Endgame vs Queenless Position
Queenless positions can retain middlegame character. Queen endings can be genuine endgames.
The phase is defined by changed tasks, not one material threshold.
Exact Theory vs Practical Principle
Exact theory
- defined position;
- provable result/method.
Practical principle
- reliable guidance;
- not exhaustive proof.
A practical ending can transition into exact theory.
Evaluation vs Theoretical Result
the related guide may say:
- White is better.
the related guide may later establish:
- exact draw;
- exact win;
- exact loss.
Evaluation and theoretical result are related but not identical layers.
Opposition vs Objective
Opposition is a tool for controlling access.
The objective may be:
- key square;
- pawn;
- promotion route;
- penetration.
Do not take opposition when outflanking wins instead.
Key Square vs Critical Square
Key square
- canonical K+P rule under defined configuration.
Critical square
- broader editorial term for a decisive endgame square.
Zugzwang vs No Good Move
Zugzwang requires:
- obligation to move causes the deterioration.
A position with only bad moves is not automatically zugzwang.
Triangulation vs Waiting Move
Waiting move:
- broader tempo-preserving move.
Triangulation:
- specific maneuver that transfers the move while returning to comparable geometry.
Opposition vs Corresponding Squares
Opposition is one geometric relation.
Corresponding-square theory can encode much more complex mutual-zugzwang networks.
Rule of the Square vs Exact Pawn Race
The square is a shortcut under simple conditions.
It can fail as a complete method when:
- double-step matters;
- checks matter;
- other pawns/pieces intervene;
- promotion comes with tempo;
- king interference changes route.
Raw Promotion Count vs Race Result
Promoting first does not always mean winning.
Continue through:
- check;
- capture;
- skewer;
- queen trade;
- mutual promotion.
Knight Endings vs Pawn Endings
Knight endings share:
- king activity;
- reserve tempi;
- passed-pawn logic.
But knight route tempi and forks can radically alter pure pawn-ending conclusions.
Closed Position vs Knight Automatically Better
Even in an ending:
- stable outpost;
- pawn placement;
- king;
- distant passer
matter more than the slogan.
Bad Bishop vs Lost Bishop Ending
A bishop can be structurally restricted yet defend, sacrifice for a pawn or activate after a break.
Opposite-Colored Bishops vs Automatic Draw
OCB endings have drawing tendencies.
They can still be won with:
- separated passers;
- multiple pawns;
- king entry;
- two-wing overload.
Wrong Bishop + Rook Pawn vs Automatic Draw
The famous draw requires defining conditions.
The stronger side can still win if:
- defending king is cut off from corner;
- additional material creates another route;
- another passer/weakness changes the geometry.
Bishop Pair vs Endgame Result
The bishop pair is an imbalance. It is not a theorem of victory.
Rook Activity vs Pawn Count
An active rook may compensate for a pawn.
A passive rook can make an extra pawn difficult to use.
There is no fixed exchange rate.
Open/Behind Rule vs Rook Activity
Rook behind the passed pawn is a strong heuristic.
Exceptions can arise through:
- side checks;
- frontal defense;
- king cut-off;
- checking distance;
- tactical geometry.
Lucena vs "Advanced Pawn = Lucena"
Lucena is a defined theoretical family. Near-Lucena positions can draw if one defining condition is absent.
Philidor vs "Put Rook on Third Rank"
The rank barrier has a purpose:
- prevent king advance;
- then switch to rear checks at the correct moment.
Copying the square without the timing can lose.
Vancura vs Generic Rook-Pawn Draw
Vancura is a specific active side-checking defensive geometry.
Not every rook-pawn draw is Vancura. Not every similar-looking position holds.
Rook Ending vs "All Rook Endings Are Drawn"
Rook endings have high drawing resources. Many are theoretical wins.
Never use the cliché as evaluation.
Queen Ending vs Material Counting
Queen checks and king shelter can dominate one or several pawns.
Queen vs Advanced Pawn vs "Queen Always Wins"
Advanced rook/bishop pawns and stalemate/checking geometry can create exceptions.
Exact king/pawn/queen placement matters.
Rook/Bishop vs Rook and Specialist Endings
R+B vs R and R+N vs R are specialist theoretical families.
Do not infer exact result from nominal material or a generic usually draw statement.
Two Knights vs Bare King vs Two Knights vs Pawn
Bare K+2N vs K cannot force mate against correct defense.
K+2N vs K+P can contain forced wins because the defending pawn changes mobility/tempo.
These are different theoretical families.
Troitsky/Troitzky Line vs Guaranteed Practical Win
The line/family is a theoretical guide for two-knights-vs-pawn winning zones.
FIDE move-count constraints and exact placement can change practical/rule-constrained status.
Fortress vs Legal Draw
Fortress:
- positional/theoretical construction.
Legal draw:
- achieved through an applicable rule.
Do not write:
"the game is drawn by fortress."
Stalemate vs Dead Position
Stalemate:
- side to move has no legal move;
- king not in check.
Dead position:
- mate impossible for both through every legal continuation.
Perpetual vs Repetition
Perpetual check is one practical mechanism.
Repetition is the legal position-recurrence concept.
Repetition can be forced without checks.
Theoretical Win vs Rule-Constrained Win
A no-move-count theoretical win can become a draw under the 50/75-move rules.
Record assumptions.
Tablebase vs Engine Evaluation
Tablebase:
- exact covered-state result under its metric/assumptions.
Engine:
- search/evaluation system;
- not exhaustive proof outside tablebase coverage.
WDL vs DTZ vs DTM
WDL
- result class.
DTZ
- distance to next zeroing move under Syzygy optimization/rule context.
DTM
- distance to mate.
They answer different questions.
Cursed Win vs Unconditional Win
A cursed win is theoretically winning absent the rule constraint but drawn in the represented 50-move-rule context.
Do not label it simply win without qualification.
Blessed Loss vs Draw
A blessed loss is theoretically losing absent the rule constraint but saved by the represented move-count context.
The five-valued WDL distinction contains more information than simple W/D/L.
DTZ-Optimal vs Human-Friendly
A DTZ-optimal move can look strange because it optimizes zeroing progress, not simplicity or shortest mate.
Human pedagogy should preserve WDL while preferring reproducible methods where possible.
Seven-Piece Coverage vs Permanent Limit
As of 2026-08-26, standard Syzygy/Stockfish probing is built around up-to-seven-piece coverage.
Coverage claims must remain:
- implementation-specific;
- date-stamped.
Do not make architecture depend permanently on 7.
Memorization vs Understanding
Memorize:
- canonical exact positions.
Understand:
- why the method works;
- how to reach it;
- how one-square differences change it.
Endgame heuristics help you reach the truth; exact theory tells you when the truth is already known.
How to Use This Reference
Each pair in this guide separates concepts that are close enough to be confused but different enough to change a decision. When studying a pair, do not memorize the wording. Find or create two legal positions: one in which the first concept is decisive and one in which the superficially similar concept is not.
The Anti-Dogma Test
For any endgame rule of thumb, ask four questions:
- What exact condition makes this heuristic work?
- What common exception breaks it?
- What resulting position would prove the rule irrelevant here?
- Can exact verification settle the dispute?
This converts slogans such as “rook behind the passer,” “opposite bishops draw,” or “take the opposition” into conditional tools.
Separate Result, Evaluation and Difficulty
A position can be a theoretical draw yet extremely difficult to defend, or a theoretical win that requires many only moves. Tablebase result, human evaluation and practical difficulty are different statements. Keep them separate whenever you annotate an endgame.
Separate Geometry From Naming
Named positions are shortcuts for recognizable conditions. A near-Lucena is not Lucena; a rook-pawn defense with side checks is not automatically Vancura; a blocked setup is not automatically a fortress. Names should follow verified geometry, never replace it.
<!-- ASSET: chess-g-223-chart-01 -->