallest family is Spatial Placement, with two genres: pattern placement and tile laying. The former judges on color or shape matching; the latter on edge matching. Between them, the ways of arranging tiles flat on a surface are essentially exhausted.
Tiles as a material, though, afford far more than flat arrangement. Flicking, stacking, flipping, throwing. There is an old East Asian pastime of slapping a card down so the gust of air flips another card over; it sits in the same lineage. All of these belong to the dexterity domain, which Parlourite treats as out of scope. Spatial Placement stops at two not because few games use tiles, but because the “arrange them on a plane” cut fits in two.
Where the choices were hard
The hardest work went into separating two mechanisms that both read as “controlling territory.”
One is placing your own pieces on a board to surround an opponent, or enclosing empty regions to claim them as your own. The other is counting pieces or cubes in each region and awarding the region to whoever has the most. Both get described as “area control,” but the predicates they ask of the engine are different. The first wants connectivity checks and flood fills; the second wants per-region tallies and tie-breaking. With the mechanism axis fixed first, the two could sit in different families: the former in Abstract, the latter in Decision & Economy.
A similar question came up in Card & Hand. Trick-taking and climbing / shedding games share the surface procedure of playing cards in turn. The temptation is to merge them, but the victory conditions diverge: tricks won versus hand reduced to zero. Merging by procedural similarity makes design judgments fuzzier later on. Kept apart.
The boundary between Spatial Placement and Abstract came down to a single question: do you move the piece after placing it? If yes, you need movement rules and capture resolution; if no, you need placement constraints and endgame scoring. The two look adjacent but share almost no required machinery.
Hybrids
Most real titles don’t fit into a single genre. An engine-building core with a set-collection scoring layer. A trick-taking game with an auction phase in front. A cooperative game with movement and capture tactics inside. At mid weight and up, hybridity is the norm rather than the exception.
The convention: one genre as the primary, up to two more as secondaries. The primary is the one that directly produces the win condition and occupies the majority of play time.
If three or more secondaries feel necessary, that’s a signal that our granularity is wrong, not that the title is unusually complex. Once implementation samples accumulate, the classification itself gets another pass.
Where we are
We originally set an internal goal of multiple representative implementations per genre. Honest assessment: not met. There are not enough titles yet to fill the matrix evenly.
That said, several of the genres considered harder in designer circles, engine building and tile laying among them, do have implementations. With those in place, the remaining genres look reachable on the same ground. The remaining distance is a volume problem, not a design problem.
The classification itself is a hypothesis to be validated by implementation. The current 21-and-5 is not a final shape. At milestones in the implementation count, we expect to revisit it inductively. Adding genres, regrouping families, both remain possible.
As implementation continues, should we encounter a mechanism that genuinely troubles the taxonomy, we would rather first sit down and enjoy the unexpected game than hurry to redraw the categories.
Appendix: The 21 Genres
Abstract Family (5)
| Genre | Core predicate |
|---|---|
| Alignment | Line or row formation on a grid |
| Connection | Connected path between designated targets |
| Territory | Enclosed region ownership |
| Combat Movement | Moving pieces to capture or threaten |
| Racing | First to reach a goal position |
Spatial Placement Family (2)
| Genre | Core predicate |
|---|---|
| Placement Pattern | Color or shape matching with neighbors |
| Tile Laying | Edge matching between placed tiles |
Card & Hand Family (5)
| Genre | Core predicate |
|---|---|
| Trick-Taking | Winning tricks by card rank or suit |
| Ladder / Shedding | Reducing hand to zero |
| Set Collection | Gathering matching card groups |
| Drafting | Selecting from rotating pools |
| Deck / Bag Building | Constructing one’s own draw source |
Decision & Economy Family (6)
| Genre | Core predicate |
|---|---|
| Auction / Bidding | Price discovery for resources |
| Worker Placement | Exclusive action-slot selection |
| Area Majority / Wargame | Region-wise count comparison |
| Engine Building | Compounding production chains |
| Push Your Luck | Voluntary continuation under risk |
| Action Programming | Pre-committed action sequences |
Social & Solo Family (3)
| Genre | Core predicate |
|---|---|
| Hidden Role / Deduction | Identity inference under incomplete information |
| Cooperative | Shared win condition against the system |
| Solo Puzzle | Single-player optimization |