OpenChess

Again I quote this from memory. Again it seems that even superior engines take more than 1s (OK, sometimes they find it in 2 or 3) to realise some things. E.g., when post-analysing in MultiPV mode, my move would not make the top 3 right away, but after making it, the score would jump (not all my moves are the most trenchant, of course). Maybe king safety issues? If nothing else, the game might be of human interest.
[White "BB+"][Black "GNU Chess 5.07"][Result "1-0"]1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Bg5 a6 8. Na3 Be7 9. Nc4 Nd4 10. Bxf6 Bxf6 11. Nd5 Bg4 12. Qxg4 Nxc2+ 13. Kd1 Nxa1 14. Ncb6 Rb8 15. h4 h5 16. Qh3 Be7 17. g3 O-O 18. Be2 f5 19. Bxh5 fxe4 20. Qe6+ Kh8 21. Bg6 Rf3 22. Bxe4 Rf6 23. Nxf6 Qxb6 24. Qf5And after a few checks, the White king made its way to g2.

I actually failed to realise that 21. Bg6 was threatening Qg4 with mate to follow, and just played it on feel. I don't like the whole idea with 11... Bg4, not just because the knight is trapped (it is rather inconvenient to hunt it down), but that 14. Ncb6 really ties Black up.

BB+ wrote: it seems that even superior engines take more than 1s (OK, sometimes they find it in 2 or 3) to realise some things. E.g., when post-analysing in MultiPV mode, my move would not make the top 3 right away, but after making it, the score would jump

This is true for all engines and for long periods of time, that's why analyzing with several engines is very effective (say, Rybka doesn't have the best move in her top 6 after analyzing every top move for +1 minute, Naum after 20 seconds suggests the best move that shocks Rybka after forcing the variations, etc.)