This is Volume IV of The Complete Raumschach Theoretical Series (IRF, 2026). This paper constitutes the first systematic middlegame theory ever published for Raumschach. This work assumes familiarity with the game’s rules, notation, and foundational opening principles. All analyses are derived from geometric first principles and await computer verification.
The middlegame in Raumschach begins — as in flat chess — when the opening’s developmental phase concludes and the position’s unique character begins to assert itself. In Raumschach, this transition typically occurs between moves 8 and 12. By this point both sides have usually deployed their Unicorns and Bishops, contested Level C in some fashion, and — if they have played well — secured the c-column against the most immediate threats. The explosive central struggle that follows is what this paper addresses.
Two aspects of Raumschach middlegame theory make it fundamentally different from its flat-chess counterpart. First, the sheer richness of available directions: where a Queen in flat chess commands 27 squares from the center of an 8×8 board, a Queen in Raumschach commands up to 26 directional rays across a 5×5×5 space — fewer squares per ray, but projecting across all three axes simultaneously. Every positional judgment must account for threats and resources that exist on levels the player cannot see in any single two-dimensional diagram. Second, the absence of castling means that king safety is a permanent, open structural problem that cannot be resolved by a single move. The king is always somewhat exposed; the question is always how exposed, and which side can exploit the exposure first.
The method of this paper is, as in the opening theory volume, derivation from geometric first principles. Every claim is grounded in the geometry of piece movement across the 5×5×5 board. Readers are encouraged to verify all lines in a physical or digital Raumschach setup, and to correct any errors they find.
The opening ends and the middlegame begins when the following conditions are approximately met:
At this moment, five key questions determine the middlegame’s strategic character. They are presented here as the Five Middlegame Diagnostics — the first things a Raumschach player should assess when the opening phase concludes:
Count the pieces and pawns each side has on Level C, and assess how well they are supported. An unsupported piece on Level C is a liability; a supported piece is an outpost. Two supported pieces on Level C amount to a fortress. The player who owns Level C in the middlegame typically dictates the game’s pace and direction.
Is the c1-column (Ac1–Bc1–Cc1–Dc1–Ec1) fully blocked, partially open, or entirely open? A fully open c1-column is a structural danger for both sides — it provides a line of force connecting the King’s starting square with the opponent’s territory. A partially open column is a dynamic tension. A fully closed column is stable but not necessarily permanent. Assess which pieces are doing the closing, and whether a single capture or move could re-open the column.
The Dual Unicorn System — both Unicorns on widely separated squares within their complementary color classes on Level C — is the single most powerful middlegame configuration. The two White Unicorns (Bb1 and Be1) are on different color classes, meaning they can cover different and complementary regions of the board. If White’s Unicorns are coordinated and Black’s are not, White has a lasting structural advantage. If both are uncoordinated, the first to coordinate wins the strategic battle.
Count the total piece and pawn activity on each of the five levels. A player with more active pieces on Levels C and D has an ascending positional advantage. A player with more active pieces on Levels A and B is playing defensively. The “level majority” is the Raumschach equivalent of space advantage in flat chess, and it determines who is pressing and who is defending.
In Raumschach, pawns can be attacked from five directions rather than two, and the three-dimensional board makes pawn chains far harder to maintain than in flat chess. Identify all pawns that lack adjacent pawn support on the same level, especially any pawns on Level C or Level D — these are typically the most exposed and the most strategically significant targets in the middlegame.
The correct starting position, verified against Dickins’ A Guide to Fairy Chess (1971), is as follows.
DEFINITIVE STARTING POSITION (per Dickins) LEVEL D — Rank 5 (Black’s second rank): [ 🨢a5 | ♗︎b5 | ♕︎c5 | 🨢d5 | ♗︎e5 ] LEVEL B — Rank 1 (White’s second rank): [ ♗︎a1 | 🨢b1 | ♕︎c1 | ♗︎d1 | 🨢e1 ] White: Bishops on Ba1 and Bd1. Unicorns on Bb1 and Be1. Queen on Bc1. Black: Unicorns on Da5 and Dd5. Bishops on Db5 and De5. Queen on Dc5.
White’s Bishops stand on Ba1 = (2,1,1) and Bd1 = (2,4,1). Their first-move Level C destinations are:
Critically, ♗︎Bd1–Cc1 closes the c-column on move 1. This means c-column closure on the first move is available not only to the Knight (Star Jump) but also to the d-file Bishop. The Third Opening Principle from Volume III should be understood to include the Bishop’s Flank as a valid first-move solution to the c-column problem. White’s a-file Bishop (Ba1) cannot reach Cc1 on move 1 — to get from Ba1=(2,1,1) to Cc1=(3,3,1) requires a change of +2 in the file coordinate, which is not a valid single Bishop step. Only Bd1 closes the c-column on move 1 by Bishop.
From Cb1, the Bishop’s most important diagonal is Cb1–Dc1–Ed1, aimed at Black’s territory via the c-file. From Cd2 (Bd1’s most active non-c1 destination), the Bishop commands Dc3 and reaches deeply into the board. Both Bishop starting diagonals offer meaningful Level C presence from move 1.
White’s Unicorns start on Bb1 = (2,2,1) and Be1 = (2,5,1). The Unicorn’s color class is determined by three invariants that each move preserves: (level+file) mod 2, (file+rank) mod 2, and (level+rank) mod 2. Computing for both:
| Square | (lvl+file) mod 2 | (file+rank) mod 2 | (lvl+rank) mod 2 | Color class |
|---|---|---|---|---|
| Bb1 = (2,2,1) | (2+2)=0 | (2+1)=1 | (2+1)=1 | (0,1,1) |
| Be1 = (2,5,1) | (2+5)=1 | (5+1)=0 | (2+1)=1 | (1,0,1) |
The two White Unicorns are on complementary color classes. Bb1 belongs to class (0,1,1) and Be1 belongs to class (1,0,1) — their reachable squares are entirely disjoint from each other, and together they cover 60 of the 125 board squares. Each Unicorn governs its own half of the board; they do not compete for the same squares and can coordinate to cover complementary territory simultaneously.
This has three immediate consequences for middlegame theory:
On Level C (level = 3), the squares accessible to each White Unicorn are:
🨢Bb1, class (0,1,1): Squares satisfying (3+file) mod 2 = 0 (file odd: a, c, e) and (file+rank) mod 2 = 1 (file and rank different parity, so rank even): Ca2, Ca4, Cc2, Cc4, Ce2, Ce4.
🨢Be1, class (1,0,1): Squares satisfying (3+file) mod 2 = 1 (file even: b, d) and (file+rank) mod 2 = 0 (both even, so rank even): Cb2, Cb4, Cd2, Cd4.
Together the two White Unicorns can reach 10 of Level C’s 25 squares — a substantial combined presence across both wings of the board’s center. The remaining 15 Level C squares are accessible to Queens, Rooks, Bishops, and Knights of either side, but to no Unicorn. In particular, Cc3 (the absolute center) is not in any Unicorn’s color class and can never be occupied by a Unicorn of either color.
Note the mirroring with Black’s Unicorns: Black’s Dd5 (class (0,1,1)) covers the same six squares as White’s Bb1, and Black’s Da5 (class (1,0,1)) covers the same four squares as White’s Be1. This means every Level C square reachable by a White Unicorn is also reachable by the corresponding Black Unicorn — the Level C Unicorn contest is always fought between matched pieces.
Level C is the fulcrum of the Raumschach middlegame. Every strategic plan ultimately relates to it — either establishing presence on it, attacking the opponent’s presence on it, or using pieces on it to project force toward the opponent’s home levels. This section classifies the four types of Level C control and the plans appropriate to each.
The most stable form of Level C control: a piece on a Level C square supported by a pawn via the upward-forward or upward-sideways pawn capture geometry. The general rule for direct pawn support: a White pawn at (B, x, r) directly defends the Level C square (C, x, r+1) via the upward-forward capture. It also defends (C, x±1, r) via upward-sideways captures. Knowing this allows exact calculation of which pawn positions provide genuine fortress support:
| Level C Square to Defend | Supporting Pawn Needed | Type of Support |
|---|---|---|
| Cc3 (absolute center) | Bc2 (upward-forward: ♙︎Bc2–Cc3) | Direct — pawn at Bc2 captures forward-upward to Cc3 |
| Cc2 (front center) | Bc1 (upward-forward: ♙︎Bc1–Cc2) | Direct — pawn at Bc1 captures forward-upward to Cc2 |
| Cb3 | Bb2 (upward-forward: ♙︎Bb2–Cb3) | Direct |
| Cd3 | Bd2 (upward-forward: ♙︎Bd2–Cd3) | Direct |
Note that the absolute center Cc3 can serve as a fortress for any piece type except a Unicorn — no Unicorn of either color can ever occupy Cc3, as established in Section III. A Queen, Knight, Bishop, or Rook on Cc3, supported by a pawn at Bc2, constitutes the most powerful possible Level C fortress.
A Unicorn or other piece on Level C that lacks direct pawn support. This is a dynamic, temporary configuration — strong because the piece is powerful, but unstable because it can be driven away or traded. The plan for the player with the Floating Unicorn is to establish a pawn on the supporting square as quickly as possible, converting the Floating Unicorn into a Fortress. The plan for the opponent is to attack the Floating Unicorn before pawn support arrives, forcing either an exchange or a retreat that concedes Level C.
A Floating Unicorn at Cc2 is a target only for Black pieces that can reach it. Cc2 belongs to color class (0,1,1) — shared by Black’s Unicorn 🨢Dd5. This means 🨢Dd5 can threaten a White Unicorn on Cc2 via a subsequent triagonal move; 🨢Da5, being class (1,0,1), cannot. The threat to a White Unicorn on Cc2 may come from Black’s Bishops, Knights, Queen, Rooks, or from 🨢Dd5 — but 🨢Da5 is permanently excluded. This gives the White Unicorn on Cc2 a partial immunity: one of Black’s two Unicorns can never touch it.
A pawn on Level C that lacks a friendly piece on Level C. As argued in Volume III, this is the least stable form of Level C control and should generally be avoided as an opening stratagem. A Wedge Pawn at Cc2 can be attacked by Black Bishops, Knights, Queens, and Rooks in the middlegame, and it cannot retreat. Its value is static: it blocks the square and prevents the opponent from using Cc2. However, a Wedge Pawn is not always bad: if supported by a piece on Level B (ready to ascend and recapture if the pawn is captured), it constitutes a legitimate positional asset. The error is in treating a lone Wedge Pawn as a substitute for a piece on Level C.
Rarely but occasionally arising, the Level C Vacuum occurs when both sides have been unable to establish stable Level C presence. In this configuration, the first side to plant a defended piece on Level C wins the strategic battle decisively. A Vacuum position should be treated as an emergency: all resources should be directed toward occupying Level C before the opponent does.
Once a player has established a Fortress or a stable Floating Unicorn on Level C, the following concrete plans are available:
The Triagonal Advance: from Cc2, a Unicorn can advance to Dd3 and then Ee4, reaching Black’s home level in two more moves and attacking Black’s E-level pawns directly. This is the most forcing plan available from a Level C outpost.
The Level C Pivot: a Queen, Rook, or Knight placed on Cc3 (the absolute center, uniquely powerful for the Queen since Unicorns cannot access this square) commands all 26 adjacent directions simultaneously. From this position, it can switch focus between threats on all flanks, making it nearly impossible for the opponent to prepare a single defensive plan. Note that Unicorns cannot access Cc3 and must use adjacent squares (Cc2, Cc4, Cd2, Cd4, etc.) as their Level C bases.
The Ascending Pawn Break: with a Unicorn established on Level C, the player can support an ascending pawn break on the b- or d-files, sending a pawn from Level B to Level C to create a passed pawn chain. The Unicorn on Level C covers the pawn’s advance and helps maintain the outpost.
The Dual Unicorn System — both White Unicorns operating on Level C with wide coverage of their complementary triagonal domains — is the most powerful positional achievement in Raumschach. This section develops its theory in full.
White’s two Unicorns are on complementary color classes: Bb1 on class (0,1,1) covering 30 squares, and Be1 on class (1,0,1) covering a different 30 squares. Together they cover 60 of the board’s 125 squares. Because they cover entirely different territory, the two Unicorns do not compete for squares — they coordinate naturally. A single Unicorn is “blind” to 95 of the 125 board squares. Two Unicorns on complementary classes are collectively blind only to 65 squares (the 15 Level C squares outside either class, plus 50 others), and see 60 squares together.
Placing the two Unicorns on well-separated squares within their respective classes maximises combined coverage: one Unicorn’s triagonal rays sweep the board’s odd-file half while the other sweeps the even-file half. The key placement question is which pair of Level C squares — one from each class — provides the broadest combined influence over the board’s center.
Unicorn 1: Cc2 — the front center of Level C. Reached on move 1 from Bb1 via (+1,+1,+1). Class (0,1,1). Controls the main triagonal toward Dd3 and Ee4; also covers Ca4, Ba3, and other squares along its eight triagonal rays.
Unicorn 2: Cd2 — the d-file of Level C, rank 2. Reached on move 1 from Be1 via (+1,−1,+1). Class (1,0,1). The Unicorn at Cd2 controls Dc3, Ec4, Ce3, and other squares via its triagonals — complementary territory to Cc2’s coverage. Together Cc2 and Cd2 place one Unicorn on the c-file and one on the d-file of Level C’s rank 2, projecting force along four distinct triagonals toward Black’s half of the board.
Both squares are reachable on the very first two moves: 1. 🨢Bb1–Cc2 then 2. 🨢Be1–Cd2, or 1. 🨢Be1–Cd2 then 2. 🨢Bb1–Cc2. This is Raumschach’s most forcing opening plan and the strategic heart of the Unicorn Surge opening.
Cc2 + Cb2: One Unicorn on the c-file (🨢Bb1–Cc2) and one on the b-file (Be1 cannot reach Cb2 on move 1 — Cb2 is class (1,0,1) ✓ but Be1=(2,5,1) to Cb2=(3,2,2) requires (+1,−3,+1), not a valid single triagonal step. Be1 reaches Cb2 via Cd2 then Cb4 then Cb2 — three moves). This configuration therefore requires more preparation.
Cc4 + Cd2: The c-file Unicorn (from Bb1 via multiple moves or from a repositioned square) reaches Cc4, while Be1 goes to Cd2 on move 1. Cc4 is class (0,1,1) ✓ for Bb1. This is a valid alternative placing the first Unicorn deeper into Level C.
Ca2 + Cd2: Bb1 retreats down the a-file triagonal to Ca2 (reachable via Aa2 first) while Be1 takes Cd2. Covers different wings of the board but requires more time to achieve.
The only impossible configurations are those where both Unicorns are assigned to squares of the same color class — since the two Unicorns are permanently on different classes, they can never both occupy, say, two (0,1,1) squares simultaneously. Beyond that, any assignment of one Unicorn to a (0,1,1) square and the other to a (1,0,1) square is a valid Dual Unicorn configuration. With Bb1 and Be1 as the Unicorn starting squares, Cc2 + Cd2 is not only possible but is the Premier Configuration.
Should a player trade a Unicorn when given the opportunity? The answer depends on five factors:
| Factor | Trade Favorable When… | Trade Unfavorable When… |
|---|---|---|
| Level C status | Your Unicorn is on Level C and trading removes the opponent’s Level C piece | Both Unicorns are on Level C; trading gives up your outpost for free |
| King safety | The opponent’s Knight or Bishop is attacking your King’s c-column approaches | Your Unicorn is the only piece controlling key squares near the c-column |
| Coordination status | Your Unicorns are not coordinated; trading one clears clutter | You have achieved the Dual Unicorn System; do not surrender it lightly |
| Pawn structure | After the trade, your remaining pieces dominate the resulting pawn structure | After the trade, the opponent’s pieces dominate the open lines |
| Material count | You are ahead in material and want to simplify | You are behind in material and need the Unicorn’s activity to create complications |
A Unicorn that is hemmed in by its own pawns — with no triagonal direction available that isn’t blocked by a friendly piece — is called a “Dead Unicorn.” This is the Raumschach equivalent of the “bad bishop” in flat chess. Signs of a Dead Unicorn:
How to avoid the Dead Unicorn: before advancing pawns on Level B, check that the pawns do not close all triagonal directions of the adjacent Unicorn. As a rule of thumb, never advance the Bb2 pawn to rank 3 if the Bb1 Unicorn has not yet moved — this can seal the Bb1 Unicorn’s key forward triagonals. Similarly, never advance the Be2 pawn if the Be1 Unicorn remains on its starting square. The two Unicorn starting squares require watching different adjacent pawns: Bb1 is threatened by premature advance of Bb2; Be1 is threatened by premature advance of Be2.
Raumschach can be understood as three interlocking two-dimensional games occurring simultaneously on three types of planes: horizontal planes (individual levels), vertical planes (rank-column slices and file-column slices), and diagonal planes (the triagonals). Expert Raumschach play requires switching focus between these three planes fluidly, recognizing that a threat on one plane may be answered by a resource on a completely different plane.
| Level | Typical occupants in middlegame | Strategic character |
|---|---|---|
| A | White King, White Rooks, some White pawns | White’s home. Defensive level. Rooks should be seeking open files to become active. King should be in a corner if possible (Aa1 or Ae1 are safest corners, away from the central c-column). |
| B | White pawns, possibly some White pieces | White’s staging area. Pawns form the structural backbone here. Pieces on Level B are “developed but not yet active” — they should ascend to Level C as soon as possible. |
| C | Pieces from both sides contesting the center | The decisive strategic level. The player who dominates Level C dictates the game. All middlegame plans refer back to Level C status. |
| D | Black’s staging area, pieces from both sides | Black’s equivalent of Level B. A White piece penetrating to Level D (especially the Unicorn via the triagonal advance) is a powerful attacking resource. Level D pressure typically precedes a direct attack on Level E. |
| E | Black King, Black Rooks, some Black pawns | Black’s home. The target. A White piece reaching Level E is typically delivering check or winning material. |
Every file — a, b, c, d, e — forms a vertical column of five squares rising from Level A to Level E. The c1-column (Ac1–Bc1–Cc1–Dc1–Ec1) is the most important because it connects White’s King’s starting square with Black’s territory along a single orthogonal line. This column is what is meant throughout this series by “closing the c-column.” Note the precision: the White King at Ac1 and the Black King at Ec5 are on the same file (c-file) but different ranks (rank 1 vs. rank 5) — they are not on the same column, and a Rook cannot attack both on a single move. The danger is along the c1-column and separately along the c5-column; they do not directly connect.
The b-column and d-column flank the central c-column and are the most natural routes for Unicorn penetration after the c-column is closed. A Unicorn on Cc2 that cannot advance along the c-column can pivot along a different triagonal and threaten Black from a flank approach.
The a-column and e-column are the wing columns. A Rook on Aa1 can slide up the entire a1-column (Aa1–Ba1–Ca1–Da1–Ea1) — a significant long-range attacking resource that becomes available once the a-file is opened.
The most complex planes in Raumschach are the triagonals — the space diagonals along which Unicorns and Queens move. The four main space diagonals of length 5 (corner-to-corner) are:
All four main triagonals pass through Cc3. However, since no Unicorn can occupy Cc3 (it lies outside all Unicorn color classes), this four-way intersection benefits Queens, Rooks, Bishops, and Knights — not Unicorns. A Queen on Cc3 commands all four main triagonals simultaneously, plus all twelve edge-diagonal directions, plus all six orthogonal directions — 26 directional rays from the single most powerful square on the board.
Attacking the King in Raumschach is profoundly different from flat chess because the King is permanently on a central file (c), has 26 possible escape directions, and yet is constantly exposed to threats along the c1-column. This section develops the first theory of attacking play in Raumschach.
The most direct attacking plan exploits the open c1-column. If the c1-column between White’s starting square and Black’s territory (Ac1–Bc1–Cc1–Dc1–Ec1) is open or can be opened, a Rook or Queen placed on the column delivers immediate threats. The full plan:
The C-Column Hammer
Clear the c1-column by trading or driving away blocking pieces, then double Rook and Queen on the column to deliver decisive vertical pressure.
Step 1: Move the White King away from Ac1 (to Aa1 or Ae1 — a corner). The King must vacate first; a King left on Ac1 while the column is opened is as endangered as the Black King it threatens.
Step 2: Open the c1-column by moving or trading away pieces on Bc1, Cc1, Dc1.
Step 3: Place a Rook on Bc1.
Step 4: Double Rooks (second Rook to Ac1 after King has moved).
Step 5: The Rooks ascend the column level by level, creating threats at each rung.
The most sophisticated attacking plan: aligning Queen and Unicorn along a triagonal aimed at the opponent’s King. Black’s King begins at Ec5 = (5,3,5). The triagonals through Ec5 are found by subtracting (±1,±1,±1) repeatedly. Two reach useful lengths:
Note that Ec5 is not on the main triagonal (which ends at Ee5=(5,5,5)). Black’s King is at the c-file corner of Level E, not the e-file corner.
Triagonal Battery Assault Plan
This requires clearing the path to Dd4 for the Unicorn and supporting the Queen’s approach to Ce3. The King at Ec5 has many escape squares — the battery alone is insufficient; it must be combined with piece support on adjacent escape squares.
Rather than a sustained attack, the Vertical Fork Offensive uses tactical strikes — vertical forks on multiple pieces — to win material and create an overwhelming material advantage. The ideal vertical fork piece is the Knight, because its (0,1,2) leap geometry creates forks across levels that are genuinely hard to see. A Knight at Cc3 can simultaneously attack pieces on different levels if they happen to be reachable by the (0,1,2) geometry. The plan: establish a Knight on Level C (specifically Cc3 or an adjacent square), then look for vertical fork opportunities as the opponent’s pieces populate Levels D and E.
Note — The Three-Dimensional Multi-Fork: Pluto’s Attack — the Cb5 Triple Fork (Opening 8, Volume III) demonstrates a fork geometry that cannot be classified as either purely vertical or purely horizontal. The Knight on Cb5 = (3,2,5) attacks three Black pieces on two different levels: the King at Ec5 via (+2,+1,0) and the Rook at Ea5 via (+2,−1,0) are both on Level E (level coordinate +2), while the Unicorn at Dd5 via (+1,+2,0) is on Level D (level coordinate +1). The fork therefore spans two levels simultaneously — it has a level component in every attack direction, yet the three targets fan out across different files and ranks rather than lying on a shared file or triagonal. This is a genuinely three-dimensional multi-fork, distinct from the vertical fork (same file or triagonal, different levels) and from any two-dimensional analogue. It exploits both the level axis and the file-rank plane simultaneously, which is only possible because the Knight’s (0,1,2) geometry generates attack vectors that vary independently in all three coordinates. It should be trained as a separate tactical theme from the vertical fork.
A slower but strategically profound attacking plan: the Unicorn ascends the main triagonal one step at a time — 🨢Cc2–Dd3–Ee4 — with each step creating a new threat that forces reactive play from the opponent.
The Staircase
Position: White Unicorn on Cc2, supported by pawn on Bc1 (or Bc3). Black’s pieces are developing but have not challenged Level C.
Step 1: 🨢Cc2–Dd3. Unicorn reaches Level D center. Threatens Ee4 next move, attacking Black’s E-level pawns.
Step 2: If Black has not blocked, 🨢Dd3–Ee4. Unicorn reaches Level E, attacking Black’s home pawns and approaching the Black King.
Step 3: The threat of further advances forces Black into permanent defensive posture.
The Staircase works because each step is a genuine threat the opponent must address, and each response slightly weakens Black’s position. By the time the Unicorn reaches Level E, Black has spent several moves reacting — allowing White to seize initiative elsewhere simultaneously.
When the opponent is attacking on Level D or E, the most dynamic defensive response is not to defend directly but to launch an equal or greater threat at the opponent’s home levels, forcing a choice between completing their attack and defending their own King. The Counter-Level Strike is most effective when your own pieces are well-placed on Level C, the opponent’s King is not adequately defended on Level A, and the opponent’s attack requires multiple more moves to become decisive.
When a piece is under attack on a given level, the Level Shift defense retreats it to a lower level while simultaneously placing it on a more active square. This is more powerful than a lateral retreat because the retreating piece changes its triagonal attack vectors entirely, potentially threatening new squares the attacker cannot easily reach. Example: a White Unicorn on Cc2 under attack retreats to Bb3 (one level down, different file and rank) — a Level Shift. From Bb3, the Unicorn attacks squares it could not reach from Cc2, and the attacking piece may no longer have it in its sights.
The most reliable defensive technique against c-column attacks: filling the c1-column with multiple pieces, making it impossible for the opponent to use Rooks or Queens effectively along it. A fully locked c1-column might look like: pawn at Bc1, piece at Cc1, piece at Dc1 — three of the five squares occupied. This is highly effective but costly in terms of piece activity: all pieces on the c1-column are primarily defensive. The player using this technique must ensure sufficient active pieces elsewhere for counterplay on the flanks.
Because there is no castling in Raumschach, the King must find its own safe haven in the middlegame. The safest positions for the White King are the Level A corners: Aa1 and Ae1 — furthest from Black’s attacking forces, and protected by the board edge on two sides. The King’s Corner Retreat should be planned from move 5 or 6 — not in a crisis, but as a quiet preparatory maneuver. The plan: King moves from Ac1 to Ab1 (one step), then Aa1 (second step), vacating the c1-column and finding shelter in the corner.
| King Position | Safety Rating | Notes |
|---|---|---|
| Ac1 (start) | ★★ | Exposed on the c-column; avoid in the middlegame |
| Aa1 (corner) | ★★★★★ | Safest corner; board edges on two sides |
| Ae1 (opposite corner) | ★★★★★ | Equally safe; board edges on two sides |
| Ab1 | ★★★★ | Good transition square en route to Aa1 |
| Ad1 | ★★★★ | Good transition square en route to Ae1 |
| Ac2 | ★★★ | Off the c-column but exposed to Level B attacks |
| Ba1 or Be1 | ★★★ | Higher level; more exposure to attack from Level C |
The Queen is the most powerful piece in Raumschach with 26 directional rays — but also the most valuable, and its early loss is catastrophic. In the middlegame, the Queen should typically be used in one of three roles:
Rooks are difficult to activate in Raumschach because their orthogonal movement only becomes useful when lines are open. A Rook is active when at least one of the following holds: it occupies an open file on its level; it occupies a column with only enemy pieces ahead of it; it is on the rank adjacent to the opponent’s back rank. The most effective way to activate Rooks is the Level A open-file strategy: advance a Level A pawn off the a- or e-file, opening the a1-column or e1-column for the corner Rook. This creates a Rook that can slide the entire column from Aa1 to Ea1, threatening Black’s home-level corner.
Two Bishops on the same side cover 24 edge-diagonal directions combined (12 each), though with substantial overlap. The Bishop pair is valuable when: the position is open on multiple levels (Bishops need clear diagonals); the pawns are not blocking the key diagonals; the opponent has traded Unicorns (removing competition for diagonal control). In closed positions with blocked levels, Bishops lose their value rapidly and Knights become comparatively stronger — analogous to flat chess’s “Knight vs. Bishop in closed positions” principle, extended to three dimensions.
Coordination is easy to gesture at and hard to measure: pieces “work well together” is a judgment every strong player makes constantly and few can state precisely. Raumcapa’s coordination term offers one workable definition, narrow but concrete: for each own piece, sum the influence of every other own piece reaching the square that piece stands on. A piece standing where its own army already projects maximum influence earns a full bonus; a piece in a corner that none of its own pieces can currently reach or reinforce earns nothing. The bonus rewards a piece for being where its own side is already strong — not for the piece’s own mobility, which is measured separately, but for whether it has chosen a square its teammates can also use, defend, or build on.
This gives the Level C outpost discussion above (Section IV) a sharper edge: a piece on Level C with no other own piece able to reach that square is contributing its individual mobility but nothing to the network, while the same piece on a square covered by a coordinating Unicorn or pawn is contributing both. Two pieces of equal individual activity are not equally valuable if only one of them stands somewhere the rest of the army can reinforce.
Piece values in Raumschach differ substantially from their flat chess counterparts, for reasons grounded in geometric coverage and the Normal Form’s particular pawn structure (the complete derivation, including Maack’s 1919 coverage analysis, appears in Volume I, Section IX).
| Piece | Value (Pawns) | Reasoning |
|---|---|---|
| Pawn | 1.0 | Standard reference unit |
| Knight | 5.0 | Coverage table (1,452 squares in SIII5) nearly matches the Rook; leap over the double pawn rank is a concrete practical advantage from move 1 |
| Bishop | 5.5 | Highest non-Queen coverage in SIII5 (1,752 squares); vertical ramp diagonals unique to three dimensions substantially increase value over the flat-chess Bishop |
| Unicorn | 3.0 (individual) | Color-binding to 30 squares per complex gives the lowest officer coverage (784 squares). Pair value somewhat above the simple sum on coordination grounds (the Dual Unicorn System) — but not the 7.0+ once claimed on the assumption the pair forces spacemate; Volume V’s exhaustive verification finds it cannot, so the premium reflects middlegame coordination only |
| Rook | 4.5 | Cumbersome in the closed Normal Form opening (10 pawns per side); value rises steeply on open columns but activation cost must be counted |
| Queen | 15.0 | 26-directional range; triagonal component gives unique forced-spacemate potential |
These figures rest on two independent sources — Maack’s 1919 geometric analysis and this series’ own positional experience — plus a third, wholly independent check: Raumcapa’s mobility-weighted evaluation function lands within a tenth or two of every figure above and reproduces the exact ranking. The full comparison, and the Unicorn’s particularly clean case, are given in Volume I, Section IX.
Knowing what a piece is worth is only half the practical question; the other half is knowing when to trade it. Raumcapa’s own documentation puts the underlying judgment well: a strong player does not calculate three moves ahead to decide whether an exchange is premature; they read the position’s character — material balance, king exposure, piece activity, line control — and feel the right appetite for simplification. That feel decomposes cleanly into four readable signals, each one confirmable in any Raumschach position without calculation, and each independently encoded in Raumcapa’s evaluation with its own weight.
| Signal | Raises appetite for trading when… | Lowers appetite for trading when… |
|---|---|---|
| 1. Material trajectory | You are already ahead by more than a pawn — even a mildly unfavorable trade is welcome, because simplification favors the side with the larger remaining force. | You are behind by more than a pawn — only a clearly winning trade is worth entering; an even trade helps whoever already has more material. |
| 2. King safety asymmetry | Your own King is the more exposed of the two — thinning the attacking force is worth a small material discount. | The opponent’s King is the more exposed of the two — keep material on the board as firepower rather than simplifying while a target exists. |
| 3. Piece activity | The piece under consideration is doing less than your other pieces — trading a passive piece for anything is rarely wrong. | The piece under consideration is your best-placed piece — trading it for a nominally equal but more passive enemy piece improves the opponent’s position, not yours. |
| 4. Open-line monopoly | The trade does not give up a fully open line — a piece commanding a clear ray (column, rank, or diagonal) is worth more than its point value for as long as it holds that line. | The piece holds the only open line of its kind on the board — surrendering it forfeits an asset no point count captures, unless the capture itself travels productively along that very line. |
None of the four signals decides an exchange alone; each shifts the bar a trade must clear, and a sound decision reads all four together. Raumcapa’s implementation makes the relative weights explicit — material trajectory shifts the bar by 40 centipawns, open-line monopoly by 25, piece activity by up to 30, and king safety asymmetry by 8 — which is itself useful strategic information: material trajectory and holding an open line matter roughly three to five times as much as king-exposure asymmetry when the four signals disagree. A position where several signals point the same direction calls for a confident decision; a position where they pull against each other — ahead on material but with the more exposed King, say — is precisely the kind of position where exchange decisions separate strong players from weak ones, because no single rule resolves it. That irreducible judgment is the real content of “exchange principles” — a feel for the position, not a formula.
A “pawn majority” in Raumschach must be evaluated level by level. A player with more pawns on Level C than the opponent has the most important type — Level C pawns can both advance toward the opponent’s home levels and support pieces on Level C simultaneously.
The three significant pawn majority types:
A pawn is “passed” in Raumschach if no opponent pawn can capture it on its way to promotion — a more demanding condition than in flat chess, since pawns can be attacked from five different directions rather than two. A truly passed pawn on Level C is extraordinarily powerful because it threatens to ascend through Level D to Level E and promote. An opponent facing a Level C passed pawn must dedicate multiple pieces to blockading it, giving the passer’s owner a decisive positional advantage.
Pawn chains can be horizontal (standard diagonal chains within a single level), vertical (a column of pawns on different levels, same file and rank), or triagonal (pawns aligned along a space diagonal). Horizontal chains behave much as in flat chess. Vertical chains on the b- or d-files are strategically significant: a vertical chain controls a file-column and supports pieces at Level C with the potential for Level B pawns to ascend directly. The strongest pawn structure combines a horizontal chain on Level B with a vertical outpost pawn on Level C — the Diagonal Wedge first described in Volume III.
When one player’s pieces are concentrated on Levels C and D while the other’s are on Levels A and B, the player on the higher levels has spatial advantage and pressing initiative. The player with the Level Gap should maintain the high-level pieces, advance a Unicorn to Level E if possible, and avoid trading high-level pieces for low-level ones. The player in the lower levels should launch counter-attacks to draw the opponent’s high-level pieces into exchanges, use the Knight (which leaps levels) to disrupt Level C positions, and attempt to send a pawn to Level C.
If one player retains both Unicorns while the opponent has traded both for Bishops or Knights, the first player has “Unicorn Dominance” — control of all the triagonal directions across all levels. This is an enormous positional advantage in open positions. Conversely, “Bishop Dominance” — two Bishops against two Unicorns — is advantageous in closed positions where the Bishops work along clear open diagonals and the Unicorns’ long-range triagonals are blocked. A player with Bishop Dominance should close the position with pawns.
A subtle but powerful imbalance arises when a player’s pawn structure has placed many pawns on squares of one particular parity pattern, inadvertently blocking most of a Unicorn’s triagonal paths. This creates permanent weaknesses in the unblocked direction — squares the opponent’s pieces can occupy without fear of pawn challenge. The opponent should post pieces on these weakened squares as permanent outposts. Before advancing any Level B pawn, verify that the advance does not inadvertently create a Dead Unicorn by blocking all of the adjacent Unicorn’s forward triagonals.
Raumschach begins with 10 pawns per side across a 125-square board. The relative pawn density is somewhat lower than flat chess, meaning open lines and diagonals arise more easily and persist longer. Players expecting the closed, maneuvering positions of flat chess pawn-chain theory will be surprised by how quickly lines open in Raumschach, especially across level transitions.
Plan Type A
Arising from the Unicorn Surge opening (1. 🨢Bb1–Cc2). White has a Unicorn on Cc2 with potential for the Staircase plan. Note that only 🨢Bb1 can reach Cc2 on move 1 via (+1,+1,+1) — 🨢Be1’s natural Level C square is Cd2, reached via (+1,−1,+1). Black’s most principled reply is 1…🨢Dd5–Cc4, seizing the c-file of Level C with the Unicorn that shares class (0,1,1) with White’s Cc2 Unicorn. Alternatively, Black may play 1…🨢Da5–Cb4 (entering Level C on the b-file), or advance the pawn ♙︎Dc4–Cc4.
Position Assessment (after approximately 6 moves)
White: Unicorn at Cc2, pieces developing on Levels A–B, pawn at Bc3. Black: Unicorn at Cc4 or Cd4, developing on Levels D–E. Evaluation: approximately equal with slight White advantage from the first-move initiative.
White’s Plan Goal 1: Execute the Staircase (🨢Cc2–Dd3–Ee4) as soon as safe. Condition: no Black piece should be on a square that can immediately recapture at Dd3 before White consolidates there. Goal 2: Simultaneously advance King to a corner (♔︎Ac1–Ab1–Aa1). Pre-emptively avoids the C-Column Hammer if Black opens the c1-column. Goal 3: Activate second Unicorn (Be1) to Cd2 — the Premier Dual Unicorn Configuration places one Unicorn at Cc2 (class 0,1,1) and the other at Cd2 (class 1,0,1). Once both Unicorns are on Cc2 and Cd2, combined coverage spans both White Unicorn color-class domains on Level C. Goal 4: Double Rooks on the a1-column after the King vacates to Aa1. Black’s Counter-Plan Black’s correct middlegame plan is to contest the Staircase before it begins. If White plays 🨢Cc2–Dd3, Black should have already placed a piece at Dd3 or controlled Ee4 to block the approach. Key defensive square: Dd4 — a Black piece there blocks the Staircase’s second step. Note that 🨢Dd5–Cc4 creates an opposing Unicorn on the same color class as White’s Cc2 Unicorn; this sets up potential Unicorn exchanges if either Unicorn advances toward the other’s territory. Black may also pursue a mirrored Staircase, advancing toward White’s home levels rather than its own: 🨢Cc4–Bb3–Aa2, the same triagonal direction in reverse, threatening White’s Level A pawns exactly as White’s Staircase threatens Black’s Level E pawns.
Plan Type B
Arising from Percy’s Defense, the Spatial Knight opening (1. ♘︎Ad1–Ac3, 2. 🨢Bb1–Cc2, 3. ♘︎Ab1–Cc1). White has a Knight on Ac3, a Unicorn on Cc2, and the c1-column closed by a Knight on Cc1.
White’s Plan Goal 1: Redeploy the Knight from Ac3 toward Level C. Best route: ♘︎Ac3–Cd3 via (+2,+1,0). Verify: Ac3=(1,3,3), (+2,+1,0)=(3,4,3)=Cd3 ✓ From Cd3=(3,4,3): (+2,−1,+2)=(5,3,5)=Ec5 ✓ — the Black King! A Knight at Cd3 delivers check to the Black King at Ec5. This is a powerful intermediate destination. Goal 2: Once Knight reaches Cd3, the position features: — Knight at Cd3 threatening Black King at Ec5 (Knight check!) — Unicorn at Cc2 covering Level C from the c-file — Knight at Cc1 closing the c1-column Combined pressure on Black’s King position is severe. Black’s Counter-Plan Black must prevent the Knight from reaching Cd3. Key blocking squares: Bd4 or Cd4 — any piece there prevents the Knight’s approach via that path. Black should also consider moving the King away from Ec5 early (to Ee5 or Ed5, off the c-file), away from the Knight’s attack vectors.
Plan Type C
Arising from the Bishop’s Flank opening (1. ♗︎Ba1–Cb1 or 1. ♗︎Bd1–Cc1). White has a Bishop on the b-file flank (Cb1) or on the c-column itself (Cc1). When the Bishop opens on Cc1, the c-column is closed from move 1 — this is a structural achievement equivalent to the Star Jump Knight. When the Bishop opens on Cb1, the c-column must be addressed on move 2 or 3.
White’s Middlegame Plan from 1. ♗︎Ba1–Cb1
Bishop at Cb1 = (3,2,1). Active diagonals from Cb1=(3,2,1):
(+1,+1,0) → (4,3,1)=Dc1
(+1,0,+1) → (4,2,2)=Db2
(0,+1,+1) → (3,3,2)=Cc2 ← covers the key Unicorn outpost!
Goal 1: Develop Unicorn to Cc2 on move 2 (🨢Bb1–Cc2).
The Bishop at Cb1 supports the Unicorn at Cc2 diagonally.
Goal 2: Develop Be1 Unicorn to Cd2 on move 2 or 3.
With Bishop on Cb1 and Unicorns on Cc2 and Cd2, White achieves the
Premier Dual Unicorn Configuration with Bishop support from the flank.
Goal 3: Close the c1-column on move 3 via Star Jump Knight (♘︎Ab1–Cc1).
Goal 4: Advance the a-file diagonal: ♗︎Cb1–Dc1 (move 4 or 5), pressing
into Black’s staging area and controlling the c-file from Level D.
The Bishop’s Flank (Ba1 form) middlegame is slightly slower on c-column
closure — that must happen on move 3. The compensation is the diagonal
support of both Unicorn outposts from Cb1 and the long diagonal toward
Black’s territory along ♗︎Cb1–Dc1.
White’s Middlegame Plan from 1. ♗︎Bd1–Cc1
Bishop at Cc1 = (3,3,1). The c-column is closed from move 1.
Diagonals from Cc1:
(+1,+1,0) → (4,4,1)=Dd1 (Level D)
(+1,-1,0) → (4,2,1)=Db1 (Level D)
(+1,0,+1) → (4,3,2)=Dc2 → (5,4,3)=Ed3 (long diagonal into Black’s half)
(0,+1,+1) → (3,4,2)=Cd2 ← the natural Be1 Unicorn square!
From Cc1, the Bishop covers Cd2 diagonally — providing immediate
support for the Be1 Unicorn’s Premier Configuration square.
Goal 1: Develop Bb1 Unicorn to Cc2 on move 2.
Goal 2: Develop Be1 Unicorn to Cd2 on move 2 or 3.
Bishop at Cc1 supports this directly: covers Cd2 via (0,+1,+1) ✓
(Cc1 does not cover Cc2 — that square lies along a Rook-type line,
(+0,0,+1), not a Bishop diagonal.)
Goal 3: With c-column closed by Bishop from move 1, all remaining moves
can be devoted to piece development and Level C occupation.
Plan Type D
Arising from Pluto’s Attack, the Cb5 Triple Fork opening (1. ♕︎Bc1–Cb2, 2. ♘︎Ab1–Bb3, 3. ♘︎Bb3–Cb5†). By move 4, White has won a Rook or Unicorn. The material advantage is decisive in the long run, but the position is structurally unusual: the Queen may be on Cb5 or may have just captured there; the Unicorns are still on Bb1 and Be1; the Bishops are still on Ba1 and Bd1; and the c1-column is entirely open. White must convert the material advantage without being punished for the undeveloped position and the open c-column.
White’s Plan Immediate priority (moves 4-6): CLOSE THE C-COLUMN. The c1-column is open, the White King is on Ac1, and Black still has pieces that can exploit the vertical line. White must place a piece on Cc1 within two moves of winning material. Best route: ♘︎Ad1–Cc1 (Star Jump, closes column immediately) or ♗︎Bd1–Cc1 (Bishop’s Flank move, same effect). Priority 2 (moves 5-8): Develop both Unicorns. 🨢Bb1–Cc2 (Premier Configuration square) and 🨢Be1–Cd2. The material advantage means White does not need to attack — simply developing while maintaining the extra piece is sufficient to win. Priority 3 (moves 6-10): Activate the Queen. If the Queen is on Cb5, it is well-placed but somewhat exposed. Consider ♕︎Cb5–Cc5 (retreating to a supported central square) or ♕︎Cb5–Cb3 (stepping back along the rank, continuing to eyeball Level C). Priority 4: King safety. White’s King should move to a corner (♔︎Ac1–Ab1–Aa1) once the c-column is closed. Do not begin this plan before the c-column is secured. Black’s Counter-Plan Black is down a Rook (≈4.5 cp) or Unicorn (≈3 cp) and has a fully developed position. The correct response is aggressive counterplay: Option 1 (if Rook was lost): Open lines immediately. Advance Level D and E pawns to create threats. Black has two Unicorns and Bishops and can generate tactical threats faster than White can consolidate. Option 2 (if Unicorn was lost): Target the c1-column while it is open. ♕︎Dc5–Dc1 attacks the White King directly along the c1-column. White MUST address this threat before developing. Key idea for Black in both cases: the more chaos in the position, the better the chances of recovering material through counter-tactics. Black’s structural development advantage (all pieces on their home squares, ready to deploy in any direction) partially offsets the material deficit. White must not allow Black to attack the open King before the c1-column is closed and the King is sheltered.
The endgame in Raumschach begins when Queens and multiple pieces have been exchanged and the position simplifies to Kings, Rooks (or Unicorns), and pawns. The transition from middlegame to endgame is marked by the same question as in flat chess: which side benefits from simplification?
Simplify when: you have a material advantage; you have a Level C passed pawn; your King is safer than the opponent’s; you have the Dual Unicorn System (caution: this means simplifying toward a favorable middlegame or an endgame with other material present, not trading down to a bare King-and-two-Unicorns endgame, which Volume V shows cannot force spacemate on its own).
Avoid simplification when: you are behind in material; your Level C position is dominant but your King is unsafe; the opponent has a passed pawn that will become decisive in a simplified endgame.
| Endgame Configuration | Winning Potential | Key Principle |
|---|---|---|
| Two Unicorns + King vs. King | ★ Draw | Two Unicorns on complementary color classes together cover 60 squares, but exhaustive verification finds zero spacemate positions anywhere on the board — combined reach too sparse to pin down a cornered King’s escape squares at once (Volume V) |
| Rook + Unicorn vs. King | ★ Draw | Exhaustively verified: zero spacemate positions exist anywhere; the 12 face-diagonal directions neither piece covers always provide an escape route (Volume V, Section XIII). Excellent middlegame restriction value; not a bare-endgame mating weapon. |
| Two Rooks vs. King | ★★ Draw from most positions, narrow winning zone | Exhaustive verification finds only 14,015 of 1,728,659 White-to-move positions are forced wins; the 3D lawnmower cannot complete confinement against best defense. Strong restriction value while other material remains (Volume V, Section VII) |
| Queen vs. King | ★★★★★ Decisive | Queen’s 26-directional range is overwhelming in open positions |
| Rook vs. King (no other pieces) | ★ Draw | Exhaustively verified: no legal position exists in which a Rook check and King support together eliminate every one of the defending King’s escape squares without the two Kings standing illegally adjacent (Volume V) |
| Two Bishops vs. King | ★★ Draw from most positions, narrow winning zone | Both Bishops start on opposite color complexes (Ba1 even, Bd1 odd) and together cover the entire board, but exhaustive verification finds only 1,972 of 1,660,585 White-to-move positions are forced wins, all already close to a corner. A reversal of the flat-chess Bishop-pair mate, not an analogue of it (Volume V) |
| Single Unicorn vs. King | ★ Draw | A single Unicorn’s 30-cell limitation makes it insufficient to force spacemate alone |
In flat chess, the King becomes a powerful attacking piece in the endgame. In Raumschach, the King has 26 directional moves and is similarly powerful — but must be used with greater care because of the three-dimensional geometry of spacemate. A King moving upward (to higher levels) in the endgame enters more dangerous territory. The ideal endgame King position is Level B, file b or d — centralized enough to support pawns on both flanks but not so high as to be exposed to checks from Level C or D pieces.
This paper has established the first systematic middlegame theory for Raumschach. The key contributions are: five diagnostic questions for position assessment, four types of Level C control with associated plans, a full analysis of the Dual Unicorn System, a theory of the three strategic planes, four attack methods and four defensive techniques, piece coordination principles, pawn structure theory, a framework for strategic imbalances unique to Raumschach, and concrete plans for the middlegames arising from the named openings (Plan Types A–D, the fourth covering the Pluto’s Attack (Cb5 Triple Fork) middlegame).
Several themes emerge as uniquely Raumschach-specific:
The Dual Unicorn System is the dominant strategic concept of the Raumschach middlegame — and its geometry must be understood precisely. White’s two Unicorns (Bb1 and Be1) are on complementary color classes (0,1,1) and (1,0,1), covering 60 squares together. Their coordination is about placing one Unicorn from each class on Level C, covering complementary triagonal domains. The Premier Configuration — one Unicorn at Cc2 (class (0,1,1), reached by Bb1 on move 1) and one at Cd2 (class (1,0,1), reached by Be1 on move 1) — provides optimal mutual coverage of both domains across Level C simultaneously. Any configuration that assigns both Unicorns to the same color class is suboptimal, as the two Unicorns would then cover identical territory.
The absolute center Cc3 is a Queen’s square, not a Unicorn’s square. All four main triagonals of the board pass through Cc3, making it the single most powerful square for Queens, Rooks, Bishops, and Knights. But no Unicorn of either color can ever reach Cc3 — it lies outside both Unicorn color classes. Strategic plans that aim to place a Unicorn on Cc3 are therefore geometrically impossible and should be replaced by plans targeting Cc2 (for Bb1) or Cd2 (for Be1).
The c1-column is the game’s most dangerous open line. The White King at Ac1 and the Black King at Ec5 share the same file (c-file) but different ranks — they are not on the same orthogonal column. The c1-column (Ac1–Bc1–Cc1–Dc1–Ec1) is the structural danger that must be closed on one of the first three moves. Moves that close it on move 1 include: the Star Jump (Knight to Cc1), the Bishop’s Flank (♗︎Bd1–Cc1), and at cost of early Queen exposure, ♕︎Bc1–Cc1.
The research directions established in Volume III remain open: computer engine analysis, deep variation theory, complete response catalogs, and the theory of specific endgame configurations. This middlegame volume adds: precise material values of Raumschach pieces, theory of Raumschach spacemate patterns in the 5×5×5 space, and full color-class analysis of strategic squares at each level.
Raumschach is 119 years old and has never had a theoretical literature. These volumes are the beginning of one. Every analysis here is offered as a starting point — a scaffold on which more precise and better-verified theory can be built. The geometry of three-dimensional chess is profound, beautiful, and deeply underexplored. May this paper inspire others to go further.