8

Has there been any analysis done on if white wins if it gets a free move in the starting position (let’s say e4)? Then, the game continues as normal.

From a human perspective it seems playing e4 and d4 as the first two moves would give white an advantage. But is it enough to be winning?

Edit: After thinking for over 1.4 trillion nodes, Sf dev thinks the following after white plays e4 and d4. Not conclusive, but worth further investigation with such a high evaluation.

67/92 1,495,349,364k -1.34

e7-e6 Bf1-d3 d7-d5 Nb1-c3 Bf8-b4 e4-e5 b7-b6 h2-h4 Qd8-d7 Qd1-g4 Bb4-f8 h4-h5 Bc8-a6 Bd3xa6 Nb8xa6 a2-a3 h7-h6 b2-b4 Ng8-e7 Nc3-d1 c7-c5 c2-c3 c5-c4 Ng1-e2 Na6-c7 a3-a4 O-O-O b4-b5 f7-f6 Bc1-f4 f6-f5 Qg4-f3 g7-g5 h5xg6/ep Ne7xg6 g2-g3 Bf8-e7 Nd1-e3 Be7-g5 Rh1-h5 Rd8-g8 Ke1-d2 Ng6xf4 Ne2xf4 Qd7-e8 Rh5-h1 a7-a6 b5xa6 Nc7xa6 Kd2-e2 Bg5xf4 Qf3xf4 Kc8-b7 Rh1-h4 Na6-c7 Qf4-f3 Qe8-c6 a4-a5

1
  • Since people are posting evaluations, imo Leela is the most natural tool here since it actually gives expected scores. It gives c6 d6 e5 and so forth for a sort of KID setup, however, the expected score is less than 70% so Leela thinks the drawing chances for Black may be a bit higher than the winning chances for White, take from that what you like.
    – koedem
    Apr 1 at 4:07

3 Answers 3

10

Unlikely.

This position has been analyzed in the Lichess cloud analysis with Stockfish 14.1. At depth 41, White has an advantage of +0.9 - which is significantly larger than the starting position, but not enough to consistently win.

Edit: Per this link, Stockfish wins against itself at bullet time controls about 25% of the time when the eval is +0.9. At long time controls - which would certainly be the case if we're looking for an objective evaluation of whether this position is winning - games are drawn more often, so it's probable that the position after 1.e4 with White to move is not winning for White.

6
  • +0.9 turns into +1.2 consistently at higher depths. If +1.2 turns into consistent +1.4, I'm inclined to say it's simply a horizon effect and white is winning. Mar 28 at 10:41
  • 3
    @edit_profile or it could turn into 0.00. See github.com/official-stockfish/Stockfish/pull/3582. At +0.9, in Stockfish vs. Stockfish games at bullet time controls, about 25% of games are won. There are fewer decisive games at longer time controls (which is almost the same as higher depth) as well.
    – Allure
    Mar 28 at 11:48
  • What percentage of games are won when the evaluation is +1.2? I think its worth investigating a bit more especially since the evaluation increases with depth. Mar 29 at 10:13
  • @edit_profile have you checked the link?
    – Allure
    Mar 29 at 11:14
  • @Allure Firstly, which plot are you looking at in the link to deduce that 25% of games at +0.9 eval are won? Secondly, it's worth noting that we have to understand what percentage are won by Stockfish when the true eval (or, let's say, high-depth eval after long time thinking) is +0.9 or +1.2, not when the eval Stockfish sees while playing bullet is +0.9 or +1.2 Apr 4 at 20:11
3

For humans especially, e4 d4 without the opponent playing any moves is extremely helpful for taking control over the center and to easily develop the knights and bishops fast.

The 1 extra move for humans will most probably result in a higher win rate as the headstart may crush black and make black feel uncomfortable due to the lack of possible development and space, but for computers its probably a different story and the advantage will not be that high

For computers, it is most likely that white cannot win consistently or with extremely high win rates, but it gives white a slight advantage, but we're humans, who wouldn't want another free move? :)

After computer analysis

  • At depth 49 (cannot go any higher), white is winning by + 0.3 advantage by Stockfish 14 cloud analysis here, if given no headstart

  • At depth 41 however, white is winning by + 0.9 advantage here if given the 1 move headstart

  • At depth 30-41+, the advantage goes over 1.0 and possibly can reach +1.4 advantage

  • Yet there is also a chance this can go down to 0.0, which highly unlikely though as white should win more games than when using black

The difference is quite significant, but black can still find ways to draw or even win if played well, so it is not very likely that it is winning for white, for computers it might be, for humans unlikely, as we always make mistakes

1
  • The question isn't asking about humans (or even computers, really, though the strongest computers are bound to be the best proxy), it's asking what the theoretical result is. In other words, when you say stuff like "most likely white cannot win consistently" or "there is a chance this can go down to 0.0", you are asked to give a percentage. Apr 4 at 20:15
3

According to stockfish 14.1 white wins.
Match conditions:

  • Each engine uses 2 threads.
  • Time Control is 90m + 30s increment.
  • Ponder ON
  • Start fen is rnbqkbnr/pppppppp/8/8/4P3/8/PPPP1PPP/RNBQKBNR w KQkq - 0 1

Game duration: "03:57:35"

What a wonderful game. White got the space advantage early but black managed to launch a dangerous kingside attack as in the king's indian defense opening formation. However white has somewhat neutralized it plus a queenside play.

[Event "white has e4 + side to move, ponder ON"]
[Site "?"]
[Date "2022.03.28"]
[Round "1"]
[White "Stockfish 14.1.2"]
[Black "Stockfish 14.1.1"]
[Result "1-0"]
[FEN "rnbqkbnr/pppppppp/8/8/4P3/8/PPPP1PPP/RNBQKBNR w KQkq - 0 1"]
[GameDuration "03:57:35"]
[PlyCount "128"]
[SetUp "1"]
[Termination "adjudication"]
[TimeControl "5400+30"]

1. d4 {+1.36/39 135.6s} c6 {-1.21/38 206.3s} 2. Bd3 {+1.40/38 0.002s}
d6 {-1.22/37 78.47s} 3. Nf3 {+1.27/38 27.62s} e5 {-1.11/37 52.92s}
4. O-O {+1.24/40 160.1s} Nd7 {-1.08/37 137.9s} 5. Re1 {+1.38/39 113.1s}
Ngf6 {-1.15/38 0.002s} 6. c4 {+1.28/35 102.7s} g6 {-1.20/40 280.2s}
7. d5 {+1.16/40 24.04s} Bg7 {-1.24/38 100.2s} 8. Nc3 {+1.29/41 255.5s}
O-O {-1.06/41 568.3s} 9. h3 {+1.44/36 342.7s} cxd5 {-1.35/42 547.7s}
10. cxd5 {+1.30/43 0.001s} Ne8 {-1.19/37 272.7s} 11. a4 {+1.50/34 133.8s}
h6 {-1.23/39 102.9s} 12. a5 {+1.44/38 271.1s} f5 {-1.48/39 0.002s}
13. exf5 {+1.50/37 68.86s} gxf5 {-1.38/36 25.74s} 14. Nd2 {+1.50/38 144.7s}
Qh4 {-1.34/41 201.0s} 15. Bf1 {+1.58/35 99.56s} f4 {-1.28/36 7.31s}
16. Nf3 {+1.49/39 188.5s} Qh5 {-1.30/39 122.7s} 17. Ra4 {+1.39/37 18.41s}
b6 {-1.47/37 118.8s} 18. b4 {+1.62/38 112.0s} a6 {-1.48/37 200.6s}
19. axb6 {+1.67/38 217.4s} Nxb6 {-1.49/39 0.001s} 20. Ra5 {+1.60/38 316.6s}
Bf6 {-1.50/39 153.2s} 21. Ne4 {+1.75/39 219.7s} Bf5 {-1.89/39 219.3s}
22. Ba3 {+1.89/41 438.7s} Rf7 {-1.79/45 0.002s} 23. b5 {+2.07/39 76.60s}
Bxe4 {-1.87/42 130.3s} 24. Rxe4 {+2.24/39 15.36s} axb5 {-2.26/46 174.9s}
25. Rxa8 {+2.23/42 0s} Nxa8 {-2.35/43 164.0s} 26. Bxb5 {+2.28/43 74.38s}
Rg7 {-2.54/40 58.74s} 27. Rb4 {+2.57/38 52.25s} Nac7 {-2.41/41 71.34s}
28. Bf1 {+2.65/39 171.2s} Qxh3 {-2.34/41 153.8s} 29. Qd3 {+2.43/39 19.80s}
Qh5 {-2.17/37 42.16s} 30. Rb8 {+2.47/37 105.3s} Be7 {-2.90/48 1058.0s}
31. Bb4 {+2.92/47 38.86s} Kf7 {-1.95/35 66.24s} 32. Rb7 {+3.16/40 38.41s}
Kg8 {-2.11/41 170.6s} 33. Rb8 {+3.04/44 33.88s} Kf7 {-2.48/37 77.99s}
34. Rb7 {+3.12/42 183.5s} Kg8 {-2.66/41 116.4s} 35. Nd2 {+3.07/46 124.1s}
Qg4 {-2.75/38 77.56s} 36. Rb8 {+3.04/44 18.64s} f3 {-2.97/39 78.67s}
37. g3 {+3.08/44 69.91s} h5 {-3.00/39 0.001s} 38. Ba5 {+3.10/43 124.1s}
h4 {-3.19/42 32.63s} 39. Bxc7 {+3.25/42 76.84s} Kf7 {-3.19/42 36.22s}
40. Ba5 {+3.23/42 137.3s} hxg3 {-3.24/42 0.002s} 41. Qxf3+ {+3.12/43 153.4s}
Nf6 {-3.19/41 0.003s} 42. Qxg4 {+3.35/43 296.4s} gxf2+ {-3.31/43 0.001s}
43. Kxf2 {+3.19/47 445.9s} Rxg4 {-3.29/45 0.001s} 44. Bc4 {+3.34/46 77.89s}
Rd4 {-3.39/39 19.81s} 45. Ke2 {+3.38/40 127.3s} Nxd5 {-3.39/42 247.7s}
46. Rb5 {+3.43/50 0.002s} Rxc4 {-3.46/43 63.02s} 47. Nxc4 {+3.55/49 229.8s}
Ke6 {-3.53/43 34.39s} 48. Ne3 {+3.64/50 131.5s} Nf4+ {-3.65/50 187.1s}
49. Kf3 {+3.71/52 0.001s} Bg5 {-3.65/45 33.68s} 50. Bb4 {+3.71/54 81.70s}
Bd8 {-3.66/43 0.002s} 51. Rb8 {+3.73/48 88.85s} Be7 {-3.78/43 15.52s}
52. Rb6 {+3.87/44 220.1s} Nh3 {-3.89/45 0s} 53. Nc4 {+3.87/48 54.25s}
Ng5+ {-3.92/41 7.38s} 54. Ke3 {+3.87/44 70.50s} Nf7 {-3.92/37 42.29s}
55. Ra6 {+3.96/43 52.45s} Kd5 {-4.15/202 234.5s} 56. Nb6+ {+4.23/49 0.001s}
Ke6 {-4.20/45 49.63s} 57. Nc8 {+4.31/41 38.55s} Bf8 {-4.30/42 10.47s}
58. Ke4 {+4.59/39 58.18s} Ng5+ {-4.27/40 8.92s} 59. Kd3 {+4.69/39 52.72s}
Nf7 {-4.40/40 48.62s} 60. Kc4 {+4.71/36 0.001s} Ng5 {-4.31/38 112.7s}
61. Bxd6 {+52.63/33 37.88s} Bxd6 {-48.61/38 96.52s} 62. Rxd6+ {+55.35/44 0.001s}
Kf5 {-4.33/32 23.07s} 63. Kd3 {+56.76/44 26.45s} e4+ {-59.05/40 92.55s}
64. Ke3 {+152.65/44 0s} Ke5 {-58.71/38 26.43s, White wins by adjudication} 1-0
11
  • 2
    White wins or white won one game? A "match" usually means a series of multiple games. Mar 28 at 19:17
  • 1
    @PaŭloEbermann, the game is adjudicated if there are 2 successive moves with a score of 9 pawns or worst.
    – ferdy
    Mar 29 at 0:28
  • 1
    @VladimirFГероямслава, the title of this topic does not say anything about a match of multiple games. Although multiple games would be nice, the problem is what position are we going to start, and it takes time and computer resources to do so, even this single game takes almost 4 hours. If you want to know the result from running multiple games then do so.
    – ferdy
    Mar 29 at 0:40
  • 2
    @ferdy The word "match" was mentioned by you, not by me. I asked whether "white wins" (which implies a repeated win to me) or just won one game you tried. Mar 29 at 5:05
  • 5
    Well, there is more to English, but OK. It means the position is won. I understand this is a strong indication it might be, because in the default position the engines normally draw, but in my physics studies they always repeated to me: "one measurement = zero mrasurements". Mar 29 at 5:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.