White 447 vs Black 4
White starts from setup 447 (RNNQKRBB), Black from setup 4 (QBBNNRKR). In the encyclopedia since 2026-10-21: 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. 447, Black by position No. 4, 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 #447 R N N Q K R B B
King on the e-file, rooks on a, f, queen on the d-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 447 gives White +0.28 (Morozov Balance Index, 400M nodes).
Black: setup #4 Q B B N N R K R
King on the g-file, rooks on f, h, queen on the a-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 4 gives White +0.38 (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 58.
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 -30 and +4 cp before settling — a reminder that quick engine checks overstate imbalances, and why the ladder goes deep.
For comparison, in symmetric play setup 447 gives White +0.28 and setup 4 gives White +0.38: 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 largest swing was 35 cp (around move 5), ending at 0 cp — the position stays playable for both sides.
Engine self-play agrees: 10 games, White won 0, drawn 9, Black won 1 (White scored 45.0%).
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 | 13 | -30 | 1. e4 d5 2. exd5 c6 3. dxc6 Nf6 4. g3 Nxc6 |
| S1 | 500 000 | 19 | +2 | 1. g3 d5 2. Bxd5 Bh3 3. d4 c6 4. Bb3 h5 |
| S2 | 2 000 000 | 21 | -5 | 1. Nc3 b6 2. e4 f5 3. g3 Bb7 4. d3 Rf7 |
| S3 | 8 000 000 | 29 | +4 | 1. g3 c6 2. c4 d5 3. cxd5 cxd5 4. Nc3 d4 |
| S4 | 32 000 000 | 39 | 0 | 1. g3 d5 2. Bxd5 Bh3 3. f4 Bxf1 4. Kxf1 c6 |
| S5 | 128 000 000 | 42 | -6 | 1. g3 d5 2. Bxd5 Bh3 3. f4 c6 4. Bb3 Bxf1 |
| S6 | 400 000 000 | 49 | 0 | 1. g3 d5 2. Bxd5 Bh3 3. f4 c6 4. Bb3 Bxf1 |
| S7 | 800 000 000 | 54 | 0 | 1. g3 d5 2. Bxd5 Bh3 3. f4 c6 4. Bb3 Bxf1 |
| S8 | 1 600 000 000 | 58 | 0 | 1. g3 d5 2. Bxd5 Bh3 3. f4 c6 4. Bb3 Bxf1 |
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: +10
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 +48, Black +79, difference -31. 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, 9 draws, 1 Black wins, White scored 45.0%.
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. g3 d5 2. Bxd5 Bh3 3. f4 c6 4. Bb3 Bxf1 5. Kxf1 e5 |
| 2 | -18 | 40 | 1. Nc3 b6 2. d4 Nf6 3. Nd3 Bb7 4. f3 c5 5. dxc5 Ne6 |
| 3 | -24 | 39 | 1. c4 b6 2. d4 Nf6 3. Nd3 c5 4. d5 e6 5. g3 Bb7 |
Main-line play-out
30 moves, 5 000 000 nodes per half-move. Largest |evaluation| 35 cp, mean 6.8 cp, final 0 cp.
| # | White | cp | Black | cp |
|---|---|---|---|---|
| 1 | g3 | +4 | d5 | -15 |
| 2 | Bxd5 | -7 | Bh3 | -12 |
| 3 | Nc3 | -14 | Nf6 | -31 |
| 4 | Bh1 | -24 | c6 | -25 |
| 5 | f4 | -35 | Bxf1 | -22 |
| 6 | Kxf1 | -33 | e5 | -23 |
| 7 | Nd3 | -23 | exf4 | -15 |
| 8 | gxf4 | -34 | Re8 | -12 |
| 9 | e4 | -12 | Ne6 | -4 |
| 10 | e5 | +10 | Nd7 | +2 |
| 11 | Qg4 | -8 | g6 | +3 |
| 12 | f5 | 0 | Nef8 | 0 |
| 13 | e6 | -1 | fxe6 | 0 |
| 14 | Re1 | 0 | h5 | 0 |
| 15 | Qg2 | 0 | Bc7 | 0 |
| 16 | fxg6 | 0 | Re7 | +5 |
| 17 | Ke2 | +14 | Qe8 | 0 |
| 18 | Kd1 | 0 | Qxg6 | 0 |
| 19 | Qe2 | 0 | Rg7 | 0 |
| 20 | Bxa7 | 0 | Qg4 | 0 |
| 21 | Bf3 | 0 | Qh3 | 0 |
| 22 | Qe3 | 0 | b6 | 0 |
| 23 | a4 | -6 | c5 | -8 |
| 24 | b3 | 0 | Rh6 | 0 |
| 25 | Qxh6 | 0 | Qxf3+ | 0 |
| 26 | Kc1 | 0 | Bxh2 | 0 |
| 27 | Ne4 | 0 | Bg3 | 0 |
| 28 | Nxg3 | 0 | Qxg3 | 0 |
| 29 | a5 | 0 | c4 | -7 |
| 30 | bxc4 | 0 | Nc5 | 0 |
Engine self-play
10 games at 500 000 nodes per move: White won 0, drawn 9, Black won 1 — White scored 45.0%. Average length 42.8 moves.
Sample game 1 (1/2-1/2)
1. Nc3 b6 2. e4 f5 3. d3 Rf7 4. f4 O-O 5. e5 c5 6. Qe2 d6 7. d4 Nc7 8. Nd3 c4 9. Nf2 b5 10. g3 Bb7 11. Bxb7 Qxb7 12. Nfd1 e6 13. Ne3 Rd7 14. Qg2 Qa6 15. Kf2 b4 16. Ne2 Qb5 17. a3 a5 18. axb4 axb4 19. Rfd1 dxe5 20. dxe5 Rxd1 21. Rxd1 Ba7 22. Ke1 Nc6 23. Rd6 Ne7 24. Nd1 Bxg1 25. Qxg1 Nc8 26. Nd4 Qb7 27. Rc6 Nb5 28. Rc5 Nxd4 29. Qxd4 Ne7 30. Kf2 c3 31. Qd6 Qe4 32. Ne3 cxb2 33. Qxe7 b1=Q 34. Rc7 Qxe3+ 35. Kxe3 Qg1+ 36. Kd3 Qd1+ 37. Ke3 Qg1+ 38. Kd2 Qxh2+ 39. Kc1 Qg1+ 40. Kb2 Qd4+
Sample game 2 (0-1)
1. Nc3 b6 2. d3 f5 3. e4 Rf7 4. g3 O-O 5. f4 e5 6. fxe5 c5 7. d4 Bb7 8. Nb3 cxd4 9. Bxd4 Ne6 10. Qe2 Nxd4 11. Nxd4 Bxe5 12. O-O-O f4 13. gxf4 Rxf4 14. Rxf4 Bxf4+ 15. Kb1 Nf6 16. h4 Be5 17. h5 Qe8 18. h6 g6 19. a3 Kh8 20. Qd2 Qb8 21. Ndb5 Bf4 22. Qe1 Bc6 23. Qe2 Ng8 24. Bg2 Bxh6 25. Bh3 Nf6 26. Nd6 Ne8 27. Nc4 Qc7 28. a4 Bg7 29. Nb5 Qb7 30. Qe1 a6 31. Nd4 Bxe4 32. Bxd7 Nf6 33. Be6 Qc7 34. Nd2 Bb7 35. Bb3 Re8 36. Ne6 Qc6 37. Nf1 Bc8 38. Rd8 Bxe6 39. Rxe8+ Qxe8 40. Qxe6 Qxe6 41. Bxe6 h5 42. b4 Bf8 43. c3 g5 44. Kc2 g4 45. Kd3 Kg7 46. Ne3 Kh6 47. Bc8 a5 48. Nc4 axb4 49. cxb4 Bxb4 50. Nxb6 Kg5 51. Ke2 Ne4 52. Nd5 Ba5 53. Ne3 Bb6 54. a5 Bxa5 55. Bxg4 hxg4 56. Nxg4 Kxg4 57. Ke3 Nd6 58. Kd4 Nb5+ 59. Kd5 Na7 60. Kd4 Kf4 61. Kd5 Bc7
Play it in the CH960 app
Open this page on a phone with the CH960 app installed, or find pair 447 / 4 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): qbbnnrkr/pppppppp/8/8/8/8/PPPPPPPP/RNNQKRBB w AFfh - 0 1
Mirror twin: White 704 vs Black 958 — a different position (castling squares are fixed), not evaluated by the search.