This is a puzzle I saw in a magazine awhile back. White's goal is to win one of the two black bishops. Black's goal is to avoid this for 8 moves. If white can capture a black bishop in under 8 moves then white wins, otherwise black wins.

I can't seem to get the "insert position" to delete the Kings off the board...

White only has a queen on d1
Black only has 2 bishops, one on c8 and one on f8

Black to move.

Gimme the Hint!

ozzie, good endgame study! [it is actually 1/3 puzzle, 1/3 problem, 1/3 study. It is NOT endgame exercise IMO - because it is primarily tactical... Q v B+B is of course drawn in general: it is a win only either in trivial cases or where the position is of a  tactical nature. Here the Bs are in a bad position relative to the Q - so tactics prevail...]

Oh if you really want to put in the diagram just place the Ks anywhere and say clearly ie in boldface big font that the Ks can be [almost] anywhere; the solution must be independent of the position of the Ks. (On second thught, I can see quibbles arising from that, so the presentation is better the way it now is...]

The magazine in whch I read this puzzle was I think Chess Life. I distinctly remember that this puzzle was used in Russian chess schools as an elementary way of teaching about calculation and board visualization. I remember finding it interesting for two reasons. First, it took me awhile to solve it. Second, it ended up being quite a good puzzle without any Kings on the board! Who would've thunk it?! :-)

The other possibility is to upload a FEN string which doesn't include a king. I don't know if that is possible though.

if course it is possible. Okay - not FEN - but the parent of FEN - the good old forsythe - comes to the rescue. Here, I'll be happy to do it for you:    :)

 2b2b///////3Q b.

After any Black move, White to capture a  Bishop in 8 moves.

or: if you want it more FEN-like [checksum of 8, slash after white's home rank also]:

"

 2b2b2/8/8/8/8/8/8/3Q4/ b.

After any Black move, White to capture a  Bishop in 8 moves.

"

[btw now I think it is an instructive exercise - as it has wide application. Something like those nontrivial 'general-special' cases where K+R wins again K+N, likewise for K+R against K+B

Also, because it is black on move while the goal is to accomplished by white, it would perhaps be more accurate to call this one a tactical theme... ]

-----------------------

btw: do you have anything similar with Q v NN with the Ns protecting each other; black K in or near centre? afaik *that* sort of position is *always* a dead and easy draw [except when it is a trivial win e.g. where white can win by playing it like K+Q v K; the config being such that the Ns cannot intervene...] - so I'd be happy to be proved wrong...

In such a K+Q v K+N+N puzzle, if one exists, the prob would be 'white to play and win', and the Ks would have to be explicitly shown; because with the knights protecting each other the Q can't just pick a knight...

That would also be most illuminating.

I appreciate your desire to name the puzzle in a very precise manner - but the fact remains that nobody has attempted to solve it yet!

I wonder if a puzzle like this is particularly good since it is not possible to just solve it with a computer. It's not even really possible to check your work with a computer. Those pesky Russians were ahead of their time!

Ozzie, thanks for the puzzle. Here is my solution. I didn't bother to write out trivial variations where Black hangs a bishop or allows an immediate fork or skewer.

::snip::

Edit: Solution removed.

Solved! The queen can force a win in every line.  :)

Hey, good job!

I've added your lines to my existing solution tree. But I disagree that ...Bg7 is the hardest defense for black. I think the toughest defense is ...Be6.

...Bg7 can be beat by an immediate key move for white which puts black in zugzwang.

ozzie> ...Bg7 can be beat by an immediate key move for white

I'm blind. ::snip:: all give the White bishop no good squares to go to so ::snip::! attacks it and wins on the spot. I'm curious to see your solution tree. :)

Edit: Faster solution removed.

Yup, *** is the key move that I had there.

Good job with the posting of all the variations. Pretty good puzzle, no?

Edit: Redacted the key move

Yeah, it's a good board vision exercise and no temptation to use an engine is a plus.

timepass> looks like you spent the time it needed

I don't even want to say how long that was.  ;)

If you want to, you can delete the solution.

Oh dear! I thought it was clear from my post that I had solved it! [I am not sure *I* solved it - actually I remembered the puzzle and the solution, because I live in a country whose relationship with the USSR was not so bad as the USA's was...] I didn't post the solution: why post a spoiler? Specially as:

I do think this puzzle is particylarly good - and not only because it is tricky, but because it is thematic [Ks anywhere, black to move, white to win...] , and so, though tactical, applicable in a lot of Q v BB cases. That's why I posted whether you knew of something similar, or something similar - meaning nontrivial thematic tactics -was at at all possible with Q v NN, Ns supporting each other...

I have some USSR material with some more of these tactical endgame themes [from Kaspayan's domination-endgame-study-compilation book: the 'theme' part of several chapters]

- I'll post them on this thread after I return from my chess vacation - if and when I feel like. Unless ozzie doesn't want them here. [those are much simpler; but important...]

At present I am both unwell and sick of chess.

Get better soon.

Reuben Fine to the contrary, Q v B+B is won in general, as shown by endgame tablebases.

Common statistical misconception.

I have had those tablebase stats since early-1990s on my machines. If you know what a newgroup is, search rec.games.chess archives for discussions on this.

Tablebases give stats on randomly placed 'legal' Q v B+B - which are won in general, but that is irrelevant to actual chess play.

Q v B+B positions that arise in a naturally played game of chess [whether by weak players or not] - statistics on them are not easy to come by, but are possible after years of experimentation..

DavidForthoffer, you are not in the John Nunn league on that; it is better left to specialists - Nunn et al -  extracting *meaningful* info from tablebases...

And keep the 50-move rule in mind...

Nothing statistical about actually arising nontrivial chess positions is automatically "shown" by tablebases, contrary to what you think....

It's hard to imagine how a KQ vs KBB endgame could ever arise naturally since such an imbalance almost never occurs in the absence of pawns.

That said, in KQ vs KBB tablebase endgames the win is only guaranteed if the side with the Queen moves first.  If the side with the bishops moves first the game will usually end in a draw via the 50 move rule. (I say "usually" because it depends heavily on the starting location of the pieces when the endgame begins.  In the default positions, the maxim holds true.)

You realize of course, that the initial posting has nothing to do with this tablebase lookup problem you are discussing...

The 2002 edition of Secrets of Pawnless Enings by J Nunn indicates that these are won by the queen, isn't this exactly what Forthoffer indicated?