MatPlus.Net

 Website founded by
Milan Velimirović
in 2006

22:01 UTC
 
  Forum*
 
 
 
 

Username:

Password:

Remember me

 
Forgot your
password?
Click here!
SIGN IN
to create your account if you don't already have one.
CHESS
SOLVING

Tournaments
Rating lists
01-Oct-2019

B P C F





 
 
MatPlus.Net Forum Helpmates When castling must be the first move
 
You can only view this page!
Page: [Previous] [Next] 1 2
(1) Posted by Rewan Demontay (Real Name: James Malcom) [Thursday, Jun 20, 2019 07:31]; edited by Rewan Demontay (Real Name: James Malcom) [19-06-20]

When castling must be the first move


Today, I recently came across the idea of castling being forced on the first move of a helpmate on Chess Stack Exchange. Thesource of my inspiration is a helpate in three that can be seen in @supercat's answer to this CSE: https://chess.stackexchange.com/q/18587/15543

Since then, I have been thinking of how to extend the idea for as long as possible. My earlier attempts have contained tons of promoted pieces. But, luckily my lastet attempt has none.

Here's my position:

h#10-White To Move And Mate Black

(= 8+13 )


Any cooks?

Also, does anyone know of any similar problems that revolve around castling and helpmates? Is a longer version of the idea here possible at all?
 
(Read Only)pid=17499
(2) Posted by Siegfried Hornecker [Thursday, Jun 20, 2019 07:37]

Any cook? Yes. easily. For example 1.-0-0 2.Sa2 Kh1 3.Re1 Kg1 4.Re7 d:e7 5.b1R e8S 6.Sc1 Sc7 mate
 
(Read Only)pid=17500
(3) Posted by Vitaly Medintsev [Thursday, Jun 20, 2019 07:44]; edited by Vitaly Medintsev [19-06-20]

Unfortunately, the are multiple cooks in 5.5 moves (so-called short solution), for example 1...0-0 2.Sc1-b3 Kg1-h1 3.Rb1-e1 Kh1-h2 4.Re1-e7 d6*e7 5.b2-b1=Q e7-e8=S 6.Ba1-h8 Se8-c7 #

On PDB Server (https://pdb.dieschwalbe.de/search.jsp), the following search query displays 2885 helpmates with at least one castling:
(not g='retro' and not g='fairies') and g='h#' and (sol='0-0' or sol='0-0-0')
 
(Read Only)pid=17501
(4) Posted by Viktoras Paliulionis [Thursday, Jun 20, 2019 12:59]; edited by Viktoras Paliulionis [19-06-20]

The longest known helpmate with white castling is h#7:

Frank Richter
10945 feenschach (207) 05-06/2014
(= 10+9 )
h#7

1.Bf1 0-0-0 2.Bg2 Rg1 3.Bh1 Rg3 4.Kf7 Rxf3+ 5.Kg6 Rg3+ 6.Kf5 f4 7.Be4 Rg5#

https://pdb.dieschwalbe.de/search.jsp?expression=probid=%27P1285983%27
 
 
(Read Only)pid=17502
(5) Posted by Frank Richter [Thursday, Jun 20, 2019 13:44]; edited by Frank Richter [19-06-20]

Thanks for citing here ... it may look easy to achieve a longer correct helpmate in legal position without promoted pieces, but it isn't ...
 
   
(Read Only)pid=17503
(6) Posted by Rewan Demontay (Real Name: James Malcom) [Thursday, Jun 20, 2019 17:20]; edited by Rewan Demontay (Real Name: James Malcom) [19-06-20]

It must be hard to make an extended one with no promoted pieces.

I had previously made this h#8 before that supposed ten, and it does contain promoted pieces.

White To Play And Mate White-h#8

(= 11+9 )


I tried applying the cook that exists in the h#10, but to no avail. Any flaws in this one? There's a few solutions to this one by my count.

#1: 1. O-O-O Nh3 2. Ba4 Re1 3. Bfe8 Re5 4. Beb5 Rd5 5. Bf7 Rd7 6. exd7 Ng5 7. d8=N
Nxf7 8. Nxf7# or 7. d8=Q/R Nh3 8. Q/Rxf8#

#2: 1. O-O-O Nh3 2. Bb4 Rg1 3. Bxe7 Rg5 4. Bxf8 Rf5 5. Re1 Rxf7 6. e7 Rxe7 7. Rxe7
Ng5 8. Bxg7#

#3: 1. O-O-O Nh3 2. Bb4 Rg1 3. Kb2 Rg5 4. Rc1 Rf5 5. Rc8 Rxf7 6. Ba4 Ng5 7. Bxf7 Nh3
8. Rxf8#
 
   
(Read Only)pid=17506
(7) Posted by Vitaly Medintsev [Thursday, Jun 20, 2019 18:04]

This one solves in 3,5 moves: 1...Bc3*g7 + 2.Kh8*g7 Ra1-a5 3.Kg7-h6 h7-h8=Q #

Have you aware about free solving software?
For instance, Olive (http://www.yacpdb.org/#static/olive) solves this one in 0.002 sec
 
 
(Read Only)pid=17507
(8) Posted by Vitaly Medintsev [Thursday, Jun 20, 2019 18:20]

One more h#7 with white castling as mating move in twin 'b)'

Mirko Degenkolbe & Steven B. Dowd, Die Schwalbe 2010
(= 10+9 )
h#7 a) Add black Pc4, b) Move c1 b2

a) 1.Be2 Bb2 2.Bd1 Bc1 3.Bb3 a×b3 4.c×b3 c4 5.b2 Sc3 6.b×c1=B Se4 7.B×d2+ K×d2#

b) 1.Be2 Bc1 2.Bd1 Bb2 3.Bb3 Bc1 4.B×c4 Bb2 5.B×a2 c4 6.B×b1 Bd4 7.Ba2 0-0-0#

http://www.yacpdb.org/#386266
 
   
(Read Only)pid=17508
(9) Posted by Rewan Demontay (Real Name: James Malcom) [Thursday, Jun 20, 2019 18:40]

No, I was not aware of that. I'll have to implement it next time I'm on the right computer. Thanks!

I honestly do try to find cooks-I guess I just suck at it.

Here's a probable failure of a fix:

(= 11+9 )


#1: 1. O-O-O Nh3 2. Ba4 Rg1 3. Beb5 Rg5 4. Bfe8 Rd5 5. Bec6 Rd7 6. exd7 Ng5 7. d8=N
Nf7 8. Nxf7#

#2: 1. O-O-O Nh3 2. Kb2 Rg1 3. Rc1 Rg5 4. Rc8 Rf5 5. Ba4 Rxf7 6. Bgxf7 Ng5 7. Beb5
Nh3 8. Rxf8#
 
   
(Read Only)pid=17509
(10) Posted by Rewan Demontay (Real Name: James Malcom) [Thursday, Jun 20, 2019 18:45]; edited by Rewan Demontay (Real Name: James Malcom) [19-06-20]

Annnnnddd I just found a cook:
1. O-O-O Nh3 2. Kb2 Ng5 3. Rc1 Nxf7 4. Rc8 Nd6 5. Bgf7 Nxe8 6. Rxe8 Rg1 7. Rxf8#

I'm currently experimenting with this setup. Th idea here is to delay the White rook's attack range entrance onto the 8th rank.

(= 11+9 )


EDIT: And already any hope of a h#8 is fricked here. See below. Looks like I need an entirety difference matrix.

1. O-O-O Nh3 2. Kb2 Ng5 3. Ra1 Nxf7 4. Bb5 Nd6 5. Bf7 Nxc8 6. Ra8 Nd6 7. Rxf8#

Although, if this 7 upholds, it might make a nice little puzzle, given how tricky it was for me to find the right sequence.
 
   
(Read Only)pid=17510
(11) Posted by Frank Richter [Thursday, Jun 20, 2019 19:23]

This is easily cooked with 4.N:h7 R:h7#

The Degenkolbe/Dowd is an interesting find.
 
   
(Read Only)pid=17511
(12) Posted by Rewan Demontay (Real Name: James Malcom) [Thursday, Jun 20, 2019 19:25]

Also, Frank, the PBD link in one of your earlier replies does not work as usual for them links. Could you give the search terms please? Also, it is possible to search for for then one criteria at a time in PBD?
 
   
(Read Only)pid=17512
(13) Posted by Frank Richter [Thursday, Jun 20, 2019 20:31]

To which criteria do you refer?

Yes, it's possible to combine them.
Example: G='h#' and STIP='h#2' and SOL='0-0' etc.
And more advanced queries are possible too, please refer to the given examples there.
Looking for a single problem you may using PROBID='P1234567'
Hope it helps.
 
   
(Read Only)pid=17513
(14) Posted by Rewan Demontay (Real Name: James Malcom) [Thursday, Jun 20, 2019 21:14]; edited by Rewan Demontay (Real Name: James Malcom) [19-06-20]

Eh, just dkrgo about it.

Also, seeing as how a mate I'm three I'm the least possible amount of moves this kind of problem can have (as far as I know), what do you think is the most economic way to do that?

This is the best that I could come up with.

(= 6+7 )


It should be easy enough to solve. It's White to move and mate.

What about the longest possible dual free version? Or does Frank's h#7 cover that?

Is an h#8 possible at all with promoted pieces?
 
 
(Read Only)pid=17514
(15) Posted by Vitaly Medintsev [Thursday, Jun 20, 2019 22:14]

to (12)
On this web page you will find all Search Expression, Boolean Operators and Wild Cards for PDB Server https://pdb.dieschwalbe.de/article.jsp?idx=3
 
   
(Read Only)pid=17515
(16) Posted by Olaf Jenkner [Tuesday, Jun 25, 2019 00:02]; edited by Olaf Jenkner [19-06-25]

A new record is easily done.
Take the zero-position of the problem "Mirko Degenkolbe & Steven B. Dowd" and add a black knight on b2, h#7,5
 
   
(Read Only)pid=17521
(17) Posted by Olaf Jenkner [Tuesday, Jun 25, 2019 00:13]

Sorry, I missed one more half move.
Put bPf3 to e4 and add a white knight on f3: h#8
 
   
(Read Only)pid=17522
(18) Posted by Olaf Jenkner [Tuesday, Jun 25, 2019 00:19]

The next one:
Put the white knight f3 to e5 and add a black knight on f3: h#8,5

I should go to sleep...
 
   
(Read Only)pid=17523
(19) Posted by Olaf Jenkner [Thursday, Jun 27, 2019 19:07]

Mirko Degenkolbe gave me some information about this "record":
The h#7 above mentioned was part of a small article by Degenkolbe and Dowd published 2010 in "Die Schwalbe".
They found nearly the same position as mine (wSd4 instead of e5)but did not publish that ugly thing.
After reading the article Hannes Baumann found the position as well.
We can say the record is h#8,5 but never published.
 
   
(Read Only)pid=17525
(20) Posted by Rewan Demontay (Real Name: James Malcom) [Saturday, Jun 29, 2019 14:40]

Do you mind putting up the h#8.6 position please? And thanks!
 
   
(Read Only)pid=17526

Read more...
Page: [Previous] [Next] 1 2

MatPlus.Net Forum Helpmates When castling must be the first move