Skip to content
Chess Variants & Alternative FormsGuide

Chess Variant Concepts Compared: Anti-Dogma Reference

Compare easily confused variant concepts—ruleset versus mode, aliases versus identity, Chess960 versus generic shuffle chess, hidden information versus blindfold, engine support versus rules authority and symmetry versus balance.

Variant vs Format

Variant:

  • changes game rules.

Format:

  • organizes games/participants.

Chess960 Swiss:

  • Chess960 = ruleset;
  • Swiss = format.

Variant vs Time Control

Atomic blitz:

  • Atomic = ruleset;
  • blitz = clock category.

Blitz is not a variant.

Variant vs Mode

Correspondence Chess960:

  • Chess960 = ruleset;
  • correspondence = mode.

Blindfold orthodox:

  • orthodox ruleset;
  • blindfold mode.

Variant vs Historical Relative

Shogi:

  • independent chess-family tradition.

Not:

  • a modern modification of FIDE chess in the same sense as Three-Check.

Variant vs Fairy Composition

Atomic game:

  • playable variant.

Circe helpmate:

  • fairy composition task.

Playable variant vs fairy-composition context.

Ruleset vs Platform

Antichess on two services can differ.

Same label:

  • not enough.

Compare:

  • king;
  • check;
  • castling;
  • capture;
  • stalemate;
  • promotion;
  • draw.

Alias vs Identity

Antichess / Losing / Giveaway / Suicide:

  • overlapping family of names.

Do not merge records by string.

Rule Delta vs Full Ruleset

Rule delta:

  • concise comparison against a named reference.

Full Ruleset Profile:

  • authoritative definition including edge cases.

Both are useful.

FIDE Chess960 vs Shuffle Chess

Chess960:

  • bishops opposite colours;
  • king between rooks;
  • Black mirrors White.

Random back rank without constraints:

  • shuffle chess;
  • not necessarily Chess960.

Chess960 vs Theory-Free Chess

Randomized start:

  • removes universal orthodox opening sequence.

Specific start:

  • can develop preparation/theory.

Therefore: Chess960 ≠ theory-free.

Chess960 c-Side vs "Where King Started"

Castling side is defined by final geometry:

  • c-side → Kc/Rd;
  • g-side → Kg/Rf.

Not by:

  • king physically moving left/right.

Chess960 Castling vs Two-Piece Move Only

FIDE explicitly permits:

  • both pieces move;
  • transposition;
  • king-only;
  • rook-only.

Chess960 FEN vs Ordinary KQkq Assumption

Castling rights can be ambiguous.

Use:

  • declared Chess960 FEN convention;
  • X-FEN/Shredder-FEN where applicable.

Freestyle Brand vs Ruleset

Current FIDE:

  • Freestyle = Chess960/FIDE Fischer Random in championship context.

But:

  • event brand;
  • ruleset

remain separate metadata.

Pocket vs Captured Material List

Pocket:

  • future legal resource.

It is part of position state.

Crazyhouse vs Bughouse

Crazyhouse:

  • captured piece → your pocket.

Bughouse:

  • captured piece → partner pocket.

Same drop mechanism, different state graph.

Drop vs Move

Drop:

  • reserve → empty board square.

No origin square.

Crazyhouse vs Shogi

Both:

  • captured material reused.

Different:

  • board;
  • pieces;
  • promotion;
  • ownership conventions;
  • pawn-drop restrictions;
  • independent histories.

Similar mechanic ≠ same game.

A queen on board may actually be:

  • original queen;
  • promoted pawn.

Crazyhouse capture can revert the latter to pawn in some profiles.

State needs identity.

Three-Check Board State vs Full State

Same pieces/squares can have:

  • 0 checks delivered;
  • 2 checks delivered.

Not same game state.

Double Check vs Two Checks

Current Lichess Three-Check:

  • one move threatening king = one checking event,

even if two pieces check.

KOTH Center vs Move into Check

King reaches d4/d5/e4/e5:

  • wins only if move is legal.

KOTH king remains royal.

Antichess vs Material Sacrifice

Orthodox sacrifice:

  • gives material for compensation.

Antichess:

  • losing material may be the direct objective.

Different evaluation model.

Mandatory Capture vs Prefer Capture

Antichess:

  • if capture exists, quiet move is illegal.

This is not a heuristic.

Non-Royal King vs Check

Current Lichess Antichess:

  • king is ordinary capturable piece;
  • no check constraint;
  • no castling.

Do not preserve royal-king legality accidentally.

Atomic Capture vs Ordinary Capture

Ordinary:

  • capturer occupies target square.

Atomic:

  • capturer also explodes in current Lichess profile.

Position transition is fundamentally different.

Atomic Blast vs All Adjacent Pieces

Current Lichess:

  • adjacent non-pawn pieces explode;
  • adjacent pawns survive collateral blast.

Capturing/captured pawns still die.

Atomic Adjacent Kings vs Orthodox Kings

Current Lichess Atomic can allow adjacent kings due explosion/capture legality.

Do not import orthodox king-adjacency rule mechanically.

Horde Material vs Symmetric Evaluation

Current Lichess Horde:

  • 36 White pawns;
  • normal Black army;
  • different win conditions.

Material equality is irrelevant.

Asymmetry vs Imbalance

Asymmetric setup can be intentionally balanced.

Symmetry:

  • not required for fairness.

Racing Kings vs Check

Current Lichess:

  • giving check is illegal.

Therefore:

  • check tactics do not merely lose value;
  • they become illegal move generation.

Racing Kings White-First Rule vs Ordinary Win

White reaching rank 8 does not always end immediately:

  • Black receives compensation reply;
  • reaching rank 8 too can draw.

Blindfold vs Hidden Information

Blindfold:

  • display removed;
  • full objective state knowable from move history.

Kriegspiel/Fog:

  • rules deliberately hide information.

Four-Player Chess vs One Ruleset

4PC is a family/platform ecosystem.

Chess.com itself has:

  • Teams;
  • FFA;
  • Solo;
  • variant/custom modes.

Do not freeze one as universal Four-Player Chess.

Team vs Temporary Cooperation

Fixed team:

  • rules encode ally.

FFA alliance:

  • strategic/social behavior;
  • not permanent team state unless rules say so.

Bughouse vs Four-Player Single Board

Both:

  • four players.

Bughouse:

  • two boards + transfer.

4PC:

  • one expanded board in common implementations.

8×8 vs Chess Requirement

Chess-family games can use:

  • 9×9;
  • 9×10;
  • 10×8;
  • 10×10;
  • hexagonal;
  • multi-board;
  • wrap-around geometry.

Compound Piece Name vs Movement

Archbishop, Cardinal, Hawk, Elephant, Chancellor, Marshal:

  • names vary.

Store movement definition.

Board Edge vs Wrap Geometry

Cylinder:

  • a/h edges connected.

Orthodox:

  • hard boundary.

Same coordinate distance intuition fails.

Chess Family vs Orthodox-Centric Hierarchy

Shogi, xiangqi, janggi, makruk:

  • independent living competitive traditions.

Do not describe them as weird chess.

Historical Reconstruction vs Current Rulebook

Shatranj/chaturanga reconstruction:

  • evidence-dependent historical model.

Modern shogi/xiangqi:

  • living standardized traditions.

Different epistemic status.

Shogi Drops vs Crazyhouse Drops

Shogi:

  • integral historical rule with specialized restrictions.

Crazyhouse:

  • modern drop-chess rule family.

Mechanism comparison is useful; equation is not.

Xiangqi Cannon vs Rook

Non-capturing:

  • rook-like.

Capturing:

  • exactly one screen required.

One piece cannot be represented by ordinary rook movement alone.

Correspondence vs Engine Policy

Correspondence mode does not imply:

  • engines allowed;
  • engines forbidden.

Authority/event decides.

ICCF vs Every Correspondence Event

ICCF rules:

  • one institutional framework.

National/platform events:

  • can differ.

Blindfold vs Accessibility Blind Mode

Blindfold:

  • cognitive no-board challenge.

Accessibility blind mode:

  • interface conveys full game state nonvisually.

Do not conflate.

Simul vs Multiplayer Variant

Simul:

  • one player participates in many separate games.

Four-Player Chess:

  • multiple players share one game state.

Consultation vs Bughouse

Consultation:

  • several people jointly choose one move for one side.

Bughouse:

  • each teammate controls own board and moves independently.

Material Odds vs Variant

Odds alter starting resources. They can be classified as:

  • handicap/alternative form,

without making an entirely new movement ruleset.

Variant Strategy vs Universal Piece Values

No fixed table:

  • Atomic knight = X;
  • Crazyhouse pawn = Y

as universal truth.

Values are state/ruleset dependent.

Transfer vs Guaranteed Improvement

Variant practice may enrich:

  • geometry;
  • flexibility;
  • visualization.

It does not automatically improve orthodox rating.

Platform Rating vs Variant Strength Everywhere

A rating belongs to:

  • platform;
  • pool;
  • ruleset;
  • time/mode.

Do not compare as universal strength number.

Standard PGN/FEN vs Universal Container

They can fail to encode:

  • pockets;
  • check counters;
  • teams;
  • visibility;
  • custom boards;
  • variant castling rights.

Stockfish vs Fairy-Stockfish

Official Stockfish:

  • orthodox + Chess960 support.

Fairy-Stockfish:

  • broad variant engine derived from Stockfish.

Do not assume official Stockfish natively analyzes every variant.

UCI_Chess960 vs Variant Autodetection

UCI_Chess960 must be configured consistently.

A Chess960-looking FEN does not guarantee every GUI/engine automatically switches rule semantics.

Engine Support vs Rules Authority

An engine can implement a variant incorrectly or differently.

Engine support:

  • implementation evidence.

Not:

  • canonical rules authority.

Platform Support vs Variant Importance

Lichess's eight variants:

  • product snapshot.

Not:

  • definitive list of significant chess variants.

Solved Variant vs Solved What?

Solved must specify:

  • ruleset/version;
  • board/setup;
  • weak/strong solution;
  • draw/win conventions.

Tier A/B/C vs Quality

Tier:

  • editorial depth.

Not:

  • prestige;
  • legitimacy;
  • strategic richness.

Inventor Claim vs Community Tradition

Modern variant:

  • inventor claim may be sourceable.

Historical/regional game:

  • may not have one inventor.

Apply evidence-aware historical wording: distinguish documented source, inference, community tradition and uncertainty.

Brand vs Ruleset

A branded circuit/event can use:

  • an existing ruleset.

Brand ownership does not equal:

  • game-mechanic authorship.

Symmetry vs Balance

Symmetric armies:

  • can have first-move advantage.

Asymmetric armies:

  • can be balanced through compensating rules.

Random Start vs Random Game

Chess960:

  • pre-game randomness;
  • deterministic play afterward.

Dice chess:

  • can include in-game stochasticity.

Different design dimensions.

Rule Reset

Switching from Atomic to Antichess to orthodox:

  • explicitly reset objective, king rules, capture rules and special state.

Habit leakage is a major variant error.

Key idea

The safest sentence in variant chess begins with the exact ruleset profile, not the variant's display name.

How to Use This Reference

For each pair, ask for the defining condition, not merely the vocabulary difference. Then construct one positive example and one near-miss. This turns terminology into a discrimination skill.

The most useful recurring questions are:

  • did a legal game rule change or only the mode/format?
  • is this name identifying a ruleset or a platform implementation?
  • is the relationship historical ancestry or deliberate derivation?
  • is a mechanism shared while the full games remain different?
  • is a data/engine feature proof of rule authority or only implementation support?

Use the reference as an anti-autopilot check whenever a familiar chess word appears inside an unfamiliar ruleset.