The position that no engine can solve.

Sort:
AntonioEsfandiari

The glorious game of knights and queens! Far too complex for any modern engine to understand.  It is very close to an equal position but Stockfish gives black +28 or so out of the opening, and then it starts going haywire in the middlegame and has no clue what is going on.

Tournament Here: 100/250 Registered: https://www.chess.com/tournament/creative-mind-knights-and-queens

 

KT_17

I am not playing that

Image result for no gif

KT_17

Image result for scared cat of cucumber gif

MickinMD

Considering it is a position that would never be reached in a game, no wonder engines can't solve it!

They're designed to solve real chess positions.

Note that in chess puzzle composition, it's required that the position is one that can be reached in an actual game - even if it requires strange moves to get there.

KT_17

exactly

AntonioEsfandiari

@MickinMD Of course it is not a regular chess position hahaha.  It is a chess VARIANT, and my point was, while engines can play at the grandmaster level of every other variant (except bughouse) they cannot play well in this one, and likely won't be able to without a neural-network-based engine.

AntonioEsfandiari

@mickinmd you are mistaken, there are plenty of "studies" and compositions that ARE not reachable chess positions, and used strictly for teaching an idea or technique.  Most chess variants will teach you something about a specific aspect of chess!  For example, 3 check chess teaches you about piece activity, weak squares, center control, timing, etc.  King of the hill teaches you about squares, center control,  risk vs reward, etc.  Crazyhouse teaches you all about weak squares, piece activity, king safety, and makes regular chess tactics seem simple.     This chess variant Knights and Queens teaches you all about space, center control, prophylaxis, king safety, piece activity, every possible queen or knight tactic, knight movement and maneuvering, outposts, and the exponentially increasing value of protected pawns as they march towards advancement!  

AntonioEsfandiari

Also this variant helps your visualization and calculation abilities, especially regarding pawns.  This variant is actually intended to be an improvement on the variant "horde" or dunsany's chess, to make the game a lot more exciting and dynamic.  It will likely be one of the very last chess variants for engines to get good at, so it will probably continue to grow in popularity!

Christopher_Parsons
AntonioEsfandiari wrote:

@mickinmd you are mistaken, there are plenty of "studies" and compositions that ARE not reachable chess positions, and used strictly for teaching an idea or technique.  Most chess variants will teach you something about a specific aspect of chess!  For example, 3 check chess teaches you about piece activity, weak squares, center control, timing, etc.  King of the hill teaches you about squares, center control,  risk vs reward, etc.  Crazyhouse teaches you all about weak squares, piece activity, king safety, and makes regular chess tactics seem simple.     This chess variant Knights and Queens teaches you all about space, center control, prophylaxis, king safety, piece activity, every possible queen or knight tactic, knight movement and maneuvering, outposts, and the exponentially increasing value of protected pawns as they march towards advancement!  

I find Chess 960 an excellent variant to teach chess principles. Since every starting position is different, it puts less emphasis on needing opening knowledge and experience to play well. It forces you to reach a deeper understanding of positional or tactical play, exercise central control, piece placement and coordination, without the crutch of knowing openings.

HorribleTomato

Losing for white, you just need to see some dangerous queen sacrifices.

AntonioEsfandiari

Not losing for white at all!  In fact if white doesn't blunder he usually wins!  The problem is it's extremely hard not to blunder and to avoid all of the tactics that black can set up. The game is very complex and dynamic.  Tomato if your opinion is based off of your engine analysis, it is wrong, engines can't play this correctly because they dont understand the increasing value of the pawns as white gains more and more territory.  The queens become suffocated as the game goes on and you will see your feeble engine go haywire in the middlegame!  I would love for you to show me how white is losing though!  

AntonioEsfandiari

134/250 registered! https://www.chess.com/tournament/creative-mind-knights-and-queens and it's only been up a few days, filling up fast.

Christopher_Parsons

I find this game a truly unique challenge. It seems to level the playing field a lot. I lost to a 1400ish player and beat a 1900ish player. I also consider myself lucky to have a beaten a 1200ish player in a game as well. This is how I would have expected daily standard chess to be when I first came to Chess.com. 

 

Join us in anti-cheating variants. They're are like P!$$ in the cheaters' morning bowl of cereal...

accountclosed99

I joined the tournament. It says it is rated, is it daily rated?

Christopher_Parsons
essejzaf wrote:

I joined the tournament. It says it is rated, is it daily rated?

Yes. Daily rated.

i3chezee

cell 

Ashvapathi

Actually engines can't solve even regular chess opening position. That's why they cheat by using opening books which run about 20-25 moves deep with several different lines. Engines have a horizon effect and their calculations reliable only up to 8-10 moves. Beyond that, they either depend on opening books, endgame table bases or they are confused and wrong.

Ashvapathi

Coming to the OP:

I think zugzwang can be critical in a position like this, so I would say advantage white. Because white has many more pawns and if the game is closed, then white has horses to enter the position and then force some zugzwang to breakthrough.

AntonioEsfandiari

@Ashvapathi Good insight!  Yes zugzwang can actually come into effect for black OR white in this variant. And white SHOULD be slightly favored in this variant, at the higher levels, but the problem is, white is a lot harder to play and the mistakes are much more fatal.  If black blunders even a couple of times he can still retain counterplay and produce tactical opportunities all the way into the endgame.  If white blunders badly he will collapse very very fast.  White goes for a slow positional push, black can play several different plans, but he is always trying to penetrate the back ranks... Sacrificing 3-4 queens to get a queen on the back rank early in the game can easily end the game in a jiffy so white has to be very cautious.  Black can also focus on controlling the center and wait for white to overextend and then blacks queens will have the advantage of attacking both sides of the board while the knights struggle to defend both, similar to Bishop vs Knight endgame in regular chess with pawns on both sides of the board.

universityofpawns

pawn power...I think I'd rather play white actually.....