White 90 vs Black 958
White starts from setup 90 (NNRKBBRQ), Black from setup 958 (RKRNNBBQ). 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 #90 N N R K B B R Q
King on the d-file, rooks on c, g, 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 90 gives White +0.31 (Morozov Balance Index, 400M nodes).
Black: setup #958 R K R N N B B Q
King on the b-file, rooks on a, c, queen on the h-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 958 gives White +0.42 (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 49.
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 90 gives White +0.31 and setup 958 gives White +0.42: 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 g4, and 3 first moves are within 10 cp of it.
Along the engine's main line (30 moves) the largest swing was 50 cp (around move 18), ending at -10 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 | 15 | -1 | 1. g4 a5 2. c4 h5 3. h3 f6 4. d3 Nd6 |
| S1 | 500 000 | 19 | +2 | 1. g4 h5 2. h3 f6 3. c4 e5 4. d3 a5 |
| S2 | 2 000 000 | 23 | -12 | 1. g4 e5 2. c4 h5 3. g5 f5 4. gxf6 Nxf6 |
| S3 | 8 000 000 | 26 | +15 | 1. c4 h5 2. d3 f5 3. g3 a5 4. Nc2 e5 |
| S4 | 32 000 000 | 34 | +5 | 1. g4 h5 2. f4 a5 3. g5 f5 4. c4 g6 |
| S5 | 128 000 000 | 38 | 0 | 1. g4 e5 2. d3 h5 3. g5 f5 4. gxf6 Nxf6 |
| S6 | 400 000 000 | 45 | 0 | 1. g4 a5 2. c4 h5 3. f4 f6 4. d3 e5 |
| S7 | 800 000 000 | 47 | +4 | 1. g4 e5 2. d3 h5 3. g5 f5 4. gxf6 Nxf6 |
| S8 | 1 600 000 000 | 49 | 0 | 1. g4 h5 2. d4 hxg4 3. Nb3 Nc6 4. Rxg4 a5 |
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: -5
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 +34, Black +39, difference -5. 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 | +3 | 42 | 1. g4 e5 2. c4 h5 3. g5 f5 4. gxf6 Nxf6 5. d3 h4 |
| 2 | 0 | 42 | 1. d3 h5 2. c4 f5 3. g3 Nf6 4. Nc3 a5 5. f4 e6 |
| 3 | 0 | 42 | 1. c4 h5 2. g3 f5 3. d3 Nf6 4. Nc3 a5 5. f4 e6 |
Main-line play-out
30 moves, 5 000 000 nodes per half-move. Largest |evaluation| 50 cp, mean 21.6 cp, final -10 cp.
| # | White | cp | Black | cp |
|---|---|---|---|---|
| 1 | c4 | +15 | h5 | -23 |
| 2 | d3 | -6 | f5 | -4 |
| 3 | g3 | -3 | Nf6 | -12 |
| 4 | Nc3 | -15 | a5 | -12 |
| 5 | f4 | -14 | g6 | 0 |
| 6 | Nc2 | +13 | Bg7 | +17 |
| 7 | h3 | +18 | a4 | +1 |
| 8 | Qf3 | +7 | Bf7 | -7 |
| 9 | Bg2 | -18 | c6 | -18 |
| 10 | e4 | -33 | fxe4 | -32 |
| 11 | dxe4 | -39 | Bxc4 | -39 |
| 12 | Ne3 | -26 | Be6 | -31 |
| 13 | Bd2 | -34 | d6 | -40 |
| 14 | O-O | -38 | Ra6 | -44 |
| 15 | Rcd1 | -31 | Nd7 | -27 |
| 16 | e5 | -44 | Rc7 | -40 |
| 17 | Qe2 | -47 | Nf7 | -49 |
| 18 | Kh2 | -47 | dxe5 | -50 |
| 19 | f5 | -33 | gxf5 | -33 |
| 20 | Nxf5 | -31 | Nf6 | -34 |
| 21 | Nxg7 | -26 | Qxg7 | -24 |
| 22 | Ne4 | -20 | c5 | -20 |
| 23 | Rc1 | -22 | Nxe4 | -17 |
| 24 | Qxe4 | -7 | Nd6 | -13 |
| 25 | Qe1 | -13 | e4 | -5 |
| 26 | Bxe4 | -7 | Bxh3 | -14 |
| 27 | Kxh3 | -10 | Qg4+ | -17 |
| 28 | Kg2 | -14 | Qxe4+ | -12 |
| 29 | Qxe4 | -7 | Nxe4 | -5 |
| 30 | Bf4 | -9 | Nd6 | -10 |
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 36.0 moves.
Sample game 1 (1/2-1/2)
1. g4 a5 2. d4 g6 3. O-O-O f5 4. Nb3 fxg4 5. Nc5 Nf6 6. Nc3 e5 7. dxe5 Bxc5 8. exf6 Qxf6 9. Rxg4 Bxf2 10. Bxf2 Qxf2 11. Rg3 d6 12. Bh3 Ne6 13. Rf1 Qc5 14. Qd5 Re8 15. Qd2 Qc6 16. Re3 Nc5 17. a3 a4 18. Rxe8+ Qxe8 19. Rf3 Qd8 20. Qd4 c6 21. Kd1 Qe7 22. Re3 Qd8 23. Rf3 Kc7 24. Qg7+ Kb6 25. Rf4 Qg5 26. Rg4 Qe3 27. Bg2 d5 28. Rb4+ Ka7 29. Nxa4 Qg1+ 30. Kd2 Re8 31. Bf3 Qe3+ 32. Kd1 Re7 33. Qxg8 Qg1+ 34. Kd2 Qe3+ 35. Kd1 Qg1+ 36. Kd2 Qe3+
Sample game 2 (1/2-1/2)
1. g4 e5 2. O-O-O a5 3. e3 h5 4. h3 f6 5. c4 d5 6. cxd5 c6 7. d4 exd4 8. exd4 Bxd5 9. f3 hxg4 10. Qh2+ Bd6 11. Bg3 Qh6+ 12. Rd2 f5 13. Bxd6+ Qxd6 14. fxg4 f4 15. Nc3 a4 16. a3 Ne6 17. h4 Nf6 18. Nc2 c5 19. Kb1 Ra5 20. Bd3 cxd4 21. Nxd5 Nxd5 22. Bf5 g6 23. Bxe6 Qxe6 24. Re1 Qxg4 25. Nxd4 Rac5 26. Qf2 Qh3 27. Rdd1 Rc4 28. Nc6+ R8xc6 29. Rxd5 b6 30. Qd2 Kb7 31. Rd7+ Ka6 32. Qd5 f3 33. Rf7 Qxh4 34. Re8 R4c5 35. Qd3+ Qc4 36. Ra8+ Kb5 37. Rxf3 Rg5 38. Qxc4+ Rxc4 39. Re3 Rh4 40. Ka2 Rg1 41. Re5+ Kc6 42. Ra7 Rf4 43. Re6+
Play it in the CH960 app
Open this page on a phone with the CH960 app installed, or find pair 90 / 958 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): rkrnnbbq/pppppppp/8/8/8/8/PPPPPPPP/NNRKBBRQ w CGac - 0 1
Mirror twin: White 869 vs Black 4 — a different position (castling squares are fixed), not evaluated by the search.
Current place in the live balance rating: 302.