The Lemon Fool

.  --- --- --- --- --- --- --- ---
8 |   |   |   |   | K |   | n |   |
   --- --- --- --- --- --- --- ---
7 |   |   | p |   |   |   |   |   |
   --- --- --- --- --- --- --- ---
6 |   |   | R |   |   |   |   |   |
   --- --- --- --- --- --- --- ---
5 |   |   |   | p |   | k | P | P |
   --- --- --- --- --- --- --- ---
4 |   |   |   |   | p | P | p |   |
   --- --- --- --- --- --- --- ---
3 | p |   |   |   | p | p | B |   |
   --- --- --- --- --- --- --- ---
2 |   |   |   |   |   | R |   |   |
   --- --- --- --- --- --- --- ---
1 |   |   |   | B |   |   |   |   |
   --- --- --- --- --- --- --- ---
    a   b   c   d   e   f   g   h

In this chess puzzle white, shown here as red, is to play and win in three moves, against any defence.

Cinelli

Looks far too easy. Where's the catch?

Black's king is firmly stuck.

But for the knight covering f6, it's mate-in-one (R-f6) with white's rather unshakeable network of pieces protecting each other. And the king can dislodge the knight: there's no reply to K-f8.

So black's only escape is to clear e4 of his own pawn. He must use the first move to vacate e3, either to e2 or xf2, then advance e4-e3 while white rids himself of the knight.

That line kills the obvious mate-in-3. So we need an alternative. k-f7 instead of f8 is no use because then black can check him at h8 on the second move after clearing e3. Various obvious rook moves (c5, d6, f6, g6) also fail to mate in the time.

But wait. This is a puzzle, not a game. We have a bishop at d1. At c2 it pins the pawn and kills black's escape. At b3 it can come into play via d5, which also kills black's escape. And it can check the king in two moves: either by sacrifice at e4 or from d7 via a4.

But in three moves? No, not a new line. But it resurrects our first line, by killing black's escape against K-f8. Because once e4 is clear - which takes black both his moves - B-c2 is mate by another route. QED.

That still leaves a puzzle: why the other white rook? If it's meant to serve as a distraction, couldn't it be a queen? Add a black piece preventing Q-a2 becoming a trivial win: maybe for instance a black pawn at b2 or rook at a1 for extra fun.

UncleEbenezer wrote:Looks far too easy. Where's the catch?

Black's king is firmly stuck.

But for the knight covering f6, it's mate-in-one (R-f6) with white's rather unshakeable network of pieces protecting each other. And the king can dislodge the knight: there's no reply to K-f8.

What about 1. K-f8 a3-a2, then 2. KxN a2-a1 Q ?

Haven't read the rest of your post, sorry if you cover this.

GS

GoSeigen wrote:

UncleEbenezer wrote:Looks far too easy. Where's the catch?

Black's king is firmly stuck.

But for the knight covering f6, it's mate-in-one (R-f6) with white's rather unshakeable network of pieces protecting each other. And the king can dislodge the knight: there's no reply to K-f8.

What about 1. K-f8 a3-a2, then 2. KxN a2-a1 Q ?

GS

That's the easy case.
3 R- f6 mate, 'cos the (other) black pawn is still blocking the escape to e4.

UncleEbenezer wrote:

GoSeigen wrote:

UncleEbenezer wrote:Looks far too easy. Where's the catch?

Black's king is firmly stuck.

But for the knight covering f6, it's mate-in-one (R-f6) with white's rather unshakeable network of pieces protecting each other. And the king can dislodge the knight: there's no reply to K-f8.

What about 1. K-f8 a3-a2, then 2. KxN a2-a1 Q ?

GS

That's the easy case.
3 R- f6 mate, 'cos the (other) black pawn is still blocking the escape to e4.

3. ... QxR??


GS

Sorry, I don't know how to disguise my text with choice of color, so I'll add some blank lines...


























1. B - e2 threatening unstoppable mate via a6 and c8 (if N moves, R-f6 mates immediately, P moves don't help, including queening)
So ... P * e2 is forced

2. R - e6 threatening mate on E5
... K * e6 is forced

3. P - f5 is now mate, as it is defended by the R on f2, and K can't move to e5 or d6 as the P move has exposed those squares to B on g3

forlesen wrote:1. B - e2 threatening unstoppable mate via a6 and c8 (if N moves, R-f6 mates immediately, P moves don't help, including queening)
So ... P * e2 is forced

Not quite the whole story. Black can reply with P*f2. Then after white moves to a6, black can open that escape route at e4.

Of course you can then still mate at d3, which is the same outcome as my mundane line's mate at c2 if black opens e4.

Though from your imaginative use of the more unlikely pieces, I guess yours is the solution he was expecting.

Agreed, I was just going to add that additional line.

By the way, Uncle E, what is your technique for the hidden text? Presumably you're using Font colour, but what is the correct colour value to use?

forlesen wrote:By the way, Uncle E, what is your technique for the hidden text? Presumably you're using Font colour, but what is the correct colour value to use?

I can never remember the colour value, but I do know how to find it! The answer is to go to a post that uses the technique, such as UncleEbenezer's earlier in this thread, and click on the " button to start a reply that quotes it, then look for it in the quote. In this case, it's on the second line of the quote:

[color=#dFf8F8]

The exact value is not especially critical, by the way, so you'll probably get somewhat different values from different posts.

Edit: Don't forget to close or move away from the reply started with the " button without submitting it!

Gengulphus

Thanks Gengulphus

Testing, testing...


Yes, that works!

forlesen wrote:By the way, Uncle E, what is your technique for the hidden text? Presumably you're using Font colour, but what is the correct colour value to use?

If you want the exact colour, you could look at the page source, or use your choice of graphics utility to sample the background colour.
Alternatively, if you hit the quote button to quote my post, you'll see the approximation I guessed and used.

Note that it's not a reliable technique: not every browser will take any notice - even those that display visually in the first place. Fine for casual use, but not if your purpose is a bit more important than a minor spoiler.

UncleEbenezer wrote:Though from your imaginative use of the more unlikely pieces, I guess yours is the solution he was expecting.

Uncle, I don't see that yours was a solution at all, though. You didn't answer my last post about 3. ... QxR.

Forlesen, very nice; and cinelli, great puzzle, thanks.

GS

UncleEbenezer wrote:

forlesen wrote:By the way, Uncle E, what is your technique for the hidden text? Presumably you're using Font colour, but what is the correct colour value to use?

If you want the exact colour, you could look at the page source, or use your choice of graphics utility to sample the background colour.
Alternatively, if you hit the quote button to quote my post, you'll see the approximation I guessed and used.

Note that it's not a reliable technique: not every browser will take any notice - even those that display visually in the first place. Fine for casual use, but not if your purpose is a bit more important than a minor spoiler.

The backgrounds of posts alternate between two shades of light blue.

Examples:
Black: ##########
The lighter blue is ECF3F7: ########## (Invisible)
The darker blue is E1EBF2: ########## (Just visible)

Julian F. G. W.

Examples:
Black: ##########
The lighter blue is ECF3F7: ########## (Just visible)
The darker blue is E1EBF2: ########## (Invisible)

Julian F. G. W.

jfgw wrote:
The backgrounds of posts alternate between two shades of light blue.

Examples:
Black: ##########
The lighter blue is ECF3F7: ########## (Invisible)
The darker blue is E1EBF2: ########## (Just visible)

Julian F. G. W.

So, trying to optimise this, a colour like E7EFF5 should split the difference between the two background colours used.

##########

Yes, that is very hard to read.

Finally, back on topic, I neglected to thank Cinelli for an excellent puzzle. Even after thinking of the first mating line above, and realising it was improbable enough to be a likely contender, I still spent quite a while failing to see how to achieve mate if black did the obvious thing and took the bishop.

forlesen wrote:

jfgw wrote:
The backgrounds of posts alternate between two shades of light blue.

Examples:
Black: ##########
The lighter blue is ECF3F7: ########## (Invisible)
The darker blue is E1EBF2: ########## (Just visible)

Julian F. G. W.

So, trying to optimise this, a colour like E7EFF5 should split the difference between the two background colours used.

##########

Yes, that is very hard to read.

Until it's quoted and so has a pale yellow background, when all of those are just visible - though I wouldn't say readable! So maybe one ideally needs the colour value for quote backgrounds as well, and to average between all three. Or rather, the two colour values for quote backgrounds and to average between all four, as nested quotes seem to be done with alternating background colours as well (see the 5th post of the thread for an example containing a triply-nested quote). Those colour values don't seem to be visible in the page source, unfortunately, at least to my rather inexpert eye (I did know HTML quite well at one point, to the point that I could write simple pages in HTML fairly efficiently by hand, and indeed I still can, but that was many years ago and the language has moved on a long way since then - so I'm anything but an expert now!).

It should also be said that I don't think complete invisibility is desirable - barely visible that something is there but unreadable is better, because uninitiated new users are less likely to be fooled into thinking that it's just a bit of white space left as a separator for some unknown reason, and thus to look for a method to read it.

Finally, I'll note that even if one goes to the trouble of making stuff completely invisible by matching the colour used to where the post will appear in the alternation, as jfgw did above, it can still go wrong in the topic review that appears at the bottom of the page when one is composing a reply, because posts appear in reverse time order there. While I've been composing this reply, his examples as seen there get the "Invisible" and "Just visible" labels entirely the wrong way around! (But they'll presumably be the right way around for whoever composes the next reply after this one...)

Gengulphus

Gengulphus wrote:Until it's quoted and so has a pale yellow background, when all of those are just visible - though I wouldn't say readable! So maybe one ideally needs the colour value for quote backgrounds as well, and to average between all three. Or rather, the two colour values for quote backgrounds and to average between all four, as nested quotes seem to be done with alternating background colours as well (see the 5th post of the thread for an example containing a triply-nested quote).

Why would anyone intentionally quote some text that was designed to be invisible due to its "spoiler" nature? ("Hey, this isn't meant to be discussed but let's discuss it anyway!").

Seems like a scenario you shouldn't have to jump through hoops to cater for.

Scott.

Well solved. forlesen was the first to come up with the key move Be2, threatening 2 Ba6 and 3 Bc8 mate. However black responds he can’t escape mate in three. This was a Fritz Giegold puzzle.

Cinelli