White 281 vs Black 683
White starts from setup 281 (NRKBBNRQ), Black from setup 683 (QRKNBNRB). In the encyclopedia since 2026-10-21: neither side starts better.
|cp| ≤ 5 at 400M · 800M · 1.6B · 4B nodes, two independent engines agree
The two setups
White: setup #281 N R K B B N R Q
King on the c-file, rooks on b, g, queen on the h-file; no bishop starts in a corner; the knights start away from the centre. The king already stands on a castled square, so one of the castlings moves only the rook. In ordinary symmetric play setup 281 gives White +0.16 (Morozov Balance Index, 400M nodes).
Black: setup #683 Q R K N B N R B
King on the c-file, rooks on b, g, queen on the a-file; one 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 683 gives White +0.41 (Morozov Balance Index, 400M nodes).
Why the pair is balanced
At the deepest rungs (400M, 800M, 1.6B nodes) the evaluation never left ±4 centipawns of equality, 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 agreed early: from 100 thousand to 128 million nodes the evaluation never left ±15 cp.
For comparison, in symmetric play setup 281 gives White +0.16 and setup 683 gives White +0.41: with both sides on the same rank the first move is worth something. In this pair Black's own setup absorbs that edge completely.
The engine's first choice for White is g4; the alternatives are slightly weaker but keep the game level.
Along the engine's main line (30 moves) the largest swing was 61 cp (around move 12), ending at +51 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 | -1 | 1. g4 d6 2. d4 c5 3. c3 cxd4 4. cxd4 |
| S1 | 500 000 | 19 | -2 | 1. g4 a5 2. c3 g6 3. f4 d5 4. Qxd5 Bc6 |
| S2 | 2 000 000 | 27 | -15 | 1. g4 g6 2. c3 e5 3. e3 d5 4. d4 b5 |
| S3 | 8 000 000 | 31 | -5 | 1. g4 g6 2. c3 e5 3. e3 d5 4. d4 b5 |
| S4 | 32 000 000 | 34 | +6 | 1. g4 e6 2. d4 d5 3. e3 a5 4. f4 b5 |
| S5 | 128 000 000 | 42 | 0 | 1. g4 g6 2. c3 e5 3. Nc2 a5 4. d4 exd4 |
| S6 | 400 000 000 | 46 | 0 | 1. g4 g6 2. c3 e5 3. Nc2 Nfe6 4. e3 b5 |
| S7 | 800 000 000 | 50 | -4 | 1. g4 g6 2. c3 e5 3. Nc2 d5 4. d4 exd4 |
| S8 | 1 600 000 000 | 55 | 0 | 1. g4 g6 2. c3 d5 3. Qxd5 Bc6 4. Qa5 e5 |
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: -4
- Lc0 0.32.1: -16
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 +20, difference +13. 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 | 40 | 1. g4 g6 2. c3 e5 3. e3 d5 4. d4 b5 5. Nc2 a5 |
| 2 | -18 | 40 | 1. g3 g6 2. c3 a5 3. e4 b5 4. Ne3 b4 5. d4 d6 |
| 3 | -30 | 40 | 1. d4 g6 2. c3 b5 3. g4 d5 4. f4 b4 5. e3 a5 |
Main-line play-out
30 moves, 5 000 000 nodes per half-move. Largest |evaluation| 61 cp, mean 39.2 cp, final +51 cp.
| # | White | cp | Black | cp |
|---|---|---|---|---|
| 1 | g4 | -12 | g6 | -6 |
| 2 | c3 | +6 | e5 | -1 |
| 3 | e3 | -11 | d5 | -17 |
| 4 | d4 | -18 | b5 | -10 |
| 5 | Nc2 | -6 | Nd7 | 0 |
| 6 | f4 | +32 | e4 | +40 |
| 7 | Nd2 | +44 | Bf6 | +51 |
| 8 | Qg2 | +48 | Be7 | +44 |
| 9 | Be2 | +45 | a5 | +42 |
| 10 | Qh3 | +42 | Rh8 | +51 |
| 11 | f5 | +51 | Qc6 | +41 |
| 12 | Bg3 | +55 | Nb6 | +61 |
| 13 | Nb3 | +55 | Nb7 | +57 |
| 14 | fxg6 | +58 | Qxg6 | +44 |
| 15 | Qf1 | +44 | a4 | +42 |
| 16 | Nd2 | +38 | Qc6 | +41 |
| 17 | Qf4 | +43 | Qd6 | +35 |
| 18 | Qf1 | +30 | Qc6 | +30 |
| 19 | g5 | +52 | h6 | +39 |
| 20 | gxh6 | +38 | Rxh6 | +44 |
| 21 | Qf4 | +47 | Rg6 | +35 |
| 22 | Bg4+ | +35 | Bd7 | +36 |
| 23 | Qxf7 | +45 | Qf6 | +38 |
| 24 | Qxf6 | +47 | Bxf6 | +51 |
| 25 | Bh5 | +47 | Rg7 | +37 |
| 26 | O-O-O | +54 | O-O-O | +52 |
| 27 | Rdf1 | +55 | Be7 | +53 |
| 28 | Be5 | +46 | Rh7 | +47 |
| 29 | Rf7 | +54 | Rxf7 | +52 |
| 30 | Bxf7 | +49 | Nd6 | +51 |
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 37.8 moves.
Sample game 1 (1/2-1/2)
1. d4 d5 2. g4 b5 3. Nb3 g6 4. e3 a5 5. Bf3 a4 6. Nc5 e6 7. Nd3 Nd7 8. Ng3 c5 9. Ne2 Nc6 10. Bd2 cxd4 11. exd4 Nxd4 12. Nxd4 Bxd4 13. O-O f6 14. Rbc1 Ne5 15. g5 Bc6 16. gxf6 Nxf3+ 17. Qxf3 O-O 18. Bg5 e5 19. Qg3 e4 20. Ne5 Bxe5 21. Qxe5 Rbe8 22. Qg3 Qc8 23. Rcd1 Qe6 24. Rd4 h6 25. Bxh6 Rxf6 26. Re1 Rf5 27. Kh1 Rh5 28. Be3 Rh3 29. Qf4 Rf8 30. Qc7 Rc8 31. Qf4 Rf8 32. Qc7 Rc8 33. Qf4
Sample game 2 (1/2-1/2)
1. d4 b5 2. c3 a5 3. g4 d5 4. Nc2 Bc6 5. f4 Nd7 6. Nd2 b4 7. Bh4 bxc3 8. bxc3 Rxb1+ 9. Kxb1 e6 10. Ka1 g6 11. e3 Nb7 12. Bf3 Nd6 13. c4 Qa6 14. Rc1 Nf6 15. c5 Nfe4 16. Nxe4 Nxe4 17. Qd1 Nxc5 18. Ne1 Ba4 19. Qd2 Ne4 20. Bxe4 dxe4 21. Qc3 Qb6 22. Nc2 f5 23. Na3 Kd7 24. g5 Qc6 25. Nc4 Bb5 26. Nxa5 Qxc3+ 27. Rxc3 Bg7 28. Nb3 Ra8 29. Nc5+ Kd6 30. Nb7+ Kd7 31. Nc5+ Kd6 32. Nb7+ Kd7
Play it in the CH960 app
Open this page on a phone with the CH960 app installed, or find pair 281 / 683 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): qrknbnrb/pppppppp/8/8/8/8/PPPPPPPP/NRKBBNRQ w BGbg - 0 1
Mirror twin: White 585 vs Black 465 — a different position (castling squares are fixed), not evaluated by the search.