Roelof Berkepeis's video: chess : Getting Stronger

Move animation of a beautiful chess game. It's a razor sharp Sicilian Najdorf Defence by GM Anand, where he sacrifices two pieces for an attack. Replay the game at https://lichess.org/S6LZ3fXi Music film: "Something's Coming On" by Joe Cocker, Live at Woodstock 1969. 2019@ArtFlow - all rights respected --- This game recently appeared in a YouTube video by Suren : "Viswanathan Anand Makes A Move Which Stockfish 10 Failed To Find", see https://www.youtube.com/watch?v=KX0DnT_OHj8 .. this is a wonderful example for testing a chess engine ! Indeed Anand, being a human, understands the game, not just concrete variations like StockFish 10. I did some research, using asmFish (an SF 10 version, programmed in assembler code, which runs 10-15% faster) the move 24...Nc7 is suddenly found best at depth 26 (it gives -3.31 while +1.85 at depth 25) but only when i set MultiPV 20 : this UCI setting outputs the [20] best lines when searching. When setting MultiPV 16 or lower, Nc7 never appears and is not considered at greater depths .. using MultiPV 1 (the SF docs say 'Leave at 1 for best performance') runs fast upto depth 40, but Nf6 keeps being 'best move' with value -0.99 .. For SF to look over the horizon, it needs a high MultiPV setting ! [Event "Corus Group A"] [Site "Wijk aan Zee, Holland - https://lichess.org/S6LZ3fXi"] [Date "2006.01.14"] [White "GM Sergey Karjakin"] [Black "GM Viswanathan Anand"] [WhiteElo "2660"] [BlackElo "2792"] [Round "1"] [Result "0-1"] 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Be3 e5 7. Nb3 Be6 8. f3 Be7 9. Qd2 O-O 10. O-O-O Nbd7 11. g4 b5 12. g5 b4 13. Ne2 Ne8 14. f4 a5 15. f5 a4 16. Nbd4 exd4 17. Nxd4 b3 18. Kb1 bxc2+ 19. Nxc2 Bb3 20. axb3 axb3 21. Na3 Ne5 22. h4 Ra5 23. Qc3 Qa8 24. Bg2 Nc7 25. Qxc7 Rc8 26. Qxe7 Nc4 27. g6 hxg6 28. fxg6 Nxa3+ 29. bxa3 Rxa3 30. gxf7+ Kh7 31. f8=N+ Rxf8 32. Qxf8 Ra1+ 33. Kb2 Ra2+ 34. Kc3 Qa5+ 35. Kd3 Qb5+ 36. Kd4 Ra4+ 37. Kc3 Qc4+ 0-1