White 204 vs Black 935
White starts from setup 204 (QBNRKNBR), Black from setup 935 (RKBRNQNB). In the encyclopedia since 2026-10-15: neither side starts better.
|cp| ≤ 5 at 400M · 800M · 1.6B · 4B nodes, two independent engines agree
How to play with different setups
White sets up the pieces by position No. 204, Black by position No. 935, mirrored on their own side; the pawns stand as usual.
Moves as usual: White starts; everything else follows the Chess960 rules.
The two setups
White: setup #204 Q B N R K N B R
King on the e-file, rooks on d, h, queen on the a-file; no bishop starts in a corner; the knights start on central files. Castling follows the Chess960 rule: king and rook end on the classical squares. In ordinary symmetric play setup 204 gives White 0.00 (Morozov Balance Index, 400M nodes).
Black: setup #935 R K B R N Q N B
King on the b-file, rooks on a, d, queen on the f-file; one bishop starts in a corner; the knights start on central files. Castling follows the Chess960 rule: king and rook end on the classical squares. In ordinary symmetric play setup 935 gives White +0.78 (Morozov Balance Index, 400M nodes).
Why the pair is balanced
At the deepest rungs (400M, 800M, 1.6B nodes) the evaluation is exactly 0.00 each time, reaching search depth 54.
Admission rule: |cp| ≤ 5 at four budgets up to 4 billion nodes, no one-sided drift along the ladder, and confirmation by two independent engines — Berserk and Lc0 — each below its own median on symmetric positions. The same corridor is required at every deeper budget (8 G, 16 G) at which the pair has been evaluated.
The shallow rungs wandered between -6 and +37 cp before settling — a reminder that quick engine checks overstate imbalances, and why the ladder goes deep.
For comparison, in symmetric play setup 204 gives White 0.00 and setup 935 gives White +0.78: with both sides on the same rank the first move is worth something. In this pair Black's own setup absorbs that edge completely.
White is not forced into a single path: the best first move is d4, and 3 first moves are within 10 cp of it.
Along the engine's main line (30 moves) the largest swing was 57 cp (around move 14), ending at +19 cp — the position stays playable for both sides.
Engine self-play agrees: 10 games, White won 0, drawn 10, Black won 0 (White scored 50.0%).
Verification ladder
Nine independent searches of the starting position, each with a larger node budget. Evaluation in centipawns from White's point of view.
| Stage | Nodes | Depth | Eval, cp | Main line |
|---|---|---|---|---|
| S0 | 100 000 | 14 | +37 | 1. f3 f5 2. d4 |
| S1 | 500 000 | 19 | -3 | 1. d4 a5 2. f3 g5 3. a4 Ra6 4. Nd2 f5 |
| S2 | 2 000 000 | 26 | +2 | 1. d4 g5 2. f3 a5 3. Nd3 f5 4. a4 d6 |
| S3 | 8 000 000 | 28 | -1 | 1. d4 a5 2. c3 g5 3. f3 f5 4. Bf2 e5 |
| S4 | 32 000 000 | 33 | -6 | 1. d4 a5 2. f3 g5 3. Nd3 d6 4. Nd2 e5 |
| S5 | 128 000 000 | 40 | -6 | 1. d4 a5 2. f3 g5 3. Nd3 d6 4. a4 e5 |
| S6 | 400 000 000 | 45 | 0 | 1. d4 a5 2. f3 g5 3. c3 f5 4. a4 d5 |
| S7 | 800 000 000 | 54 | 0 | 1. d4 a5 2. f3 g5 3. c3 f5 4. a4 d5 |
| S8 | 1 600 000 000 | 53 | 0 | 1. d4 a5 2. f3 g5 3. c3 f5 4. a4 d5 |
Deep verification
The search ladder reaches 1.6 billion nodes; admission needs a 4-billion rung and two further engines on top of that. Beyond even that, every pair was re-checked at budgets up to ten times larger and in thousands of self-play games. This section shows what each check returned for this pair.
| Nodes | Eval, cp |
|---|---|
| 4 000 000 000 | 0 |
| 8 000 000 000 | not checked |
| 16 000 000 000 | not checked |
This pair stays within the ±5 cp corridor up to 4 000 000 000 nodes.
Independent engines, run on the same position:
- Berserk 14: -3
- Lc0 0.32.1: -9
Independent confirmations: 3 of 3. An engine counts as confirming when its reading of the pair is closer to equality than its own reading of a typical symmetric Chess960 position (+10 for Berserk, +18 for Lc0). Where only one engine confirms, the pair is balanced according to that engine alone, and this page says so.
Own weight of each setup, measured against the classical array: White +41, Black +78, difference -37. A balanced pair needs White's setup to be weaker by about the price of the first move.
Extended self-play: 10 games, 0 White wins, 10 draws, 0 Black wins, White scored 50.0%.
Best first moves
The 3 best first moves for White, searched together at 400 000 000 nodes.
| # | Eval, cp | Depth | Main line |
|---|---|---|---|
| 1 | 0 | 42 | 1. d4 a5 2. c3 g5 3. f3 f5 4. a4 e5 5. Qa3 Qxa3 |
| 2 | 0 | 42 | 1. f3 a5 2. d4 g5 3. c3 f5 4. a4 e5 5. Qa3 Qxa3 |
| 3 | -7 | 41 | 1. a4 e5 2. a5 d5 3. c3 f5 4. a6 b6 5. d4 e4 |
Main-line play-out
30 moves, 5 000 000 nodes per half-move. Largest |evaluation| 57 cp, mean 30.0 cp, final +19 cp.
| # | White | cp | Black | cp |
|---|---|---|---|---|
| 1 | d4 | -1 | a5 | -11 |
| 2 | c3 | -13 | g5 | -9 |
| 3 | f3 | -13 | f5 | -17 |
| 4 | h4 | -7 | gxh4 | -4 |
| 5 | Bf2 | -10 | Ra6 | -4 |
| 6 | Rxh4 | +22 | e5 | +22 |
| 7 | dxe5 | +26 | Bxe5 | +29 |
| 8 | Nd3 | +30 | Bg7 | +27 |
| 9 | Rxh7 | +20 | Ne7 | +26 |
| 10 | Nf4 | +32 | Be5 | +23 |
| 11 | Nd3 | +31 | Bf6 | +23 |
| 12 | g3 | +26 | Bg5 | +25 |
| 13 | a4 | +37 | Re6 | +39 |
| 14 | Rh5 | +57 | Qg7 | +39 |
| 15 | Nc5 | +43 | Rd6 | +36 |
| 16 | Rxd6 | +34 | Nxd6 | +33 |
| 17 | Bd3 | +30 | Qg6 | +35 |
| 18 | Rh3 | +36 | Nc6 | +34 |
| 19 | Qd1 | +44 | Ne5 | +37 |
| 20 | Bd4 | +35 | Nc6 | +36 |
| 21 | Nb3 | +42 | b6 | +53 |
| 22 | Be3 | +47 | Re8 | +54 |
| 23 | Kf2 | +41 | Ne5 | +41 |
| 24 | Bxg5 | +41 | Qxg5 | +38 |
| 25 | Nd4 | +40 | Qf6 | +34 |
| 26 | Bc2 | +41 | f4 | +37 |
| 27 | gxf4 | +29 | Qxf4 | +38 |
| 28 | Qd2 | +45 | Qf6 | +32 |
| 29 | Ne3 | +25 | Ndc4 | +26 |
| 30 | Nxc4 | +20 | Nxc4 | +19 |
Engine self-play
10 games at 500 000 nodes per move: White won 0, drawn 10, Black won 0 — White scored 50.0%. Average length 39.7 moves.
Sample game 1 (1/2-1/2)
1. f3 f5 2. d4 a5 3. c3 a4 4. Bf2 g5 5. Nd2 d6 6. b4 Ngf6 7. b5 Qh6 8. c4 Ng7 9. Qc3 Ngh5 10. e3 g4 11. Ne2 gxf3 12. gxf3 e5 13. Rg1 Qf8 14. b6 cxb6 15. Qb4 Bg7 16. e4 fxe4 17. Nxe4 Ra6 18. d5 Bh6 19. Bxb6 Nxe4 20. Bxe4 Nf6 21. Nc3 Rxb6 22. Qxb6 Nd7 23. Qa5 Be3 24. Rg2 Bb6 25. Qxa4 Qf4 26. Rb1 Rf8 27. Re2 Bd4 28. Qa3 Bc5 29. Qc1 Qh4+ 30. Kd1 Nf6 31. Rb3 Rg8 32. Kc2 Rg1 33. Qd2 Nxe4 34. Nxe4 Bf5 35. Rb1 Rg7 36. Rb5 Qh3 37. Rxc5 dxc5 38. Qc3 Bxe4+ 39. fxe4 Qf1 40. Qe1 Qf8 41. Rf2 Rf7 42. Rd2 Rf3 43. Qe2 Rf6 44. Kb2 Rb6+ 45. Ka1 Qh6 46. Rd1 Qh3 47. Rd3 Qh6 48. Rd1 Qh3 49. Rd3 Qh6
Sample game 2 (1/2-1/2)
1. d4 g5 2. f3 d5 3. c3 f5 4. Nd3 Nd6 5. Bc2 a5 6. Qc1 h6 7. h4 Bf6 8. Be3 Nf7 9. Bf2 Ra6 10. Ng3 f4 11. Nh5 Bh8 12. g3 e5 13. dxe5 fxg3 14. Nxg3 gxh4 15. Rxh4 Nxe5 16. Rf4 Nf7 17. Qe3 Ne7 18. Nc5 Rb6 19. Na4 Ra6 20. Nc5 Rb6 21. Na4 Ra6
Play it in the CH960 app
Open this page on a phone with the CH960 app installed, or find pair 204 / 935 in the app's encyclopedia, then choose how to play.
- Local game — two players on one device, with a clock.
- With a mentor — a clearly labelled training bot with hints, at your level.
- Challenge a friend — an online game by link; both players get the same pair, colours swap on rematch.
Play this pair against the computer, right here Play a friend
FEN (Shredder, castling rights by rook files): rkbrnqnb/pppppppp/8/8/8/8/PPPPPPPP/QBNRKNBR w DHad - 0 1
Mirror twin: White 668 vs Black 18 — a different position (castling squares are fixed), not evaluated by the search.