I don't believe you at all. Any computer would rate a move that didn't extend the game in the face of an incoming mate, as a blunder.
Perfect play but mated in 22 moves?
Sort:
Kupov3 wrote:
I don't believe you at all. Any computer would rate a move that didn't extend the game in the face of an incoming mate, as a blunder.
Well, that's what it said when it came back from computer analysis.
Here you go, a copy of the page:
Rapid Team Match (>1500, 2days/move) - Board 1
| Score: | 1-0 | Date: | 12/2/2009 |
| Time: | 2 days | ||
| Opening: | E21: Nimzo-Indian Defense: Three Knights Variation | ||
Computer Analysis (~2500 strength)
Inaccuracies: 0 = 0.0% of moves
Mistakes: 0 = 0.0% of moves
Blunders: 0 = 0.0% of moves
Score:
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Nf3 O-O 5. Bg5 c5 6. e3 h6 7. Bh4 Qa5 8. Bxf6 Bxc3+ 9. bxc3 Qxc3+ 10. Nd2 gxf6 11. Rb1 cxd4 12. Rb3 Qa5 13. Rb5 Qxa2 14. Qg4+ Kh8 ( 14... Kh7 15. Qxd4 f5 16. Be2 Nc6 17. Qc3 a6 18. Rb6 Rg8 19. O-O Qa5 20. Qxa5 Nxa5 21. Ra1 Nc6 22. c5 ) ( 14... Kh8 15. Rh5 Qa1+ 16. Ke2 Kh7 17. Qf4 d3+ 18. Kxd3 Kg8 19. Qxh6 Qa3+ 20. Ke2 Qd3+ 21. Kxd3 f5 22. Qh7# ) 15. Rh5 Qa1+ 16. Ke2 d3+ 17. Kxd3 Kh7 18. Qh4? ( 18. Qf4 Kg8 19. Qxh6 Qa3+ 20. Ke2 Qd3+ 21. Kxd3 Nc6 22. Qh8# ) ( 18. Qh4 f5 19. Rxh6+ Kg8 20. g4 f4 21. Bg2 Qg7 22. Ne4 f5 23. Nf6+ Rxf6 24. Rxf6 fxe3 25. Rxf5 Nc6 26. Rg5 Ne5+ 27. Kc2 ) 18... Kg6 ( 18... f5 19. Rxh6+ Kg8 20. g4 f4 21. Bg2 Qg7 22. Ne4 f5 23. Nf6+ Rxf6 24. Rxf6 d5 25. Ke2 Nd7 26. Rxe6 Qb2+ 27. Kd1 Qb1+ 28. Kd2 Qb4+ 29. Kd1 fxe3 30. cxd5 exf2 31. Qxf2 Qxg4+ 32. Kc2 ) ( 18... Kg6 19. Rxh6+ Kg7 20. Rh7+ Kg8 21. Rh8+ Kg7 22. Qh6# ) 19. Rxh6+ Kg7 20. Rh7+ Kg8 21. Rh8+ Kg7 22. Qh6#
Inaccuracies(?!): 0 = 0.0% of moves | Mistakes(?): 0 = 0.0% of moves | Blunders(??): 0 = 0.0% of moves
Believe me at all now? ^^
No inaccuracies does not mean you played perfectly. Suppose the threshold for ?! is more than -0.3 (deviation from the "best" move) but each of your moves were -0.29 then the machine would make no comment as to inaccuracies as you slowly slide into the mire.
I agree the site analysis is pretty shoddy. I have a number of examples where it suggests better moves than mine and proceeds to give a long line of analysis but on the move after its improvement it is completely stymied by a simple move that was over its horizon (a mate in one for instance! dropping a whole Rook on another occasion).
It looks like it's suggesting plenty of alternative variations...Maybe the interface bugged out or something, or it rated your opponents play, I dunno.
That is strange.
I just input this game into Deep Rybka 3...
Afer 14.Qg4+, it says 0.00,
After 14....Kh8, it says +156.50 (and gives a variation of 14....Kh7 as +0.00 instead) and in the chess.com computer analysis it gives a couple variations but fails to reveal a mistake or blunder.
The evaluation pretty much stays at +156.50 until 18.Qh4 when it says +2.63, but after 18....Kg6 it goes back to 156.50 until checkmate with a variation of 18...f5!? 19.Rxh6 Kg8 20.g4 f4 21.Bg7 Qg7 22.Be4 as +2.63
Analysis time = 10 minutes 45 seconds.
So much for chess.com 2500-level analysis.
May 9, 2010
0
#8
how many moves deep did Rybka look to find 14...Kh8 as the error? it might have gone further than our computers look.
For completeness sake, 14....Kh8 is given the "??" and the 14....Kh7! variation continues...
15.d3 f5
16.0-0 d6
17. Qxd4 Nc6
18.Qc3 a6
19. Bb1 Qa4
20.Bc2
-0.07
Of the moves actually played, the only symbol assessments that are given are...
14....Kh8?? and
18. Qh4?! Kg6?!
erik wrote:
how many moves deep did Rybka look to find 14...Kh8 as the error? it might have gone further than our computers look.
I am not sure how to assess the "depth" of the search since I am a computer-analysis rookie and I have heard that "depth" of search is not the same as the ply-depth of the given variation.
This is actually the first game I have ever analyzed using Rybka.
I thought 12-ply was what the computer analysis on chess.com goes to but how that translates to "depth" is well beyond me. If anyone knows how I can look this up on Deep Rybka 3 w32, please let me know...I used whatever the default is.
But why would the chess.com analysis list mating variations of 14...Kh7 and not call those inaccuracies, blunders or mistakes?
I'm not sure what to think of the computer analysis (str 2500) when it gives me 0 inaccuracies, 0 mistakes and 0 blunders eventhough I got mated in 22 moves. What's up with that?