Page 1 of 1

Test position: To sac or not to sac?

Posted: Tue Jun 22, 2010 7:41 pm
by zullil
From the "Where did I go wrong here?" thread, which is currently in distress:

r2r2k1/5pp1/1pp1bn1p/p1p5/q3P2N/3PB2P/1PPQ2P1/3RR1K1 w - - 0 20

White played Bxh6. What does your favorite engine suggest?

Re: Test position: To sac or not to sac?

Posted: Tue Jun 22, 2010 7:54 pm
by Robert Flesher
zullil wrote:From the "Where did I go wrong here?" thread, which is currently in distress:

r2r2k1/5pp1/1pp1bn1p/p1p5/q3P2N/3PB2P/1PPQ2P1/3RR1K1 w - - 0 20

White played Bxh6. What does your favorite engine suggest?

I posted a reply and somehow broke your thread, sorry about that :D

My Dissident aggressor settings says it is a good move, but Ng6 is poor, Nf3 is better and white may then be winning!

Re: Test position: To sac or not to sac?

Posted: Tue Jun 22, 2010 11:18 pm
by royb
Robert Flesher wrote:
zullil wrote:From the "Where did I go wrong here?" thread, which is currently in distress:

r2r2k1/5pp1/1pp1bn1p/p1p5/q3P2N/3PB2P/1PPQ2P1/3RR1K1 w - - 0 20

White played Bxh6. What does your favorite engine suggest?

I posted a reply and somehow broke your thread, sorry about that :D

My Dissident aggressor settings says it is a good move, but Ng6 is poor, Nf3 is better and white may then be winning!
After some long think times and some interactive analysis with Stockfish I believe that while it's not necessarily winning outright (as far as I can tell with my engines and 1 hour of CPU time) these lines look promising:

After Bxh6 gxh6 Qxh6 Nh7:

28 -0.32 22.Nf3 Rd7 23.Rd2 Kh8 24.Rf1 Rg8 25.Rdf2 Rg6 26.Qf4 Rd8 27.b3 Qa3 28.Qc7 Rc8 29.Qxb6 Ng5 30.Nxg5 Rxg5 31.g4 Qb4 32.Qa6 Rcg8 33.Qxc6 Rh5 34.Kg2 Qd4 35.Rxf7 Bxf7 36.Rxf7 Rh7 37.Qf6+ Qxf6 38.Rxf6 Kg7 39.Rc6 Rgh8 40.Rxc5 Rxh3 41.Rxa5 (3533.93) (Stockfish custom settings)

or

24 -0.60 22.Nf3 Re8 23.Rd2 Re7 24.Rf2 Bd7 25.Ng5 f6 26.Nxh7 Rxh7 27.Qxf6 Qd4 28.Rf1 Qxf6 29.Rxf6 Re7 30.R1f3 Rae8 31.b3 Kg7 32.R6f4 Kh7 33.Rf6 Be6 (3636.04) (IvanHoe 57a)

Notice both believe that white has a bunch of compensation (but not quite enough!) for the sacrificed material.

This turns out to be a very interesting position! Thanks for posting it!

Re: Test position: To sac or not to sac?

Posted: Tue Jun 22, 2010 11:36 pm
by yanquis1972
well, obviously there is compensation; i think the more interesting question is is it objectively the best move. it's the kind of move any human plays in a heartbeat, & if the ensuing moves are executed well against a non-silicon or ice-veined opponent with plenty of clock it is probably good enough to win or at very least draw; it appears it would very easy to mobilize all 4 pieces against the king quite quickly, & the black queen appears hopelessly stuck on the queenside. (i suggest anyone play robert's zappa aggressor with the black pieces & try to defend the position). i didn't analyze closely but R4 seems to think black is down no matter what; without the human knowledge to complicate it was suggesting Qf2 i believe.

i sincerely doubt it's winning though; again it would be interesting to try zappa aggressor vs R4 at longish TC from this position & observe. i'm pretty confident rybka would hold and be obviously better within 10 moves but i'd be glad to be wrong.

Re: Test position: To sac or not to sac?

Posted: Wed Jun 23, 2010 1:32 am
by Robert Flesher
yanquis1972 wrote:well, obviously there is compensation; i think the more interesting question is is it objectively the best move. it's the kind of move any human plays in a heartbeat, & if the ensuing moves are executed well against a non-silicon or ice-veined opponent with plenty of clock it is probably good enough to win or at very least draw; it appears it would very easy to mobilize all 4 pieces against the king quite quickly, & the black queen appears hopelessly stuck on the queenside. (i suggest anyone play robert's zappa aggressor with the black pieces & try to defend the position). i didn't analyze closely but R4 seems to think black is down no matter what; without the human knowledge to complicate it was suggesting Qf2 i believe.

i sincerely doubt it's winning though; again it would be interesting to try zappa aggressor vs R4 at longish TC from this position & observe. i'm pretty confident rybka would hold and be obviously better within 10 moves but i'd be glad to be wrong.

Can you post the whole game again, I played Zappa Dissident aggressor Vs Rybka 4.0 and Zappa won! But I thought you sacked later in the game?

Re: Test position: To sac or not to sac?

Posted: Wed Jun 23, 2010 1:37 am
by yanquis1972
unfortunately, it wasn't my game, but the sac was immediate. i'd love to see the zappa-rykba pgn.

Re: Test position: To sac or not to sac?

Posted: Wed Jun 23, 2010 1:41 am
by zullil
Robert Flesher wrote: Can you post the whole game again, I played Zappa Dissident aggressor Vs Rybka 4.0 and Zappa won! But I thought you sacked later in the game?
It wasn't my game either, but here it is:

[Event "?"]
[Site "?"]
[Date "2010.06.22"]
[Round "?"]
[White "Hagen"]
[Black "Rybka 3.Dynamic 1-cpu w32"]
[Result "0-1"]
[PlyCount "52"]
[EventDate "2010.??.??"]

1. e4 c5 2. Nf3 Nc6 3. Bb5 Qc7 4. O-O Nf6 5. Re1 a6 6. Bxc6 dxc6 7. d3 e5 8. h3
Bd6 9. Bg5 h6 10. Be3 O-O 11. Nbd2 b6 12. a4 Rd8 13. Nc4 Be6 14. Nxd6 Qxd6 15.
Qd2 a5 16. Rad1 Bd7 17. Nh4 Be6 18. f4 exf4 19. Bxf4 Qd4+ 20. Be3 Qxa4 21. Bxh6
gxh6 22. Qxh6 Nh7 23. Ng6 Qd4+ 24. Kh1 Qf6 25. Rf1 Qxg6 26. Qxg6+ fxg6 0-1