♚Open in the app

White 221 vs Black 518

White starts from setup 221 (NQRBKNBR), Black from setup 518 (RNBQKBNR). In the encyclopedia since 2026-10-15: 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. 221, Black by position No. 518, mirrored on their own side; the pawns stand as usual.

Black · No. 518
White · No. 221

Moves as usual: White starts; everything else follows the Chess960 rules.

Printable card (A4, landscape)

The two setups

White: setup #221 N Q R B K N B R

King on the e-file, rooks on c, h, queen on the b-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 221 gives White +0.22 (Morozov Balance Index, 400M nodes).

Black: setup #518 R N B Q K B N R

King on the e-file, rooks on a, h, queen on the d-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 518 gives White +0.11 (Morozov Balance Index, 400M nodes).

Why the pair is balanced

At the deepest rungs (400M, 800M, 1.6B nodes) the evaluation never left ±5 centipawns of equality, reaching search depth 53.

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 ±12 cp.

For comparison, in symmetric play setup 221 gives White +0.22 and setup 518 gives White +0.11: 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 e4, and 3 first moves are within 10 cp of it.

Along the engine's main line (30 moves) the evaluation stayed within ±8 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 13 +12 1. c4 Nf6 2. f3 e6 3. e4 d5 4. cxd5 exd5
S1 500 000 16 +3 1. c4 Nf6 2. f3 e6 3. Bf2 c5 4. e4 g6
S2 2 000 000 24 -4 1. f3 Nf6 2. c4 e6 3. d4 Be7 4. e4 d6
S3 8 000 000 33 +5 1. f3 Nf6 2. e4 e5 3. c3 d5 4. Ne3 c6
S4 32 000 000 39 0 1. e4 e5 2. f3 Nf6 3. c3 d5 4. Ne3 c6
S5 128 000 000 41 +7 1. e4 e6 2. f3 d5 3. Nb3 a5 4. d4 a4
S6 400 000 000 47 +5 1. e4 e5 2. c3 Nf6 3. Ne3 d5 4. exd5 Nxd5
S7 800 000 000 51 +1 1. e4 e5 2. c3 Nf6 3. Ne3 d5 4. exd5 Nxd5
S8 1 600 000 000 53 0 1. c3 d5 2. f3 Nf6 3. e4 Nbd7 4. Bf2 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: -3
  • Lc0 0.32.1: +13

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 +5, Black +11, difference -6. 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 -16 42 1. e4 e6 2. f3 d5 3. c3 dxe4 4. fxe4 Ne7 5. d4 Ng6
2 -19 43 1. e4 e6
3 -19 42 1. h3 e5 2. Bh2 Nc6 3. c4 Nf6 4. Ne3 d6 5. b4 g6

Main-line play-out

30 moves, 5 000 000 nodes per half-move. Largest |evaluation| 8 cp, mean 0.4 cp, final 0 cp.

# White cp Black cp
1 f3 +6 Nf6 +8
2 e4 +3 e5 0
3 c3 +3 d5 -1
4 Ne3 -1 dxe4 0
5 fxe4 0 Nbd7 0
6 Bf2 0 Nc5 0
7 Bc2 0 Ne6 0
8 O-O 0 h5 0
9 d4 0 Ng4 0
10 Nxg4 0 hxg4 0
11 Rce1 0 a5 0
12 Qd1 0 exd4 0
13 cxd4 0 Bb4 0
14 Qxg4 0 Bxe1 0
15 Rxe1 0 O-O 0
16 e5 0 a4 0
17 Qe4 0 Ng5 0
18 Qd3 0 f5 0
19 Bxa4 0 Be6 0
20 Bb3 0 Bxb3 0
21 Qxb3+ 0 Kh7 0
22 Nc2 0 Ne4 0
23 Qh3+ 0 Kg8 0
24 Qb3+ 0 Kh7 0
25 Rf1 0 Ra6 0
26 Be3 0 c5 0
27 d5 0 Rb6 0
28 Qd3 0 Rxb2 0
29 d6 0 Qe8 0
30 Bc1 0 Rxc2 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 35.1 moves.

Sample game 1 (1/2-1/2)
1. f3 g6 2. h4 h5 3. c4 a5 4. e4 Bg7 5. d4 e6 6. Be2 d5 7. Nd2 Nc6 8. Nc2 dxe4 9. Nxe4 Nh6 10. Rd1 Nf5 11. d5 exd5 12. Rxd5 Qe7 13. Bc5 Qe6 14. b3 b6 15. Ng5 Qf6 16. Ne4 Qe6 17. Ng5 Qf6 18. Ne4
Sample game 2 (1/2-1/2)
1. c4 g6 2. b4 Bg7 3. f3 Nf6 4. g4 d5 5. h4 d4 6. Ng3 O-O 7. Nb3 e5 8. g5 Nfd7 9. c5 Nc6 10. b5 Ne7 11. c6 bxc6 12. bxc6 Nb6 13. h5 Nf5 14. Ne4 Nd6 15. Nbc5 Nxe4 16. Nxe4 Bf5 17. Bh2 Bxe4 18. Qxe4 Qxg5 19. Qg4 Qf6 20. Bb3 g5 21. Rc5 Rae8 22. Kf2 Re7 23. Ra5 Nc8 24. Rg1 Nd6 25. Rxa7 Kh8 26. Kg2 Bh6 27. Bc2 Qg7 28. Bf5 f6 29. Rb1 Qf7 30. Rbb7 Nxf5 31. Qxf5 g4 32. fxg4 Qd5+ 33. Qf3 Qd6 34. Rb5 Bxd2 35. Bg3 Bf4 36. Qd5 Rd8 37. Qxd6 Rxd6 38. Bxf4 exf4 39. Kf3 Rxc6 40. Rb4 Rc3+ 41. Kxf4

Play it in the CH960 app

Open this page on a phone with the CH960 app installed, or find pair 221 / 518 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): rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/NQRBKNBR w CHah - 0 1

Mirror twin: White 648 vs Black 534 — a different position (castling squares are fixed), not evaluated by the search.

All pairs of the encyclopedia · All asymmetric setups