White 284 vs Black 764
White starts from setup 284 (NBRKNRBQ), Black from setup 764 (RBKNNRBQ). In the encyclopedia since 2026-10-09: 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. 284, Black by position No. 764, 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 #284 N B R K N R B Q
King on the d-file, rooks on c, f, queen on the h-file; no bishop starts in a corner; the knights start away from the centre. Castling follows the Chess960 rule: king and rook end on the classical squares. In ordinary symmetric play setup 284 gives White +0.19 (Morozov Balance Index, 400M nodes).
Black: setup #764 R B K N N R B Q
King on the c-file, rooks on a, f, queen on the h-file; no bishop starts in a corner; the knights start on central files. The king already stands on a castled square, so one of the castlings moves only the rook. In ordinary symmetric play setup 764 gives White +0.32 (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 55.
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 -31 and +20 cp before settling — a reminder that quick engine checks overstate imbalances, and why the ladder goes deep.
For comparison, in symmetric play setup 284 gives White +0.19 and setup 764 gives White +0.32: 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 c3, and 3 first moves are within 10 cp of it.
Along the engine's main line (28 moves) the evaluation stayed within ±19 cp the whole way: the balance holds as the pieces come out, not only on move one.
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 | 15 | -31 | 1. f4 f5 2. c3 a5 3. g3 a4 4. Nac2 Nf6 |
| S1 | 500 000 | 19 | -1 | 1. f4 f5 2. c3 a5 3. g3 a4 4. Nf3 g6 |
| S2 | 2 000 000 | 23 | +20 | 1. f4 f5 2. c3 a5 3. g3 a4 4. O-O-O Nf6 |
| S3 | 8 000 000 | 30 | -14 | 1. f4 f5 2. g3 a5 3. c3 a4 4. Nac2 Nf6 |
| S4 | 32 000 000 | 37 | -2 | 1. c3 a5 2. f4 f5 3. g3 a4 4. Nac2 Nf6 |
| S5 | 128 000 000 | 45 | 0 | 1. f4 f5 2. c4 g6 3. d4 Nd6 4. Bd3 a5 |
| S6 | 400 000 000 | 55 | 0 | 1. f4 f5 2. c4 g6 3. d4 Nd6 4. Bd3 a5 |
| S7 | 800 000 000 | 54 | 0 | 1. f4 f5 2. c4 g6 3. d4 Nd6 4. Bd3 a5 |
| S8 | 1 600 000 000 | 55 | 0 | 1. f4 f5 2. c4 c6 3. d4 Nd6 4. Bd3 g5 |
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: 0
- Lc0 0.32.1: -17
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 +33, Black +65, difference -32. 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 | -5 | 44 | 1. c3 |
| 2 | -7 | 44 | 1. g3 f5 2. f4 g6 3. O-O-O Nf6 4. c3 a5 5. a4 Qg7 |
| 3 | -9 | 43 | 1. f4 f5 2. g3 g6 3. O-O-O Nf6 4. c3 a5 5. a4 Qg7 |
Main-line play-out
28 moves, 5 000 000 nodes per half-move. Largest |evaluation| 19 cp, mean 2.6 cp, final 0 cp.
| # | White | cp | Black | cp |
|---|---|---|---|---|
| 1 | f4 | -9 | f5 | -19 |
| 2 | g3 | -9 | a5 | -7 |
| 3 | a4 | -2 | g5 | +2 |
| 4 | c3 | +4 | gxf4 | -1 |
| 5 | Rxf4 | 0 | Nd6 | 0 |
| 6 | Bc5 | 0 | c6 | +9 |
| 7 | Bxd6 | +15 | Bxd6 | +13 |
| 8 | Rxf5 | 0 | Rxf5 | +11 |
| 9 | Bxf5 | +7 | Kc7 | 0 |
| 10 | O-O-O | +1 | Qg7 | +6 |
| 11 | Bc2 | +9 | e5 | 0 |
| 12 | Ng2 | 0 | Qf7 | +3 |
| 13 | Bb3 | +12 | Ne6 | +3 |
| 14 | Ne3 | 0 | Qe8 | 0 |
| 15 | Nc4 | 0 | Nc5 | 0 |
| 16 | Nxd6 | 0 | Kxd6 | 0 |
| 17 | Bc2 | 0 | Kc7 | 0 |
| 18 | Qg1 | 0 | b6 | +1 |
| 19 | Qf2 | 0 | Kb7 | 0 |
| 20 | Qe3 | 0 | Qe6 | 0 |
| 21 | d4 | 0 | exd4 | 0 |
| 22 | Qxe6 | 0 | Nxe6 | 0 |
| 23 | Nb3 | 0 | dxc3 | 0 |
| 24 | Rxd7+ | 0 | Kc8 | 0 |
| 25 | Re7 | 0 | Kd8 | 0 |
| 26 | Rb7 | 0 | Kc8 | 0 |
| 27 | Re7 | 0 | Kd8 | 0 |
| 28 | Rb7 | 0 | Kc8 | 0 |
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 39.0 moves.
Sample game 1 (1/2-1/2)
1. c3 c6 2. f4 a5 3. e4 a4 4. Bc5 Bd6 5. Qg1 f5 6. e5 Bxc5 7. Qxc5 d6 8. Qe3 Nc7 9. Nac2 g6 10. d4 Bc4 11. Nd3 dxe5 12. fxe5 Nde6 13. Na3 Ba6 14. Re1 Nd5 15. Qd2 f4 16. O-O-O O-O-O 17. Bc2 Qg7 18. c4 Ndc7 19. d5 cxd5 20. Qa5 Qh6 21. Kb1 dxc4 22. Nb4 f3 23. gxf3 Rxd1+ 24. Rxd1 Rd8 25. Nxa6 Rxd1+ 26. Bxd1 bxa6 27. Nxc4 Qxh2 28. a3 Qf4 29. Qxa4 Kd8 30. Ka2 Qd4 31. Bc2 Nb5 32. Qb3 Kd7 33. a4 Nbc7 34. Be4 Ke8 35. Na5 Qxe5 36. Nc6 Qf6 37. a5 Kf8 38. Qb6 h5 39. Nb4 Qd4 40. Bxg6 h4 41. Bf5 Kf7 42. Qxd4 Nxd4 43. Bh3 Kf6 44. f4 Nde6 45. Nd3 Nd5 46. Bf1 Ne3 47. Bh3 Nd5 48. Kb3 Ndxf4 49. Nxf4 Nxf4 50. Bf1 Ke6 51. Bxa6 Kd6 52. Bb7 e6 53. Be4 Kc5 54. Ka4 Nd5 55. Bg2 Nb4 56. Kb3 Nd5 57. Ka4 Nb4 58. Bh3 e5 59. Kb3 Nd5 60. a6 Kb6
Sample game 2 (1/2-1/2)
1. f4 a5 2. c3 a4 3. e4 f5 4. exf5 c6 5. O-O-O Bc4 6. d3 Bd5 7. c4 Bf7 8. Nac2 Bh5 9. Rd2 Rxf5 10. g3 g5 11. Bd4 e5 12. Bxe5 Bxe5 13. fxe5 Rxe5 14. Qg1 Re2 15. Qd4 Rxd2 16. Kxd2 Qxd4 17. Nxd4 Kc7 18. Bc2 Ng7 19. Kc3 Bg6 20. Nef3 Ra5 21. Re1 Nf7 22. Nd2 b6 23. a3 Nd6 24. Bd1 Nf7 25. Ne4 h5 26. Nf6 Kd6 27. Nf3 Rf5 28. Ne4+ Kc7 29. Nd4 Ra5 30. Nf6 Kd6 31. Rf1 Kc7
Play it in the CH960 app
Open this page on a phone with the CH960 app installed, or find pair 284 / 764 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): rbknnrbq/pppppppp/8/8/8/8/PPPPPPPP/NBRKNRBQ w CFaf - 0 1
Mirror twin: White 588 vs Black 396 — a different position (castling squares are fixed), not evaluated by the search.