This is Volume V of The Complete Raumschach Theoretical Series (IRF, 2026). This paper constitutes the first systematic endgame theory ever published for Raumschach. All analyses are derived from geometric first principles and await computer verification.
The endgame is where the geometry of Raumschach diverges most dramatically from flat chess. Three facts define the difference, and every theorem in this volume flows from them.
First: The King has 26 potential escape directions. In flat chess, a King in the center of the board commands 8 squares. In Raumschach, a King at an interior position commands up to 26 adjacent squares — across all six face-directions, all twelve edge-directions, and all eight triagonal (corner) directions. This means that confining a King to a corner, forcing it to the edge, or cutting off its retreat requires far greater force than in flat chess. Many endgames that are trivially won in two dimensions are drawn or require considerably more technique in three.
Second: The promotion square is the far corner, not the far rank. A White pawn promotes when it reaches Level E, rank 5 — the precise coordinate (E, x, 5) for any file x. This is the “far corner” of the board in the pawn’s direction of travel, requiring the pawn to advance three or four ranks AND ascend three or four levels simultaneously. The journey to promotion is geometrically longer and more dangerous than in flat chess, and the opposition of the defending King is correspondingly more complex.
Third: Stalemate is a far rarer accident. Because the King has 26 potential directions, it is almost never the case that every single one is blocked. Stalemate in Raumschach requires an extraordinary confluence of blocking pieces — a genuine curiosity rather than a routine defensive resource. This has profound implications: endgame technique that relies on stalemate tricks (a staple of flat chess endgame defense) is largely unavailable in Raumschach.
These three facts make the Raumschach endgame simultaneously more demanding (more force required to spacemate) and more decisive (fewer defensive drawing resources). The player with superior material in the endgame almost always wins — but the technique required to convert that superiority can be subtle and demanding.
This is no longer only a geometric prediction. Across the 50,000-game Raumcapa self-play corpus referenced throughout Volume III, 42,092 games — 84.2% — ended in an actual spacemate, with nearly all of the remainder (7,510, or 15.0%) drawn by threefold repetition and barely any (398, under 1%) reaching the 50- or 75-move rules. Flat chess at comparable levels of play draws far more often than that; the King’s extra escape directions evidently do not translate into extra survival chances once a real material advantage is established. The corpus also shows where those spacemates actually land: overwhelmingly on Levels B and D (13,031 and 12,991 of the 42,092 total) rather than the contested center (5,226 on Level C) or the home corners (5,292 on A, 5,552 on E). The pattern fits the theory cleanly — Level C is where control is fought over, but a King driven from it is run down one level out, in the staging territory between the center and the safety of its own corner, not in the center itself and not safely home either.
A White pawn promotes upon reaching any square on Level E, rank 5 — the five squares Ea5, Eb5, Ec5, Ed5, Ee5. These are the five squares at the “far top corner” of the board from White’s perspective: maximum level (E) and maximum rank (5). A Black pawn promotes upon reaching any square on Level A, rank 1 — the five squares Aa1, Ab1, Ac1, Ad1, Ae1. These are the “far bottom corners” from Black’s perspective.
The promoted pawn may become any piece: Queen, Rook, Bishop, Unicorn, or Knight. As in flat chess, promotion to a Queen is almost always correct, though underpromotion (typically to a Unicorn) is occasionally preferable to avoid stalemate — an even rarer scenario than in flat chess, as will be discussed in Section XVII.
Consider a White pawn that begins on Level A, rank 2 (one of two starting positions). To reach Level E, rank 5 it must ascend 4 levels and advance 3 ranks, with each pawn move doing one or the other. The minimum number of moves for this pawn is therefore 4 + 3 = 7 moves minimum. Compare this to flat chess, where a pawn needs at most 5 moves (from rank 2 to rank 7).
For a White pawn at position (Level, rank), the minimum moves to promotion at Level E, rank 5 are:
| Starting Level | Starting Rank | Levels to climb | Ranks to advance | Minimum moves |
|---|---|---|---|---|
| A | 2 | 4 | 3 | 7 |
| B | 2 | 3 | 3 | 6 |
The key implication: a passed pawn reaching Level C or D is already close to promotion in the sense that a defending King may not have enough moves to intercept it. A passed pawn on Level D, rank 4 promotes in two moves — no defending King more than one step away can stop it.
The Critical Zone for White pawn promotion is the set of squares from which a pawn promotes in three moves or fewer: Level C rank 4, Level D rank 3+, and Level E rank 4. Once a passed pawn enters the Critical Zone, the defending side faces a genuine emergency and must either blockade with a piece or accept promotion.
Unlike flat chess, where the pawn can only promote on its file, Raumschach’s pawn can shift files when capturing. A pawn that begins on the c-file can end up promoting on the b- or d-file if it has made captures along the way. This means that calculating whether a passed pawn is truly “passed” requires checking all five promotion squares (Ea5 through Ee5), not just the one on the pawn’s current file. Defenders must be vigilant about this.
Promotion to a Unicorn rather than a Queen is occasionally correct in Raumschach, when promoting to a Queen might stalemate the opponent (extraordinarily rare, but geometrically possible). In this case, promoting to a Unicorn instead wins.
The “opposition” in flat chess describes the relationship between two Kings facing each other with exactly one square between them along a file or rank. The player who does not have the move is said to “have the opposition” and holds the positional advantage. The opposition is the fundamental concept of King and pawn endgames in flat chess.
In Raumschach, the opposition must be generalized to three dimensions. This requires a new framework.
The “Chebyshev distance” between two squares in flat chess is the number of King moves required to travel from one to the other. In Raumschach, the equivalent 3D Chebyshev distance between squares (L₁, f₁, r₁) and (L₂, f₂, r₂) is:
d = max(|L₁−L₂|, |f₁−f₂|, |r₁−r₂|)
where L is the numeric level (A=1 through E=5), f is the numeric file (a=1 through e=5), and r is the rank (1–5).
This is the minimum number of King moves required to travel between two squares, since the King can move one step in any of the 26 directions simultaneously changing up to three coordinates by ±1 each.
Examples:
This is not merely a convenient definition; it is also the distance metric Raumcapa uses throughout its own positional reasoning, arrived at independently for an unrelated purpose. The engine’s influence field projects each piece’s presence outward as φ(d) = 1/(1+d), where d is exactly the Chebyshev distance defined here — chosen explicitly over the Manhattan alternative because, as the engine’s own documentation states, Chebyshev distance correctly reflects the number of King moves needed to reach a cell, where Manhattan distance over-penalizes edge cells along one axis while under-penalizing true corners. A metric this volume derives for measuring King-to-King opposition turns out to be the same metric the engine independently needed for measuring every piece’s reach. That convergence is a quiet endorsement of the metric itself, not a coincidence of two unrelated design choices.
In flat chess, the opposition occurs when two Kings are on the same file or rank with exactly one empty square between them (distance 2, same axis). In Raumschach, the direct opposition is generalized as:
Two Kings are in direct opposition when their 3D Chebyshev distance is exactly 2, they share the same axis (same file, or same rank, or same level, or same diagonal), and the player to move must yield ground.
More precisely: if the White King is at position W and the Black King is at position B, and d(W,B) = 2, then the King that must move (the one whose turn it is) cannot approach without entering the square controlled by the opposing King, and may be forced to yield a key square — just as in flat chess.
However, the Raumschach opposition is far more complex than the flat-chess opposition for one reason: the two Kings can be in “opposition” along any of 13 possible axes (3 orthogonal axes, 6 face-diagonal axes, 4 triagonal axes) rather than just 2 (file and rank). The player who understands all 13 opposition types holds a significant endgame advantage over one who only understands the standard 2D opposition.
| Axis Type | Description | Example (Kings at) |
|---|---|---|
| Orthogonal — Level axis | Same file, same rank, adjacent levels | Ac1 and Cc1 (opposition: Bc1 between them) |
| Orthogonal — File axis | Same level, same rank, adjacent files | Aa3 and Ca3 (opposition: Ba3 between them) |
| Orthogonal — Rank axis | Same level, same file, adjacent ranks | Ac1 and Ac3 (opposition: Ac2 between them) |
| Face-diagonal (6 types) | Two coordinates change simultaneously; opposition at distance 2 along a face diagonal | Aa1 and Cc1 (level+file diagonal; Bb1 between) |
| Triagonal (4 types) | All three coordinates change; opposition at distance 2 along a space diagonal | Aa1 and Cc3 (main triagonal; Bb2 between) |
The most strategically important new opposition type in Raumschach is the Level Opposition: two Kings on the same file and rank, different levels, with exactly one empty level between them. For example, White King at Bc3 and Black King at Dc3 — both on file c, rank 3, separated by Level C. The King to move must either ascend (entering the other King’s control zone) or move laterally (yielding the c-file advantage).
The Level Opposition is the dominant opposition type in Raumschach pawn endgames because it controls the pawn’s ascension path. To escort a pawn from Level B to Level C to Level D, the White King must maintain Level Opposition against the Black King on the pawn’s file — forcing the Black King to give way as the pawn ascends.
A subtler but powerful opposition type: two Kings at Triagonal distance 2 — separated by exactly one square along a space diagonal. Example: White King at Aa1 and Black King at Cc3, with Bb2 between them on the main triagonal. This opposition is relevant when a Unicorn pawn (a pawn being escorted via ascending-triagonal paths) needs the King to use triagonal opposition to clear the path.
King and Pawn vs. King (KPK) is the foundational endgame of flat chess. In Raumschach, it is similarly foundational — but far more complex, because the pawn has two directions of travel and the King has 26 escape directions. We develop the theory here for the first time.
Given a White King, a White pawn, and a Black King, when does White win (promote the pawn) and when is the game drawn? In flat chess, the answer depends on the “rule of the square” and the opposition. In Raumschach, the answer depends on a generalization of both.
In flat chess, the “rule of the square” states that the defending King can catch a passed pawn if and only if it can reach the “square” of the pawn — a geometric region defined by the pawn’s remaining promotion distance. In Raumschach, we define the analogous concept as the Promotion Cube:
Given a White pawn at position P with n moves remaining to promotion, the Promotion Cube is the set of all squares reachable by the Black King in n moves (using 3D Chebyshev distance). If the Black King is outside the Promotion Cube and it is White’s turn, the pawn promotes before the King can intercept it. If the Black King is inside the Promotion Cube, the King may be able to block or capture the pawn.
More precisely: the Promotion Cube is the set {S : d(S, any promotion square on pawn’s file) ≤ n} where n is the number of pawn moves to promotion and d is the 3D Chebyshev distance.
Example: White pawn at Dd4 needs 2 moves to promotion (advance to Dd5, ascend to Ed5 — or ascend to Ed4, advance to Ed5). The Promotion Cube for this pawn is all squares within 3D Chebyshev distance 2 of the relevant promotion squares. Any Black King within distance 2 of Ed5 is “inside the cube” and may be able to interfere; any Black King at distance 3+ is outside and the pawn promotes freely.
When the Black King is within the Promotion Cube, the White King must escort the pawn — positioning itself to maintain the Level Opposition on the pawn’s ascension file, forcing the Black King to give way at each step.
The White King “escalates” alongside the pawn, maintaining the Level Opposition on the pawn’s file at every step of the pawn’s ascent from Level B to Level E.
Example Position: White King: Bc3, White Pawn: Bc2, Black King: Dc3. White to move.
The Black King at Dc3 is directly one level above the White King — Level Opposition with White to move. Direct pawn advancement fails (Black King at Dc3 can capture pawn at Cc2 via a face-diagonal King move).
Correct plan:
1. ♔︎Bc3–Bc4! (White King advances one rank on Level B)
Now: White King at Bc4, Black King at Dc3.
If Black plays ♔︎Dc3–Dc4 (mirroring):
2. ♔︎Bc4–Cc4! (White King ascends to Level C, seizing Level Opposition!)
Now: White King at Cc4, Black King at Dc4. Black to move.
If Black plays ♔︎Dc4–Ec4:
3. ♙︎Bc2–Cc2! (Pawn ascends to Level C, supported by King at Cc4)
The Escalator Escort succeeds.
In complex KPK positions, the method of corresponding squares — traditionally a flat-chess advanced technique — might seem to offer guidance. Exhaustive verification of the Raumschach KPK state space settles the question directly: 97.2% of all White-to-move positions are forced wins, and the remaining 2.8% of drawn positions have a purely geometric characterisation (Section XVI). No table of corresponding squares is needed, and the flat-chess tools do not transfer.
The defending King in a KPK endgame has three drawing techniques available in Raumschach:
A fundamental question in any chess endgame is: with which material combinations can spacemate be forced against a lone King? The answer in flat chess is well-established. In Raumschach, with a King that has 26 escape directions and a 5×5×5 board with 125 squares, the question must be answered from geometric first principles.
| Material | Verdict | Technique | Difficulty |
|---|---|---|---|
| King + Queen | ✓ Win | Queen confines the King to a corner; King approaches; spacemate delivered by Queen | Moderate |
| King + Two Rooks | ✗ Draw (almost always) | Exhaustive verification finds only 14,015 of 1,728,659 White-to-move positions are forced wins, all already near a corner; the lawnmower / Box Method does not transfer to 3D. See Section VII. | Settled by exhaustive verification |
| King + Two Unicorns | ✗ Draw | Exhaustive verification (and independent direct scan) finds zero spacemate or stalemate positions anywhere on the board; combined coverage of 60 squares is too sparse to ever pin down all of a King’s escape squares at once. See Section IX. | Settled by exhaustive verification |
| King + Rook + Unicorn | ✗ Draw | Exhaustive verification finds zero spacemate positions anywhere in the legal state space; the 12 uncontrolled face-diagonal directions always provide an escape route. See Section XIII. | Settled by exhaustive verification |
| King + Bishop + Unicorn | ✗ Draw | Exhaustive verification finds zero spacemate positions anywhere in 7,803,603 canonical positions; the 6 orthogonal directions covered by neither piece always provide an escape route. See Section XIII. | Settled by exhaustive verification |
| King + Rook + Bishop | ✗ Draw (almost always) | Exhaustive verification finds only 19,629 of 3,390,118 White-to-move positions are forced wins, all already near a corner; the 8 uncovered triagonal directions always provide an escape route from generic positions. Max 13 moves to mate. See Section XIII. | Settled by exhaustive verification |
| King + Queen + any piece | ✓ Easy Win | The extra piece makes confinement trivial | Easy |
| King + Rook (alone) | ✗ Draw | No legal position exists in which a Rook check and legal King support together eliminate every escape square; the corner case requires the two Kings to stand illegally adjacent. See Section VIII. | Theoretical draw |
| King + Two Bishops | ✗ Draw (almost always) | Both Bishops cover the entire board between them, but exhaustive verification finds only a narrow band of already-cornered positions are won; from a generic position the King cannot be driven anywhere. See Section XII. | Settled by exhaustive verification |
| King + Single Unicorn | ✗ Draw | A single Unicorn reaches only 30 of 125 squares; cannot confine a King that has 95 squares outside the Unicorn’s color class | Theoretical draw |
| King + Single Bishop | ✗ Draw | A single Bishop reaches only one color complex (62 or 63 of 125 squares); cannot confine a King that has the complementary complex as a safe haven. See Section XI. | Theoretical draw |
| King + Single Knight | ✗ Draw | A Knight reaches at most 24 squares and cannot control adjacent squares; exhaustive scan confirms zero spacemate or stalemate positions in the legal state space | Settled by exhaustive scan |
| King + Two Knights | ✗ Draw | Exhaustive verification finds 84 forced wins out of 1,756,314 White-to-move positions, all resolved in a single move (pre-arranged mate-in-1 setups, not a forced technique). Functionally a draw, exactly as in flat chess; the extra reach across three levels does not change the result. | Settled by exhaustive verification |
It is worth being precise about how the table above came to be fully resolved, now that a real Raumschach engine exists and is documented in public. Every row above is now settled, but not by Raumcapa, and not for lack of tuning — its architecture makes it structurally the wrong tool for this particular job. Proving that a material combination forces spacemate against best defense from every possible position requires either an exhaustive tablebase (retrograde computation working backward from every spacemate position) or a deep, full-width search verifying every defensive try. Raumcapa is neither: it evaluates the current position once, statically, and never looks further than the single reply needed for its own capture calculations. It can play a perfectly good King and Rook ending without ever being able to certify whether the starting position it was given was theoretically won in the first place. The rows above were closed by a standalone tablebase generator built for the purpose, not by improving a principle-based evaluator — however well its principles happen to play.
The most important elementary endgame in Raumschach. A Queen (26 directional rays) combined with a King (26 directional moves) provides overwhelming force against a lone King. The challenge is execution — with the defender having 26 escape directions, naive confinement fails if the attacker is not methodical.
The correct technique is to confine the lone King to successively smaller regions of the board, ultimately driving it to a corner where it has at most 7 adjacent squares. Spacemate in a corner requires the Queen to cover all 7 adjacent squares simultaneously — achievable because the Queen’s 26 directions from a nearby square encompass them all.
Drive the lone Black King to a corner of the board (one of the 8 corner squares: Aa1, Aa5, Ae1, Ae5, Ea1, Ea5, Ee1, Ee5), then deliver spacemate with the Queen supported by the White King.
Step 1: Calculate the 3D Chebyshev distance from the Black King to each of the 8 corners. Drive it toward the nearest corner.
Step 2: The Queen creates a “confinement box” — a plane perpendicular to the direction of the corner — and prevents the Black King from crossing it. As the Black King approaches the corner, the Queen’s box shrinks until the King is confined to a 2×2×2 sub-cube (8 squares) near the corner.
Phase example: Black King at Cc3 (center), White King at Aa1, White Queen at Ee5.
Phase 1: ♕︎Ee5–Ce5 (cutting off Level E and Rank 5 region).
Phase 2: Shrink the box on each Black King move.
Phase 3: White King marches toward the action while the Queen confines.
Phase 4: Deliver spacemate with Black King cornered.
Because stalemate in Raumschach requires all 26 (or fewer at edge/corner) of the lone King’s squares to be blocked or attacked, stalemate is extraordinarily rare in KQK. Nevertheless, always ensure the Black King has exactly one legal move available until the spacemate is ready, then close it off.
In flat chess, K+Q vs. K can be won in at most 10 moves from any position with best play. In Raumschach, exhaustive verification across the full 2,869,412-position state space establishes the precise maximum: mate in 8 (15 plies) from the most distant won starting position, lower than the flat-chess figure despite the larger board, since the Queen’s 26 rays overwhelm the lone King’s 26 escape directions faster than the King’s extra mobility can compensate. One worst-case starting position, White to move: White King Aa1, White Queen Be4, Black King Bc3 — the most central, least confined arrangement available to the defender, and still mate in 8 with best play on both sides.
The flat-chess intuition for K+R+R vs. K is that two Rooks confine the lone King to a shrinking cuboid via the “Box Method,” eventually forcing spacemate in a corner. The geometric argument is attractive: the King has only 7 escape squares from a corner, and two Rooks covering orthogonal lines plus the White King’s support seems sufficient to close them all. Exhaustive verification finds this intuition fails in three dimensions, for the same underlying reason it failed for the single Rook.
A single Rook covers one orthogonal line (a rank, file, or column) — not a plane. In the 2D Box Method, a Rook on the 6th rank closes the entire rank because the board has only two dimensions; in 3D, the level boundary between B and C is a 5×5 plane of 25 squares, and one Rook covers only 5 of them. Both Rooks together cover at most 10 squares of a face — leaving 15 squares on every boundary face uncontrolled. The King’s 26-direction mobility means it can exploit these gaps through face-diagonals and triagonals that orthogonal Rooks cannot monitor.
At a corner, the configuration is tighter, and the arguments for a win look most persuasive. But assembling a legal configuration in which two Rooks deliver check and the White King simultaneously covers every escape square — without either Rook blocking a square the King needs covered, or the two Kings standing adjacent — is geometrically possible only in a narrow set of already-constructed positions. Forcing the game into those positions from a generic starting arrangement is what the exhaustive search shows cannot be done against best defense.
Exhaustive verification across the full 3,853,682-position state space finds only 14,015 White-to-move positions — out of 1,728,659 — in which White can force spacemate; every other position is a draw with best defense. The winning positions require at most 10 White moves from the most distant (19 plies). All are positions already close to a forced conclusion, not positions reachable from a generic central configuration against a defender who avoids the narrow winning zone. This is the same structural result as the Two-Bishop and Two-Unicorn findings (Sections XII and IX), with the Rook pair performing somewhat better — roughly 7× more winning positions than the Bishop pair — but still far short of the “reliable technique” the flat-chess analogy suggested.
The two Rooks do confine, harass, and restrict the defending King across the whole board, and a defender who plays carelessly can still be spacemated. Best defense, however, always keeps open at least one escape route from the Rooks’ orthogonal confinement — typically a triagonal or face-diagonal flight that neither Rook can monitor without moving, and that the White King cannot cover from a legal supporting distance.
The practical implication: two Rooks without other material cannot force spacemate against a lone King in Raumschach. The technique that determines Rook endgames in practice is therefore not how to convert K+R+R vs. K alone, but how to maintain enough other material — a pawn, a piece, the initiative — to exploit the defender’s errors, or to transition to a more decisive ending. The four-star middlegame rating this combination holds reflects its genuine restriction and coordination value; it does not reflect a bare-endgame mating guarantee.
In flat chess, K+R vs. K is a trivial win: the Rook confines the lone King to successively smaller strips until checkmate on the edge. In Raumschach, the 5×5×5 board with 125 squares and the King’s 26 escape directions raises the natural question: is K+R vs. K a win or a draw? It is a draw, and the reason is structural rather than a matter of insufficient technique.
The lone Raumschach King at an interior position commands 26 adjacent squares. A single Rook, even with the White King’s help, struggles to cover all the escape squares simultaneously. The lone King has many escape directions that the Rook’s orthogonal movement cannot cover on its own: specifically, the 8 triagonal directions and the 12 edge-diagonal directions, which the White King must therefore cover unassisted.
A corner square has only 7 neighbors, which makes it the best case for the defender’s King to be driven to. Even there, the geometry defeats the mating attempt. A square covering all 7 neighbors of a corner by King-adjacency alone must itself sit at the diagonally opposite corner of the local 2×2×2 cube — the unique point at chebyshev distance 1 from every one of the corner’s neighbors. But that point is also at chebyshev distance 1 from the corner square itself, since the corner and its local cube’s far point are themselves adjacent. There is no way to place the White King close enough to blanket every escape square without placing it adjacent to the Black King — which is illegal in any position, reachable or not. The same obstruction recurs at every other square on the board: nowhere does a Rook check combined with legal King support eliminate all 26 (or, at an edge or corner, fewer) escape squares at once.
King and Rook vs. lone King is a draw with best defense, for every legal position on the board. No square — corner, edge, or interior — admits a configuration in which the Rook delivers check and the White King simultaneously covers every one of the defending King’s remaining escape squares without the two Kings standing adjacent to each other, which the rules forbid. Exhaustive verification across every legal King/King/Rook placement confirms there is not one position in which the defending King has no legal reply while in check: the mating net the Rook+King combination would need to construct does not exist anywhere in the position space.
The practical implication: a Rook alone, regardless of the supporting King’s activity, cannot force spacemate against a lone King in Raumschach. The Rook can confine, harass, and restrict the defending King’s territory, and a defender who misplays badly enough can still blunder into a position with very few replies — but no sequence of legal moves can ever produce a position with zero replies. Practical players should not expend effort hunting for a forced win that the geometry does not contain; the correct technical goal with King and Rook alone is to probe for tactical chances (winning further material, provoking a different kind of error) rather than to chase a mate that cannot be constructed.
Two Unicorns versus a lone King looks, at first glance, like a fascinating endgame unique to Raumschach with no flat-chess analogue: because each Unicorn is confined to its 30-cell color class, the two White Unicorns together cover 60 of the 125 squares, on two classes that are complementary — no square is reachable by both. Exhaustive verification shows this complementary coverage never translates into a forced mate. It is, in the end, the single-Unicorn draw (Section X) with a second Unicorn added.
The Unicorn’s color class in Raumschach is determined by the parities of all three coordinates simultaneously. Specifically, a square at (Level L, file f, rank r) belongs to the parity class (L mod 2, f mod 2, r mod 2). Because a triagonal move changes all three coordinates by ±1, it flips all three parities simultaneously — so a Unicorn alternates between exactly two parity classes on successive moves. From their correct starting squares Bb1 and Be1, the two White Unicorns begin in different parity classes: Bb1 belongs to class (0,0,1) and alternates to (1,1,0); Be1 belongs to class (0,1,1) and alternates to (1,0,0). These four classes are entirely disjoint, which is precisely why the two Unicorns cover complementary, non-overlapping territory — together 60 of the 125 squares, one 30-square domain each. This part of the original analysis is correct and stands.
Exhaustive verification — across the full 4,037,081-position state space, and independently confirmed by a direct unreduced scan of every legal King/King/Unicorn/Unicorn placement — finds not one position anywhere on the board in which the two Unicorns and King deliver spacemate. The state space contains no terminal mate, no terminal stalemate, nothing: every single legal position is, and remains, a draw.
The error in the earlier reasoning was the same one corrected for the Bishop pair in Section XII: that the two Unicorns’ classes jointly covering all 125 squares, or covering “roughly 13 of the King’s 26 neighbors each” in the abstract, says nothing about whether that coverage can ever be assembled at one moment, at one position, legally. The Unicorn pair has even less to work with than the Bishop pair did: 60 reachable squares combined rather than 125, 8 short rays each rather than 12 longer ones, and a King with 26 escape directions to outmaneuver. Where the Bishop pair could at least force mate from a narrow band of already-cornered positions, the Unicorn pair cannot do so from any position — the combined reach is simply too sparse to ever pin down every one of a cornered King’s escape squares at once without illegality or a gap.
The practical implication: King and two Unicorns, without other material, is a draw against best defense, full stop — not a difficult win requiring careful technique, and not a win with a narrow non-generic exception the way the Bishop pair turned out to be. A player relying on this combination to convert an endgame should not expect it to deliver spacemate alone; it would need a third piece, or pawns, to make progress. The complementary parity-class structure remains a genuinely elegant fact about how the two Unicorns divide the board between them, and it is exactly the kind of coordination Raumcapa’s middlegame evaluation rewards as a positional asset (the Dual Unicorn System bonus, Volume III) — but as a stand-alone mating force in the bare endgame, it does not deliver.
A single Unicorn reaches only 30 of the 125 board squares — its color class. The 95 squares outside its color class are permanently inaccessible to it. This fundamental limitation means that the lone defending King can always remain on squares the Unicorn cannot reach, making spacemate geometrically impossible.
King and single Unicorn vs. lone King is a draw with best defense. The defending King can always position itself on a square outside the Unicorn’s color class (one of the 95 squares the Unicorn can never attack). The lone King need never enter the Unicorn’s color class if it moves carefully — and since the Unicorn can neither attack it nor control its path from outside the class, spacemate is impossible.
The Unicorn’s color class covers 30 squares; the defending King with 95 squares of “safe” territory cannot be cornered by a Unicorn that only controls the other 30 squares.
This theorem has a practical implication: in any endgame where one side has only a King and one Unicorn (with no pawns), the game is drawn regardless of position, provided the defender knows to stay off the Unicorn’s color class.
A single Bishop, like a single Unicorn, is confined to one color complex of the board: the 63 squares with L+f+r even, or the 62 with L+f+r odd, depending on which square it starts from. Unlike the Unicorn, whose 8 triagonal directions reach only 30 of 125 squares, the Bishop’s 12 face-diagonal directions reach roughly half the board — a far less restrictive confinement, but a confinement all the same.
King and single Bishop vs. lone King is a draw with best defense. The defending King can always remain on a square outside the Bishop’s color complex — the 62 or 63 squares (whichever the Bishop does not occupy) that the Bishop can never reach, slide through, or attack. A piece that can never attack the King’s square cannot, in combination with the attacking King alone, force spacemate.
This is a weaker safe haven than the Unicorn’s: roughly half the board rather than three-quarters. It is nonetheless sufficient. The defending King’s task is the same in kind as against a lone Unicorn (Section X) — stay off the attacker’s complex — only easier to misjudge near the boundary, since half the board, not a clear majority of it, is forbidden territory.
The practical implication is the same as for the lone Unicorn: any endgame where one side has only a King and a single Bishop, with no pawns, is drawn regardless of position, provided the defender knows to keep off the Bishop’s color complex.
In flat chess, K+B+B vs. K is a well-known but technically demanding win: the two Bishops on different colors cover all diagonal directions, and with the King’s support, spacemate can be forced in any corner. In Raumschach, the complementary-coverage structure does carry over geometrically — but exhaustive verification shows the third dimension defeats the mating technique anyway, for a different and more thoroughgoing reason than the color-complex argument addresses.
The two White Bishops do start on opposite parities — Ba1 (L+f+r=4, even) and Bd1 (L+f+r=7, odd) — and together they reach all 125 squares, exactly as the light- and dark-squared Bishops do in flat chess (Volume I, Monograph III). And the simple objection that “Bishops cannot cover triagonals, so the King always escapes” does not hold up as stated: a triagonal destination square still has a fixed parity, and whichever Bishop shares that parity can occupy or attack it from elsewhere on the board.
Exhaustive verification across the full 3,785,608-position state space (reduced via the board’s 48-element symmetry group) finds only 1,972 White-to-move positions — out of 1,660,585 — in which White can force spacemate; every other position is a draw with best defense. The winning positions form a narrow band already close to a corner, with forced mate completing in at most 7 moves from any of them. From a generic or central starting position, the two Bishops cannot drive the defending King anywhere: best defense holds a draw.
The static color-complex argument explains why no square is permanently safe from both Bishops, but it says nothing about whether the Bishops can ever force the King into a position where that coverage matters. The King’s triagonal step is never blocked mid-flight by a Bishop — correct, as the static objection says — and across the bulk of the board’s 125 squares, that turns out to be enough: the King simply has too much room, and too many directions a Bishop pair alone cannot shepherd, to ever be compelled toward a corner against correct defense. The handful of positions that are won are exactly the ones where the King is already trapped near a corner by the starting configuration, not positions the Bishops fought their way to.
The practical implication inverts the flat-chess intuition entirely: a Bishop pair without other material is, with best defense, not a winning combination in Raumschach. A player with only King and two Bishops against a lone King should expect a draw, not a technical win requiring patience — the technique the flat-chess Bishop pair relies on, methodically herding the King toward any corner, does not transfer to three dimensions. The 1,972 positions where mate is forceable are worth knowing as a practical curiosity (an opponent who blunders a King into one of them can still be punished), but they do not constitute a general winning method.
Rook and Unicorn together cover all 14 direction families that neither covers alone: the Rook covers 6 orthogonal directions, the Unicorn covers 8 triagonal directions. The combination therefore spans every direction on the board except the 12 face-diagonals. The geometric argument for sufficiency is as follows: together they can attack all 26 adjacent squares of the lone King simultaneously from nearby positions, since the King’s 26 neighbors fall into orthogonal, triagonal, and edge-diagonal subsets, and whatever the two sliders leave uncovered the King can cover.
Exhaustive verification settles the question decisively: King + Rook + Unicorn vs. lone King is a theoretical draw. Not only are there no forced wins from generic positions — there are no spacemate positions anywhere in the legal state space at all. The brute-force corner scan, an independent check of every arrangement with the defending King in any of the eight corners, confirms zero mate positions. The 12 face-diagonal directions, covered by neither the Rook nor the Unicorn, provide the defending King with escape squares that the attacking King cannot legally cover from a non-adjacent position.
Exhaustive verification across 7,874,975 canonical positions finds zero spacemate positions anywhere in the legal state space — not even the narrow already-cornered exceptions present in K+R+R vs. K and K+B+B vs. K. The combination covers 14 of 26 directional families but the uncontrolled 12 face-diagonal directions are sufficient for the defending King to always maintain at least one escape route against legal White King placement. This overturns the claim in earlier versions of this series that Rook + Unicorn forces spacemate and constitutes the most important mixed-piece endgame combination.
The practical implication: Rook + Unicorn without other material is not a mating weapon against a lone King. The two pieces have excellent coordination value in middlegame positions and are effective at gaining material against an opponent who has other pieces — the Rook + Unicorn pairing for area control remains the most efficient two-piece attacking combination in the middlegame (theory1, §VIII). That claim concerns their joint coverage of 14 directional rays in a populated position, which is accurate; it is only the bare-endgame mating power against a lone King that does not exist.
Rook and Bishop together cover all orthogonal and face-diagonal directions (18 of 26), leaving only the 8 triagonal directions uncovered. Exhaustive verification finds 19,629 forced wins out of 3,390,118 White-to-move positions (0.58%), all in positions already near a corner, with at most 13 White moves to mate from the most distant. This is the same structural result as K+R+R and K+B+B — a narrow winning zone that cannot be reached from a generic position against best defense — but with the triagonals as the persistent escape directions rather than face-diagonals or orthogonals. The defending King always has at least one triagonal neighbor that neither the Rook nor the Bishop can reach, and that the White King cannot legally cover.
The combination of one Bishop and one Unicorn spans all non-orthogonal directions of the board: the Bishop covers 12 face-diagonal directions, the Unicorn covers 8 triagonal directions. Exhaustive verification settles the question: King + Bishop + Unicorn vs. lone King is a complete theoretical draw. Zero spacemate positions exist anywhere in 7,803,603 canonical positions. The 6 orthogonal directions are covered by neither piece, and the defending King always maintains at least one orthogonal escape square that the Bishop cannot reach, the Unicorn cannot reach, and the White King cannot legally cover from a non-adjacent square.
Queen vs. Rook with both Kings present is typically won by the Queen in flat chess, though it can be drawn in some positions. In Raumschach, the Queen’s 26 directional rays give it a decisive superiority over the Rook’s 6, and this endgame is won for the Queen in almost all positions. The technique: use the Queen’s triagonal rays to attack the Rook’s position while keeping the opposing King at bay.
The flat-chess principle “place the Rook behind the passed pawn” applies directly to Raumschach but with a three-dimensional interpretation. A White passed pawn advancing up the c-column (ascending level by level on the c-file) should have a White Rook “below” it — on the same file and rank but a lower level, pushing it from behind. The White Rook at Ac3 supporting a pawn on Bc3 then Cc3 then Dc3 is the 3D version of this classical technique.
Similarly, the defending Rook should try to get “in front of” the passed pawn — on the same file and rank but a higher level — to blockade it. A Black Rook at Ec3 blockades a White pawn ascending the c3-column, since the pawn must eventually reach Ec3 and the Rook will capture it or force the pawn to deviate.
In flat chess, a Rook on the 7th rank is enormously powerful because it attacks the opponent’s pawns and cuts off the opposing King. In Raumschach, the equivalent is a Rook on Level D (one below the opponent’s home territory): a White Rook on Level D attacks Black’s Level D pawns and cuts off the Black King from retreating to Level D from Level E. The “seventh-level Rook” principle: try to advance a Rook to Level D or E as soon as possible in Rook endgames.
The race between a Rook trying to stop a passed pawn and a passed pawn trying to promote before the Rook can catch it is the most common and tense endgame situation in Raumschach Rook endgames. A key insight: a Rook can always reach any square in at most 2 moves (move to any file in one move, then to any rank/level in the second). So if the pawn needs 3+ moves to promote, the Rook can always catch it — unless blockaded or sacrificed. Passed pawns that need only 1 or 2 moves to promote are truly unstoppable by a lone Rook if the pawn’s owner controls the intervening squares.
A Unicorn can escort a pawn to promotion via the triagonal. Consider a White pawn at Bc2 and a White Unicorn at Aa1. The Unicorn’s triagonal direction (+1,+1,+1) takes it from Aa1 to Bb2 to Cc3 to Dd4 to Ee5 — the main triagonal. However, if the pawn ascends via the c2-column (ascending level by level), the promotion square Ec2 is not a promotion square (promotion requires rank 5). This illustrates the importance of pawn file choice: a pawn must eventually reach rank 5 AND Level E.
A Unicorn in a pawn endgame can execute a devastating fork: simultaneously threatening to promote an enemy pawn (by attacking the blockading piece) while creating its own promotion threat elsewhere. Because the Unicorn moves along triagonals, its forks operate across all three levels simultaneously — making them uniquely difficult to see and defend against.
The most important strategic concept in Unicorn + pawn endgames: ensure that your pawn’s promotion square is on the same color class as your Unicorn. If the pawn’s destination promotion square is on the Unicorn’s color class, the Unicorn can cover it, escort the pawn there, and even defend the promotion square from a triagonal. If the promotion square is NOT on the Unicorn’s color class, the Unicorn cannot protect the pawn at the moment of promotion — a significant defensive weakness.
Principle 1: The Active King. The White King should aim for the high ground (Level B or C) in a pawn endgame, from which it can both escort White pawns toward promotion and prevent Black pawns from reaching Level A rank 1. A King on Level C commands influence over pawns on Levels B, C, and D simultaneously.
Principle 2: The Outside Passed Pawn. In flat chess, an outside passed pawn forces the defending King to run to the wing, allowing the attacking King to eat the remaining pawns. In Raumschach, the “outside” dimension is three-dimensional: a passed pawn on the a-file AND on Level C is doubly outside — far from both the file-center and the level-center. The defending King cannot simultaneously address a threat on the a-file of Level C and threats on the e-file of Level B; it must choose, and the attacking King exploits the choice.
Principle 3: Pawn Majority Conversion. A pawn majority on one level should be converted to a passed pawn by advancing the majority forward and ascending one pawn to the next level. A pawn majority on Level B (more White pawns than Black pawns there) should be advanced to create a passed pawn before converting to Level C.
Principle 4: Don’t Rush to Promote. Unlike flat chess where promotion is almost always the priority, in Raumschach a pawn that rushes to promotion may leave behind a weakened pawn structure on lower levels that the opponent’s King exploits. Promote when the resulting Queen will decide the game immediately.
The Raumschach pawn is structurally different from its flat-chess counterpart. It advances toward its promotion rank — the single square at Level E, rank 5 for each file — by moving through faces (two non-capture directions: advance rank, or advance level) or through edges (five capture directions combining one forward step with a file shift, or combining both forward directions). From a center position such as Cc3, the pawn can move to Cc4 (advance rank) or Dc3 (advance level), and can capture on Cb4, Cd4, Db3, Dc4, or Dd3.
This geometry creates a promotion path that is two-dimensional in (level, rank) space: both coordinates must reach their maximum simultaneously. In complex KPK positions, the method of corresponding squares — a flat-chess advanced technique — might seem to offer guidance. Exhaustive verification of the Raumschach KPK state space settles the question directly: the overall win rate is high, the remaining draws have a purely geometric characterisation (Section XVI), and no table of corresponding squares is needed.
In flat chess, King and Pawn vs. lone King is one of endgame theory’s most instructive topics: roughly half of all positions are draws, drawing technique requires understanding the opposition, the rule of the square, and key squares. Exhaustive verification of Raumschach’s KPK state space (2,478,642 legal positions across all pawn levels and piece placements) produces a very different picture.
Exhaustive retrograde analysis finds that 88.4% of White-to-move positions are forced wins for White — roughly double the flat-chess rate of ~55%. The remaining 11.6% are draws, and their geometry is fully characterised: every drawn KPK position has the pawn on file a or file e. Pawns on files b, c, or d win from 100% of positions. The file-a and file-e pawns draw from 28.8% of positions each.
Win rates by pawn level: Level B (85.8%), Level C (87.5%), Level D (90.0%), Level E (90.8%). Win rates by proximity to promotion: a pawn at Level D rank 5, or Level E rank 4 (one non-capture step from promotion) wins 98.0% of the time.
Unlike flat chess where promotion runs along a single file, Raumschach’s two-dimensional promotion path produces a diagonal structure in (level, rank) space. Win rates increase as the pawn’s (level, rank) pair approaches (E, 5): symmetrically across the level=rank diagonal, since either coordinate advancing toward maximum is equally valuable. This symmetry is a direct consequence of the pawn having two equally-weighted non-capture directions.
The White King’s 26-direction mobility makes it far harder for the defending Black King to maintain a blockade in 3D. The pawn also has more attacking reach (5 capture directions vs. 2 in flat chess), making it harder to shadow the pawn without being captured. These factors together roughly double the win rate relative to the 2D game.
Draws are concentrated entirely on files a and e, where the pawn’s promotion path runs along the extreme edge of the board. The file cannot shift left (from file a) or right (from file e) without a capture, and the defending King can use the board edge to maintain a sustainable blockade more easily than in the interior. This is the 3D analogue of the a-file/h-file draw in standard KPK: a structurally similar confinement argument applies when the pawn’s forward corridor runs along a board edge.
The practical rules for Raumschach KPK: if the pawn is on file b, c, or d, White wins unconditionally — no technique is needed. If the pawn is on file a or e, check whether the Black King can maintain a blockade along the promotion corridor; with best play, 28.8% of those positions draw.
The flat-chess tools for KPK — the rule of the square, opposition, key squares, corresponding squares — do not have direct Raumschach analogues. The win rate is high enough across most positions that no technique is required; the position is won from nearly everywhere on the interior files. The traditional “corresponding squares” computation (which the research agenda previously listed as an open question) is answered by this result: no such table is needed, because the question collapses to “is the pawn on an edge file?”
In flat chess, zugzwang arises because the King’s movement options are limited: from a central position it has 8 squares to go to, and in some KPK positions all 8 are disadvantageous. In Raumschach, the King has 26 squares from a central position and 7 from a corner. Putting a Raumschach King in true zugzwang requires far greater precision and almost always requires the King to be near a board edge.
The fundamental reason is triangulation. In flat chess, triangulation means moving the King three steps to reach the same square while passing the move to the opponent — a manoeuvre that is often impossible because the King has too few available routes. In Raumschach with 26 directions, any King in an interior position can always triangulate: move anywhere adjacent and return, spending two moves to “pass”. This means interior positions almost never admit zugzwang: the side to move can always neutralise the obligation by wasting a tempo. True zugzwang in Raumschach pawn endgames is therefore a phenomenon of the board’s edges, where the King’s mobility drops enough that certain triangulation routes are unavailable.
The KPK retrograde analysis (Section XVI) provides the first exhaustive characterisation of pawn-endgame zugzwang in Raumschach. Of the 1,223,118 valid King+Pawn vs King positions, 126,614 are zugzwang positions in the strict sense: the position is a draw with White to move but a forced win for White when Black is to move (Black’s obligation to move costs a decisive tempo).
Every one of the 126,614 Black-in-zugzwang positions in KPK has the White pawn on file a or file e. Pawns on files b, c, or d produce no zugzwang positions at all. The edge file is not merely correlated with zugzwang — it is a necessary condition. This result follows directly from the 100% win rate for interior-file pawns: since every interior-file position is a forced win for White regardless of whose turn it is, no position can be zugzwang (zugzwang requires the position to be DRAW with one side to move and WIN with the other).
There are no White-in-zugzwang positions in KPK (positions where White prefers Black to have the move). White’s King and pawn combination is never disadvantaged by having the move: the ability to advance the pawn or improve the King’s position is always an asset.
The geometric reason: a pawn on file a or e must promote via a column that runs along a board edge. The defending King, when positioned correctly, can hold a blockade of this narrow corridor without the triangulation options available in the board interior. A pawn on file b, c, or d promotes via a path with enough lateral room that the White King’s 26-direction mobility overwhelms any blockade — triangulation routes always exist, and the defending King is inevitably dislodged.
Corner Zugzwang. A King in a corner (7 adjacent squares) is most vulnerable. When all 7 are unfavorable, the King is in zugzwang. This is the basis of most piece-endgame mating patterns.
Edge Zugzwang. A King on a board face has fewer than 26 squares; triangulation becomes difficult. Pawn-endgame zugzwang arises here, specifically where the pawn’s promotion corridor runs along the edge, restricting both Kings’ maneuvering room.
Pawn-Diagonal Opposition. The most important pawn-endgame zugzwang in Raumschach arises in the (level, rank) diagonal space. Because the White pawn advances toward both higher level AND higher rank (non-capture moves are “advance rank” or “advance level”), the critical squares near the promotion zone are diagonal in the (L, r) plane. The Black King must position itself to control this diagonal path while avoiding zugzwang.
When both sides have a pawn, zugzwang arises across a much wider range of positions — including some with interior-file pawns. The systematic frozen-pawn analysis (analysing King opposition for each fixed pawn pair) reveals the following structure:
Zugzwang-richness increases sharply when the two pawns are on opposite edge files: configurations with White pawn on file e and Black pawn on file a (or vice versa) produce 50–80% more zugzwang positions than same-file configurations. The reason: both Kings are simultaneously under edge-file constraints, and the space between the two promotion paths is narrow enough that no triangulation avoids the critical obligation.
Interior-file KPKP does produce some zugzwang (unlike the single-pawn case), because with two pawns on the board each King must simultaneously support its own pawn’s advance and contest the opponent’s advance — a genuine two-front dilemma that can trap the King even in the interior. However, the count of zugzwang positions for center-file KPKP is substantially smaller than for edge-file configurations, consistent with the general principle that interior King mobility reduces zugzwang opportunities.
Two distinct things sit under the heading of zugzwang: recognizing that a position is zugzwang, and engineering one through triangulation. Recognition — does the side to move already have no move that does not worsen their position? — is a comparison between static evaluations. Engineering one — choosing a route that hands the obligation to the opponent — requires reasoning about a sequence of future positions.
In Raumschach interior positions, triangulation is almost trivially available: the King has 26 directions, and moving anywhere adjacent then returning wastes exactly one tempo. Engineering zugzwang therefore requires one of the following structural conditions: (a) the King is near a board edge where certain triangulation routes are cut off; (b) the King has a specific parity constraint due to its relationship to a pawn’s coverage; or (c) a pawn blockade creates a two-front dilemma (as in KPKP) that limits effective tempo-wasting.
Raumcapa’s own architecture makes this distinction concrete. The engine carries a mobility-restriction proxy, active only in the endgame (both Queens off the board, opponent down to at most one minor piece): an opponent with very few legal replies is scored as more likely to be obligated. This is explicitly a correlate of obligation, not a certificate of it — the engine’s own documentation flags that it will sometimes reward a cramped opponent who has a perfectly good move available — and triangulation itself, the active engineering of zugzwang, is deliberately not attempted. Recognizing a cramped position is one ply of work; engineering one is a sequence of them, and an engine that evaluates only the position in front of it was never going to do the second.
Rule 1: Interior files, single pawn. If your passed pawn is on file b, c, or d, zugzwang is irrelevant — you win unconditionally regardless of turn. Do not waste time calculating opposition; simply advance.
Rule 2: Edge files, single pawn. On file a or e, zugzwang is possible. The Black King’s goal is to reach and hold the blockade square directly in front of the pawn in both the level and rank directions simultaneously. The White King’s goal is to drive the Black King away from this square, then support the pawn through.
Rule 3: Two-pawn endings. With pawns on both sides, zugzwang arises even on interior files. The King must not move passively when both pawns are still alive: each tempo wasted on inactive King movement can allow the opponent to establish a zugzwang position on the critical squares between the two promotion paths. Opposite-edge-file pawn configurations (one pawn on file a, the other on file e) are the most zugzwang-sensitive positions in all of Raumschach pawn theory.
Rule 4: Near promotion. As the pawn approaches the promotion zone (level D or level E, rank 4 or rank 5), the number of zugzwang-capable positions grows sharply. A pawn at Level D rank 5, or Level E rank 4 (one non-capture step from promotion) wins 98% of the time; many of the remaining 2% draw positions are precisely zugzwang positions where the Black King has managed to hold the blockade through the critical promotion diagonal.
For a King to be stalemated in Raumschach, all of its adjacent squares (up to 26) must be either occupied by its own pieces or attacked by the opponent’s pieces — while the King itself is not in check. In practice, the attacking side needs an enormous army of pieces to control 26 squares simultaneously without one of them checking the King. This is so demanding that in most positions, if the attacker is powerful enough to control 26 squares, they can simply choose to deliver spacemate instead.
The most realistic stalemate scenarios in Raumschach involve a lone King in a corner position (only 7 adjacent squares needed to block) where the attacker makes a careless move. In practice, a single Queen cannot attack all 7 adjacent squares of a corner King simultaneously while not attacking the King itself — multiple pieces are needed, creating the genuine stalemate risk, though still rare.
For the attacking side: be alert to stalemate only in corner positions where the lone King has 7 or fewer adjacent squares. Before each move, verify that the lone King retains at least one legal move. For the defending side: stalemate can occasionally be sought in desperate positions with a corner King, but it is almost never achievable without active attacker carelessness. Do not rely on stalemate as a defensive strategy; it will almost never materialize.
This paper has established the first systematic endgame theory for Raumschach across nineteen sections: pawn promotion geometry, the 3D opposition, the Promotion Cube, King and pawn endgame theory, piece sufficiency for spacemate, elementary endgame techniques for K+Q, K+RR, K+R, K+UU, K+U, K+B, and K+BB combinations, and the drawn status of K+RR and K+RU, mixed piece endgames, Rook endgames with pawns, Unicorn endgames, pure pawn endgames with full KPK analysis, zugzwang in three dimensions with multi-pawn theory, and the rarity of stalemate.
Five results stand out as the paper’s most significant contributions:
First: The Two-Bishop Reversal. Contrary to the flat-chess intuition that a Bishop pair forces mate, King and two Bishops versus lone King is a draw from almost every position in Raumschach. Exhaustive verification finds only 1,972 White-to-move positions, out of 1,660,585, in which the Bishops can force spacemate — all of them already close to a corner, none of them reachable by driving a King there from a generic starting position. The two Bishops do jointly cover all 125 squares, exactly as the flat-chess Bishop pair does, but coverage of every square is not the same as the ability to compel the King toward any particular one of them; the King’s unobstructable triagonal step turns out to give it enough room across the bulk of the board that best defense holds a draw.
Second: The Two-Unicorn Draw. Contrary to the original analysis, King and two Unicorns versus lone King is a theoretical draw, not a win — exhaustively verified, with not a single spacemate or stalemate position found anywhere on the board. The two Unicorns do cover complementary, non-overlapping 30-square color classes, exactly as claimed, but joint coverage of 60 squares in the abstract never translates into the ability to pin down all of a cornered King’s escape squares at one legal moment. The Dual Unicorn System remains, as the middlegame theory volume established, a genuine positional asset — it is simply not, on its own, a mating weapon.
Third: The Single-Unicorn Draw. A single Unicorn cannot force spacemate — its 30-cell color class covers only 24% of the board, leaving 76% as permanent safe haven for the defending King. This is the Raumschach equivalent of the single-Bishop draw and is equally fundamental.
Fourth: The Rook’s Structural Insufficiency. King and Rook versus lone King is a theoretical draw, settled by exhaustive verification across every legal King/King/Rook placement on the board. The intuitive corner-mate construction — White King diagonally adjacent to a cornered defender, Rook delivering check along the remaining file — fails for a precise geometric reason: the unique square close enough to cover all of a corner’s other neighbors is itself adjacent to the corner, which means the two Kings would have to stand illegally next to each other. No square on the board escapes this obstruction. A Rook can confine and harass a lone King in Raumschach, but it cannot mate one.
Fifth: Promotion Requires Six Moves Minimum. The promotion journey in Raumschach is significantly longer than in flat chess, requiring a minimum of six moves (from Level B, rank 2, or equivalently Level A, rank 3) and up to seven (from Level A, rank 2, the longer of the two actual starting configurations established in Section II). Passed pawns are therefore more valuable relative to pieces than in flat chess — they require more moves to convert, which means they monopolize more of the opponent’s attention for longer.
The research agenda that this paper leaves open is narrow. The KPK question has been resolved (Section XVI): 88.4% of White-to-move positions are forced wins, the 11.6% of drawn positions are fully characterised by a geometric rule (pawn on file a or e), and no table of corresponding squares is needed. The multi-pawn zugzwang theory (Section XVII) is now established: zugzwang in KPK is confined to edge files, zugzwang in KPKP arises across all file combinations with opposite-edge configurations being the richest, and four practical rules are given. The research agenda of this paper is complete.
With the publication of this fifth volume, Raumschach now has — for the first time in its 119-year history — a complete theoretical literature covering the opening, middlegame, and endgame phases of play. Ferdinand Maack envisioned a chess that matched the three dimensions of real conflict. In developing this theory, we have attempted to do justice to the beauty and depth of what he created. May it be the beginning of a long tradition of Raumschach scholarship.