1

Jonathan Scrantz played this game against Lichess Level 8:

[Title "Jonathan Scrantz VS Stockfish Level 8 (3000) after 9. Kh1"]
[FEN ""]
[Startply "17"]

1. e4 e5 2. Bc4 Bc5 3. b4 Bxb4 4. f4 exf4 5. Nf3 Be7 6. d4 Bh4+ 7. g3 fxg3 8. O-O gxh2+ 9. Kh1 d5 10. exd5 Bh3 11. Nc3 Bxf1 12. Qxf1 Bf6 13. Ba3 Ne7 14. Ne4 O-O 15. Nxf6+ gxf6 16. Bxe7 Qxe7 17. Bd3 Qd7 18. Qg2+ Kh8 19. Bxh7 Qxd5 20. Qh3 Kg7 21. c4 Qa5 22. Qg2+ Kh6 23. c5 Rh8 24. Nh4 Qxc5 25. dxc5 Kh5 26. Nf5 Rxh7 27. Rf1 Rg7 28. Qxg7 Na6 29. Rf3 b6 30. Rh3#

I am aware that Lichess Level 8 deliberately makes mistakes, but I wonder whether it really did so in this game or not. It seems to me that 12. ... Bf6 was the definitively losing move. But SF10+ on that position at depth 24+ thinks that Bf6 has score +1.9 compared to +1.3 for Be7, so did Black already have a losing position before move 12? If so, where is the first point at which Black made a clear mistake and what is the general (human) strategy for White to win from that point?

8
  • It is possible that they used Stockfish against itself. Feb 5 at 2:21
  • @RewanDemontay: Ah that's an interesting hypothesis. Well I'd give Jonathan the benefit of doubt since it seems that he had already prepared lines at least up to move 15, and White has a clear advantage after Bf6 but SF10+ doesn't see it until searching quite deep, so it seems plausible if he could keep the advantage. I guess he played many many times in analysis mode trying to figure out how to trick SF, and memorized many lines starting from move 9. But do you think you can answer my question? I can't trust SF that's why I'm asking humans haha..
    – user21820
    Feb 5 at 3:03
  • @user21820 The guy was using a weak lichess stockfish and gave it very little depth. It's really not hard to exploit an engine at around ~15 depth Feb 5 at 3:43
  • @edit_profile: We all know that Lichess Level 8 is not full SF. But that wasn't my question. Can you identify the first mistake?
    – user21820
    Feb 5 at 4:26
  • @user21820 My bad, didn't read the question. The game is winning for black as early as move 4. For white, the game is won by at least move 9. I'll do some more analysis for 8...d5, which is borderline winning but I haven't completely analyzed it, but it looks like white might be winning even earlier than that. Feb 5 at 6:29

1 Answer 1

1

Black had a losing position after 7...fxg3. The best way for black to go for a draw at this point was to play 7...d5, which continues 8.Bxd5 fxg3 9.O-O Nf6 10.Ba3 gxh2+ 11.Kh1 Nxd5 12.exd5 Qxd5 13.Nc3 Qa5 +=.

There are two possible continuations for black after 8.O-O, which are 8...d4 and 8...gxh2+, that may not seem initially bad, but are completely losing. I've detailed those in the following study: https://lichess.org/study/URz9uXJd

Please ask me any further questions if your confused.

2
  • Thank you for your answer, but do you have any human explanation of the winning strategy for White? Or did you just do a computer analysis? Your study claims that the best line after 9. Kh1 follows Jonathan's game, but you didn't show how to respond to 12. ... Be7, and I don't see any obvious human strategy to guarantee winning against that.
    – user21820
    Feb 5 at 9:31
  • I don't have a human explanation because the position is too complex for me. According to the analysis, there simply are too many complications and tactics in the position for there to be a single "winning strategy". After 12.Qf1, it's completely winning. 12...Be7 is responded with 13.d6 Qxd6 14.Nd5 Nc6 15. Bf4 Feb 5 at 15:46

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.