Can your chess engine see this?

And can you see this as well? Some time ago I had a fun game on the Schliemann defence and reached (strived for) a position that was something beautiful and don't happen that often in chess. I went for my initial plan and blew the whole thing up. Got a crazy attack but after some bad moves by I totally blew it, got depressed and resigned in the endgame. I think I was worst at it anyway but just the fact that I had blew what could have been my best game so far was enough for me.

 Anyway here is the position: 

I think that the first move should be spotted by anyone with some tactical knowledge. Black's second move isn't nothing new either but it is somewhat harder to spot. But I think I'm not very wrong to say that is a well known tactical motif too. 

After my loss I went to analyse this game with Crafty and ChessMaster8K. I know that they aren't the strongest engines around but the are all I got. Neither of them saw the full combination but the two of them agreed that the position was indeed better for black after my idea than what they planned.

I just want to see what you guys tell me about this position and what your engines tell about it too. After it if nobody comes to the same idea as I did I'll post just the first part of the combo and I'm hoping that somebody can tell me if their engines gives a stronger evaluation for black's position after it.

Thanks 

 Yep. This is what I think almost everybody can see. The bishop sac on h7 but there is something even more powerful for black I think. The engines I got don't see it but after I make the moves and let them analyse the position they both agree that black's position is even more winning. And besides the the second part of the combo is just so cool.

I think that if I went just with the classic bishop sac I'd won the game cause I'm used to doing it and I know when it works and why it works. With this I failed miserably. To give you a hint I'll give you the first part of the combo as it went on the game actually played 1... Bxh2+ 2. Kh1{not wanting to fall for the knight check and being drawn into the front of the pawn shelter. I think this is actually the best defence} I think that maybe now the second part of the combo becomes more apparent.

 That's just the firs part of the combo. And Kxh2 is just suicidal. White should play Kh1.

 You got it! The double bishop sacrifice. And I blew it... So can you please tell if you have a chess engine if it finds the second bishop sac for black. And if it doesn't, make him analyse the line he gives for the position and see how it evaluates it. Then make him analyse things after the second bishop sac and see how different the second analysis is.

Rybka sees 1...Bxh2 2.Kh1 and then:

    2...Bxg2+ 3.Kxg2 Qd5+ 3.Kxh2 Qxg5 (-2.91) 

    2...Qd7 3.Kxh2 (-2.75)

 That's interesting. Hiarcs gives Bxg2 a score of 3.5 and Qd7 a score of 1.6.

 Yes! Crafty and ChessMaster also give very similar evaluations at the beginning of the line.But did Rybka saw the double bishop sac?

I agree that with so similar evaluations nothing much can be said about the strongest move. But the second bishop sac is much more stylish and on top of that it is sound! 

> But did Rybka saw the double bishop sac?

Yes, see the lines I posted. It saw both alternatives.

 I'm not very knowlegable of engines but in that case I'd say that Sigma really needs some fine tuning. 

 Very nice.

Sharukin> That's interesting. Hiarcs gives Bxg2 a score of 3.5 and Qd7 a score of 1.6.

From the initial position Rybka evaluated the double bishop sac as -2.91, but once it played the moves it re-scored it as -3.47. I probably did not give it enough thinking time on the original position. I also use Hiarcs... it's a nice complement.

 I think this evaluation is much more closer to truth. I think that this sacrifice is more of a positional and deep one. And normally engines aren't very good at strategical/positional judgement. I think that human players looking at both positions immediatley have a feel that black's win is much more tangible with the double bishop sac but seeing it is really the hard part.

 If you go deeper I think Rybka will rate black's position even higher. I analysed this a lot after my loss and the win was there and it was like a very long forced mate. The thing is that engines don't see it. When I give the position to crafty after the second bishop sac he starts by saying that white is winning. After a while he gives black the advantage and after a minute or so he gives -2. something as his evaluation.

A few notes about Chessengines -

The exact score is just a question of scale. It means nothing if one program says -1.9 and another 2.5 on the same position. It's not as simple as: if it's 1 point per pawn, it's all the same in all programs. It's common for Hiarcs to give 

There are positions that are +4 by many engines and are drawn (rare. But they do happen in practical play now and then).

One program's -1.5 is another's -2 at some positions - and in many of those cases, even when the position is completely won for black, the program who stated the -1.5 is actually BETTER at playing and evaluationg the same kind of positions that were evaluated.

My suggestion is to try to  limit yourself to: "At which point does the program's evaluation is high enough to suggest a win". a +1.1 from Rybka is already enough for a win in most positions where it happens ( I said most, so engine experts - don't post the positions where Rybka has it wrong. I know them well).

Also in the same program, once you pass the score that usually means a win with correct play, differences between -2 and -3 are symbolic (meaningless. All that really matters if that's an easy win... That extra one point doesn't even mean it's any easier to win at these scores).

 Wow. Nice comments. By the way what chess engines (free) do you suggest me to get to annalyse my games? I normally prefer to analyse by myself and when I have doubts I resort to engine analysis. I do this because I have no fun at looking at lines engines spit out and not understanding why those moves are the best ones he considers. Of course that in some cases I understand why but in many positions were there are no direct threats and tactics seems to be just sleeping it can be very frustrating. But what I got is CM8K and Crafty and neither of them inspire me in great confidence. I think I'll install Arena and then install some of the free engines that are available out there.

But on the topic of the thread some people were analysing this with the same engine and got two winning lines for black. Of course if one is -2.3 and the other -3.75 (with enough time)  both are clearly winning and either one could be followed. But in this case the double bishop sacrifice was just too flashy and it could happen in an opening were I never saw it happen. By the way thanks for reviving my biggest fiasco chess-wise.