White 103 vs Black 673
White starts from setup 103 (QNBRNKRB), Black from setup 673 (BQRBKNNR). 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 #103 Q N B R N K R B
King on the f-file, rooks on d, g, queen on the a-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 103 gives White +0.17 (Morozov Balance Index, 400M nodes).
Black: setup #673 B Q R B K N N R
King on the e-file, rooks on c, h, queen on the b-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 673 gives White +0.13 (Morozov Balance Index, 400M nodes).
Why the pair is balanced
At the deepest rungs (400M, 800M, 1.6B nodes) the evaluation never left ±2 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 ±14 cp.
For comparison, in symmetric play setup 103 gives White +0.17 and setup 673 gives White +0.13: 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 g3; the alternatives are slightly weaker but keep the game level.
Along the engine's main line (30 moves) the evaluation stayed within ±25 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 | 14 | +14 | 1. b3 Nf6 2. g3 b5 3. Bxa8 Qxa8 4. Bb2 h5 |
| S1 | 500 000 | 17 | +2 | 1. b3 Nf6 2. g3 b5 3. Bxa8 Qxa8 4. Bb2 e6 |
| S2 | 2 000 000 | 21 | -11 | 1. b3 Nf6 2. g4 h6 3. d4 c5 4. d5 Ng6 |
| S3 | 8 000 000 | 29 | -14 | 1. b3 Nf6 2. g3 b5 3. Bxa8 Qxa8 4. Bb2 Ne6 |
| S4 | 32 000 000 | 34 | -3 | 1. g3 Nf6 2. c4 b6 3. Bxa8 Qxa8 4. b3 d5 |
| S5 | 128 000 000 | 43 | 0 | 1. g3 Nf6 2. c4 b6 3. Bxa8 Qxa8 4. b3 Ng6 |
| S6 | 400 000 000 | 51 | 0 | 1. g3 Nf6 2. c4 b6 3. Bxa8 Qxa8 4. b3 Ng6 |
| S7 | 800 000 000 | 55 | +2 | 1. g3 b5 2. Bxa8 Qxa8 3. b4 Nf6 4. a4 a6 |
| S8 | 1 600 000 000 | 54 | +2 | 1. g3 b5 2. Bxa8 Qxa8 3. b4 Nf6 4. a4 a6 |
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: -7
- Lc0 0.32.1: +7
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 +58, Black +123, difference -65. 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 | 42 | 1. g3 b5 2. Bxa8 Qxa8 3. b4 Nf6 4. a4 a6 5. Bb2 Ne6 |
| 2 | -18 | 42 | 1. b3 Nf6 2. Bb2 b5 3. g3 Bxh1 4. Bxf6 gxf6 5. Rxh1 h5 |
| 3 | -27 | 41 | 1. b4 Nf6 2. g3 b6 3. Bxa8 Qxa8 4. Bb2 e6 5. d3 Ng6 |
Main-line play-out
30 moves, 5 000 000 nodes per half-move. Largest |evaluation| 25 cp, mean 7.0 cp, final 0 cp.
| # | White | cp | Black | cp |
|---|---|---|---|---|
| 1 | b3 | -17 | Nf6 | -5 |
| 2 | g3 | -6 | b5 | -8 |
| 3 | Bxa8 | -8 | Qxa8 | -9 |
| 4 | Bb2 | -3 | Ne6 | +4 |
| 5 | a4 | +5 | c6 | -7 |
| 6 | Nf3 | -11 | O-O | -14 |
| 7 | d3 | -9 | a6 | -12 |
| 8 | O-O | +4 | Qb7 | +4 |
| 9 | c4 | +13 | d6 | +7 |
| 10 | Bc3 | +5 | b4 | 0 |
| 11 | Bb2 | +24 | Nc5 | +21 |
| 12 | Nbd2 | +25 | e5 | +16 |
| 13 | d4 | +13 | exd4 | +10 |
| 14 | Bxd4 | +10 | Ne6 | +10 |
| 15 | Bb2 | +6 | Re8 | 0 |
| 16 | e4 | 0 | d5 | +14 |
| 17 | Qb1 | +19 | Nc5 | +12 |
| 18 | e5 | +8 | Nfe4 | +10 |
| 19 | Qc2 | +9 | Bb6 | +1 |
| 20 | Nh4 | 0 | Qe7 | +13 |
| 21 | Bd4 | +7 | Nxd2 | +4 |
| 22 | Rxd2 | +9 | Qe6 | +6 |
| 23 | a5 | +10 | Ba7 | 0 |
| 24 | Ng2 | 0 | Qe7 | +7 |
| 25 | Bxc5 | +2 | Qxc5 | 0 |
| 26 | cxd5 | -2 | cxd5 | 0 |
| 27 | Qxc5 | 0 | Rxc5 | 0 |
| 28 | Nf4 | 0 | Rxa5 | 0 |
| 29 | Nxd5 | 0 | Bc5 | 0 |
| 30 | Re1 | 0 | Ra3 | 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 37.2 moves.
Sample game 1 (1/2-1/2)
1. g3 b5 2. Bxa8 Qxa8 3. b3 Nf6 4. Bb2 Ne6 5. c4 bxc4 6. bxc4 O-O 7. d3 Re8 8. Nf3 Nc5 9. Bxf6 exf6 10. Nc3 f5 11. Qc1 g6 12. h4 h5 13. Kg2 Bf6 14. Nd5 Bg7 15. e3 d6 16. Rge1 c6 17. Nc3 Ne6 18. d4 Rb8 19. Qd2 Qb7 20. e4 Qb2 21. exf5 Qxd2 22. Rxd2 gxf5 23. Ne2 Nf8 24. Kf1 a5 25. Rd3 a4 26. Nf4 Rxe1+ 27. Kxe1 Ne6 28. Nxh5 Bh8 29. Nf4 Nxf4 30. gxf4 Bf6 31. Rd2 Re8+ 32. Re2 Re4 33. Rxe4 fxe4 34. Nh2 Kg7 35. Nf1 Bxh4 36. Ke2 Kg6 37. f3 exf3+ 38. Kxf3 Be1 39. Ke4 Bf2 40. d5 c5 41. Kf3 Bd4 42. Nd2 f5 43. Kg2 Kh5 44. Kg3 Kg6 45. Kg2 Bf6 46. Nf3 Kh5 47. Kh3 Kg6 48. Ne1 Bb2 49. Ng2 Bf6
Sample game 2 (1/2-1/2)
1. b3 Nf6 2. Bb2 b5 3. Bxf6 exf6 4. g3 Ne6 5. Bxa8 Qxa8 6. d4 d5 7. c4 bxc4 8. bxc4 c5 9. cxd5 Qxd5 10. Nc3 Qh5 11. Nb5 O-O 12. d5 Qh3+ 13. Rg2 Nd4 14. Nxd4 cxd4 15. Qxd4 Bb6 16. Qf4 Qd7 17. Rg1 Qh3+ 18. Rg2 Qd7 19. Rg1 Rc5 20. Qf3 Qh3+ 21. Qg2 Qd7 22. Qf3 Qh3+ 23. Qg2 Qc8 24. Qf3
Play it in the CH960 app
Open this page on a phone with the CH960 app installed, or find pair 103 / 673 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): bqrbknnr/pppppppp/8/8/8/8/PPPPPPPP/QNBRNKRB w DGch - 0 1
Mirror twin: White 850 vs Black 475 — a different position (castling squares are fixed), not evaluated by the search.