Forums

An original puzzle

Sort:
Kacparov

I can't solve it, maybe someone else can? Possibly it's very easy but I don't see something :)

The puzzle is:

White has a king, a knight and a pawn. Black has a king and a queen. Set up such a position, that white wins even though it's black to move and the black king is not checked.

Note: even if white manages to promote, KQN vs KQ is usually a draw.

Conflagration_Planet

Give the answer then Borg.

Kacparov

2396? That's bullet, it doesn't count :)

I guess I'm missing something simple

waffllemaster

Tried for a bit, very tough because of drawing resources.  I'm having trouble making a position with 2 threats where there's no stalemate or drawn pawn endgame.

So I don't think it's too simple because the queen can often sac itself for either piece to make a draw pretty easily.

stubborn_d0nkey

I think I might be able to solve this with endgame tablebases, just by myself would probably be too hard

waffllemaster
stubborn_d0nkey wrote:

I think I might be able to solve this with endgame tablebases, just by myself would probably be too hard

Yeah, that'd be pretty easy, just look through the 7 million different combinations of 5 pieces on 64 squares and you'll have it in no time!

And hey, because of mirror positions and possibly more than 1 solution, you may have only 1 or 2 million to look through.  Good luck.

waffllemaster

I kept trying to make this type of thing work, but I decided it can't and has to be something else :)

Threatens two different mates, lookin good?

 

 

But there are too many checks on the 6th rank, the most annoying is just Qf6+ because of the stalemate.

Queen close to same rank or file as king eliminates lots of checks.  I also tried to use knight as fodder to hamper the queen.  But in those cases couldn't set up a threatening enough position.

stubborn_d0nkey

Wow wafflemaster, ? What I obviously meant is that there may be a line where white can queen that is a winning KQN vs KQ. Looking at all such positions (the KQN vs KQ that can arise) and seeing whether they are a win or a draw would be too much for me.

Kacparov

 

 

 

 

 

 

 

I still don't know the answer. There is one more idea (on the diagram), but black manages to save by Qe7+! or Qe8+! and then Kg7. Maybe it can be changed somehow so there is no Kg7 possibility.

heinzie

*Spoilers*

http://kirill-kryukov.com/chess/longest-checkmates/

edit: fortunately that doesn't really help any, either

stubborn_d0nkey

It may be able to give the answer

Kacparov

And still no solution ?

Conflagration_Planet

E mail one of the IMs or GMs on here, and ask them.

waffllemaster
stubborn_d0nkey wrote:

Wow wafflemaster, ? What I obviously meant is that there may be a line where white can queen that is a winning KQN vs KQ. Looking at all such positions (the KQN vs KQ that can arise) and seeing whether they are a win or a draw would be too much for me.

That was my idea too, but I was too lazy to go look up in one of my endgame books to learn when Q+N can beat Q and then try to come up with some clever way to force that position.

That would be a good way to approach the problem though.

(And putting random positions in an EGTB may take a long time).

tolikid
Kacparov wrote:

 

 

 

 

 

 

 

 

I still don't know the answer. There is one more idea (on the diagram), but black manages to save by Qe7+! or Qe8+! and then Kg7. Maybe it can be changed somehow so there is no Kg7 possibility.

If white first to move then Nf7 is mate. If not then draw


:

tolikid
martos wrote:

The original problem:

 

I think this is solution that arrives at previous puzzle:):

Kacparov

That can't be the solution as once black promotes, it's white's move - and it should be black's...

I got a clue that the pawn is on the g (or b) file, but still no solution.

azziralc

What's the answer?

Pensadorf1

???

gimmewuchagot

Wow, I have no idea, lol