internet chess club

Internet Chess Club

Grant and you will reveal!
Strange name, real hobbies
[ Sign up | Log in | Guest ] (beta)
ganstaman 52 ( +1 | -1 )
Puzzle 283 Ok, the puzzle on the homepage at the time of me writing this (search for puzzle number 283 here: )

White to move, mate in 4. I won't spoil the solution given, as it is brilliant seeming to me.

But how can black spoil white's plan of Qh4-(h or e)7-b7#? I can see black delaying the journey by 1 move, so it would be mate in 4. What did I miss, if anything?
heinzkat 0 ( +1 | -1 )
I think the answer is 1. Qh4 Rh1!
sf115 35 ( +1 | -1 )
If 1. Qh4 Rh1 white either:

2. Qxh1 a1=Q 3. Qh7 Qxa6+ 4. bxa6 b1=Q 5. Qb7++ which is a move slower than the actual solution.


2. continue with the original plan with 2. Qe7 Rh8+ 3. Nd8 Rh7 (3...Rxd8 also slows down whites mate) 4. Qb7+ Rxb7 5. axb7++ which is also in 5 moves.

So there is only a mate in 5, which is slower than the actual solution
ganstaman 3 ( +1 | -1 )
Ah, that seems to do it. Thanks.
lighttotheright 48 ( +1 | -1 )
There is an alternate solution that is not accounted for by this puzzle. This kind of thing irritates me. It turns out that my first choice was in fact correct, but the puzzle said I made a mistake. Once you know the solution, it is easy to see that my alternate path works because mine is based on the same principles that are used in the given one.

It's still mate in four, but why are alternate solutions unaccounted for? I thought that problem was fixed long ago.
heinzkat 21 ( +1 | -1 )
That problem was fixed long ago Lighttotheright, I think you're wrong here though - puzzle 283 has only one solution (for mate in four).
lighttotheright 9 ( +1 | -1 )
The alternate 1. Nd8 works and is confirmed. It is mate in four just like the given solution.
lighttotheright 46 ( +1 | -1 )
Heinzkat -- I'm not wrong. 1. Nd8 does work. The problem is that no computer can solve this particular problem from four moves out. The computer checking the solutions is wrong.

Once you input the solution, the computer suddenly sees the solution at 2 or 3 moves out depending upon the program that you use. This particular puzzle is an anti-computer one.
heinzkat 30 ( +1 | -1 )
Lighttotheright! This is the interesting part of the puzzle, which is not covered in puzzle #283, since the puzzle application didn't support underpromotions of the opposite side back then. After 1. Nd8? [if I may boldly give it a ?] c2
2. Qc1 b1=B! Black is stalemated and there can be no mate in four.

After 1. Nb4 however, 1. ... c2 2. Qc1 b1=B 3. Nd3! exd3 4. Qh1, it IS mate in four :-)
chessnovice 21 ( +1 | -1 )
... Sometimes puzzle makers don't have solutions entered entirely. Human error is prone to omissions. If you find an alternate solution, you're able to enter it. Nothing to necessarily get irritated about, I think.
heinzkat 63 ( +1 | -1 )
Also... Remember GameKnot does no checking of the puzzle at all - the puzzle application merely checks if all variations end in checkmate. If so, the puzzle is 'approved' 'correct' (twice '', ah well) an can be put up for others to solve. As you can see on the puzzle page, a lot of the puzzles still have some hiatuses then. If you think 1. Nd8 mates in four too, 'prove it'! You can do so by browsing to puzzle #283, hover over 'Options' in the right corner and click 'Alt. solution...'. This will give you the option to prove there is another way to play, that results in mate in at most the same number of moves as the original author. If all variations are correct, you can save the puzzle too, and the puzzle has been 'improved'.
lighttotheright 13 ( +1 | -1 )
1. Nd8 c2 2. Qc1 b1=B 3. Nd5 Qb2 (or any move) 4. Nc7#

It is forced checkmate. There is no stalemate with my solution. The alternate is confirmed, and not seen by computer at four moves out.
lighttotheright 9 ( +1 | -1 )
1. Nd8 c2 2. Qc1 b1=B 3. Nd5 Bb2 (or any move) 4. Nc7#

Sorry I made a slight error in copying the line. This is correct.
heinzkat 13 ( +1 | -1 )
2. ... b1=B! Not 2. ... b1=Q. Please re-evaluate your assessments. :-)

(note that I hadn't noticed the b1=B solution either. A nice brilliancy by Ado Kraemer)
heinzkat 4 ( +1 | -1 )
Lighttotheright 3. ... Bb2 is not possible in the intended line! It's 2. ... b1=B, not 2. ... bxc1=B.
lighttotheright 9 ( +1 | -1 )
I did it again.

1. Nd8 c2 2. Qc1 b1=B 3. Ne6 Bb2 (or any move) 4. Nc7#

Those knight moves can get confusing.

I think this is correct now.
lighttotheright 15 ( +1 | -1 )
Heinskat look at the position. The key to the solutiion is the knight path. The Queen sac is to block any defence by black. The solution is very simple. 1. Nb4 and 1. Nd8 both work.
heinzkat 7 ( +1 | -1 )
Note that in your line... 3. ... Bb2 is illegal. In fact, after 3. Ne6, all Black's moves are illegal - it is stalemate!
heinzkat 26 ( +1 | -1 )
OK... with a diagram then... After
1. Nd8 c2
2. Qc1 b1=B
3. Ne6
It is Black to move. Do you see any legal ones?

Interesting how Ado Kraemer can still keep us busy with his brilliant problems, 36 years after he died.
lighttotheright 27 ( +1 | -1 )
OK heinzkat. You threw me for a loop with b1. I was thinking about the c1 square. But even with b1=B, it is still checkmate in four. 1. Nd8 b1=b 2. Qxc3 and my solution still works.

heinzkat 28 ( +1 | -1 )
Well... 1. Nd8 c2
2. Qc1 b1=B is the line that makes mate in four impossible.

1. Nd8 b1=B like you give in this last post, perfectly mates in four - but there's a difference. Carefully check the notations right from the beginning of this thread, I think I haven't messed up any of them...
lighttotheright 12 ( +1 | -1 )
OK...It is still not stalemate. But it would be a mate in 5 instead. 1. Nd8 c2 2. Qc1 b1=B 3. Qxc2 and it is not stalemate. ...but it does take an extra turn to get the knight to c7.
lighttotheright 14 ( +1 | -1 )
Sorry for some of the copy mistakes. My copy function on my computer is not working. I have to reboot my whole computer to correct the problem.
heinzkat 14 ( +1 | -1 )
Exactly... And therefore, 1. Nd8 is an incorrect try. Mate in five is possible in many ways - there is only one correct starting move to mate in four.
lighttotheright 8 ( +1 | -1 )
Well, it looked correct to me at the time. And my computer was not cooperating with me either!
chessnovice 20 ( +1 | -1 )
... b1=B is pretty clever! Indeed, prolongs the Nd8 line by one move. I wouldn't worry about not considering that though, since I doubt that very many people really would.
bogg 17 ( +1 | -1 )
In case anyone else has the ... same problem that I had. The reason that 2. ... b1=B doesn't cook 1.Nb4 and does cook 1. Nd4 and 1. Nd8 is because after 1.Nd4 White has 3. Nd3 ed: 4. Qh1++.

Took me a while to see the tree within the forest.
heinzkat 8 ( +1 | -1 )
chessnovice Obviously the composer did consider it :)

bogg: indeed, see my post above too, 5/05/2008 04:47:30
chessnovice 4 ( +1 | -1 )
hienzkat True. But if you can't trust the composer, who can you trust?
bogg 9 ( +1 | -1 )
heinzkat Missed that post. Would have saved me some time had I noticed it.