MatPlus.Net

 Website founded by
Milan Velimirović
in 2006

13:56 UTC
ISC 2024
 
  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
1-Jan-2024

B P C F





 
 
MatPlus.Net Forum General all things being equal, in a retro problem...
 
You can only view this page!
(1) Posted by Eugene Rosner [Monday, Sep 29, 2014 15:34]

all things being equal, in a retro problem...


does white vs. black economy matter? In directmate problems there is a standard preference that if there is a choice between a white piece and a black piece on the board, one chooses the black. Is there a convention like this in retros?
 
(Read Only)pid=12724
(2) Posted by Nikola Predrag [Monday, Sep 29, 2014 16:48]

When White must carry the task, the skill&art is to achieve it with minimal white force. If Black helps, even less white force is desirable.
Of course, everything depends on the content.
There are various retro-types and various contents, so...
 
 
(Read Only)pid=12725
(3) Posted by Per Olin [Monday, Sep 29, 2014 20:01]; edited by Per Olin [14-09-29]

Proof games are not always considered to be retros, but I take the opportunity to add a question related to proof games.

Proof game problems have emerged from the need to determine the legality of a position. Gradually this has turned into an own popular problem genre. When we consider the economy of a proof game starting from the standard initial position, then we generally see many excessive pieces; there is much material that is not necessary for expressing the content of the problem and not needed for correctness. This excessive use of force is automatically accepted, all proof games start from the initial position and are never considered to be uneconomical.

The economical version of proof games are a->b proof games (A to B), where the starting position is composed and there is the possibility of leaving out excessive force. In a recent award, where I had an a->b proof game, the judge mentioned that this form gives more possibilities for expressing normal problem thematic ideas; further the judge mentioned that normal proof games are very limited. To this praise for artistic freedom one can only agree.

Question supporting the original question in post 1: should we rethink when it comes to economy in proof games?
 
   
(Read Only)pid=12729
(4) Posted by Ganapathi Ramaswami [Tuesday, Sep 30, 2014 09:09]

I got quite interested. Can you please show the problem from the award and also how did you computer test it?
 
   
(Read Only)pid=12730
(5) Posted by Per Olin [Tuesday, Sep 30, 2014 09:44]

The claim period for the competition in question ends tomorrow Oct. 1. Will check with the tourney director that there are no changes to the award and post later this week.
 
   
(Read Only)pid=12731
(6) Posted by Neal Turner [Tuesday, Sep 30, 2014 10:55]

'...this form gives more possibilities for expressing normal problem thematic ideas'

But are proofgamers interested in expressing normal problem thematic ideas?
 
   
(Read Only)pid=12732
(7) Posted by Per Olin [Tuesday, Sep 30, 2014 11:27]

Neal, as answer to your question: yes, normal thematic ideas and, of course, things that can not be shown in any other form than proof games. As a normal theme, take for instance my a->b proof game from Problemas April 2013, which shows four AUWs (Allumwandlungen). We have AUWs in direct mates, helpmates, selfmates, studies etc., it is a natural theme also for proof games.
 
   
(Read Only)pid=12733
(8) Posted by Joost de Heer [Tuesday, Sep 30, 2014 14:45]

But those AUWs have a completely different motivation: There's a bishop promotion because there's a bishop in the final diagram, not because the rook-part of a queen would interfere with the solution.

Usually, several non-moving pieces in a proofgame do have a function: they limit the movement of the moving pieces. Moving a rook from a1 to h1 is trivial on an empty board, but becomes harder if you have to avoid nonmovable objects.
 
   
(Read Only)pid=12738
(9) Posted by Per Olin [Tuesday, Sep 30, 2014 18:23]

Correct, 'those AUWs have a completely different motivation'. But this fact does not deprive them the status of being AUWs. The issue raised by Neal was are proof game composers interested in normal thematic ideas and I mentioned AUW as an example. Examples of normal motivations are in http://pdb.dieschwalbe.de/search.jsp?expression=PROBID='P1001352' and http://pdb.dieschwalbe.de/search.jsp?expression=PROBID='P1084087' (these links will either function or not). As promotion motivation is not the issue of this thread, let's leave it for the time being!

Again correct, 'several non-moving pieces in a proofgame do have a function: they limit the movement of the moving pieces'. A non-moving piece, e.g. if it prevents duals, is not excessive force: it has a function, it secures correctness. But my question concerns extra force in a proof game, that has no function at all, is there only because it is there in the initial game array. Is it then optimal to start from the initial game array? Is proof games the only problem genre that accepts extra force and does not condemn this as uneconomical?
 
   
(Read Only)pid=12740
(10) Posted by Dupont Nicolas [Wednesday, Oct 1, 2014 00:22]

The (classical) proof game genre has its own economy criteria, which depends on the final position and not on the initial one. Number of moves, captures, existence or not of visible promoted force, home-based side, etc. Note that ranking those criteria is not obvious – as an example concerning the second problem you quoted, what about the same theme realized with only one visible promotion but with more moves and a no more home-based white side, is it better or not?

Concerning the a->b proof game genre, I find it interesting and full of promises, but I will try to compose this way only when a strong solving program will exist – Popeye is unable to check such a bit long and complicated problem. I asked Etienne Dupuis to include a->b possibility in Euclide, but it hasn’t been implemented yet…
 
   
(Read Only)pid=12742
(11) Posted by shankar ram [Wednesday, Oct 1, 2014 04:52]

Platzwechsel (exchange of pieces) is one more "orthodox" theme that has been done in proof games..
A cyclic exchange would be interesting..
Then you have the Valladao theme: castling + promotion + en passant capture..
 
   
(Read Only)pid=12744
(12) Posted by Per Olin [Thursday, Oct 2, 2014 16:44]

Referring to kind request in post 4 and gradually coming to the point, here something that I hope will open up what I am trying to express. My earlier posts can be seen as commercial spots for the story to come. Thanks for your patience!

Normal proof game (Problem 1)

Let's start with a problem published here in MatPlus Forum in a sidetrack to a discussion. The problem shows loss of a tempo with a knight, which is possible when the knight is captured later during the solution. This is only a scheme, nobody gets excited about the content.

Per Olin
MatPlus Forum Dec. 11, 2013

(= 15+14 )

Proof game in a) 4.5 moves b) 5.0 moves

Solution: a) 1. Nb1-c3 Pd7-d5 2. Nc3xd5 Pe7-e6 3. Nd5-e7 Nb8-d7 4. Ne7xg8 Rh8xg8 5. Pc2-c3
b) 1. Nb1-c3 Pd7-d5 2. Nc3xd5 Pe7-e6 3. Pc2-c3 Ng8-e7 4. Nd5xe7 Nb8-d7 5. Ne7-g8 Rh8xg8


A to B (Problem 2)

The theme of Springaren Summer Tourney 2013, open for all types of problems, was: 'In the play predominantly the same kind of piece is used by both Black and White'. According to the judge L. Werner the tours by the knights in the winning problem are well worth having a look at.

Per Olin
1st Prize Springaren Summer Tourney 2013

(= 13+12 )

Position A

(= 10+10 )

Position B

Proof game in 10.0 moves from A to B

(Generally, the diagrams are side by side, but that seems to be impossible here)

Both parties make ten moves. The white pawn on f7 is captured by the black knight and the black pawn on g6 is captured by either white knight. The black pawn c6 is captured on c6 or after it has advanced. Further are captured one of the white knights and the white bishop.

Tries: 1) In the try 1.Sb7/a4? c5 2.Sxc5 Sb4 black captures the bishop on d8 and the pawn on f7 and returns to a2. This gives for black one pawn move and eight knight moves, total nine moves, which is one too few. Getting a tenth move for black by moving the pawn to c4 fails, as the white knight checks when capturing on c4. 2) If the play starts with 1.Sb7? Sb4 2.Sa5 Sa2 3.Sxc6 Sb4 4.La5 Sxc6, then the problem is the white bishop, which has moved away from d8. Black can capture the bishop on his return from f7 to a2 on b4 or d8, but white then needs eleven moves: two moves with the bishop, three with the knight starting on c5, six for the knight on c1 to capture on g6 and return. If here white in his 4th move does not move the bishop on d8, then it is captured on its starting square d8, which leaves white with nine moves. White can, however, not lose a move with the knight on c1 on its route to g6 and back to c1 .

Solution: When going in the other direction, white manages to lose a move by 1.Se2 Sc1 2.Sf4 Se2 3.Sxg6 Sf4 4.Sh8 Sg6 5.Sb7 Sxh8 6.Sa5 Sxf7 7.Sxc6 Sxd8 8.Sb4 Sc6 9.Sa2 Sb4 10.Sc1 Sa2.



Economy and artistic freedom

Economy concerning material in a chess problem means expressing the idea with as little force as possible. Problem 1 is in this respect clearly uneconomical; there is much excessive force not needed for the idea. Problem 2 with 13 + 12 pieces in position A fulfills the normal criteria for economy in a chess problem: there is nothing extra there.

In post 10 was mentioned that proof games are evaluated with other criteria than economy. This is correct; the question raised in this thread is that should economy (of the starting position) be part of the criteria list?

For a normal proof game, position A is the initial game array and position B is the diagram position; a normal proof game is published without position A, which is the same for all proof games. This position with 32 pieces is heavy, fixed, beautiful... but limits much the composer. Compared to Problem 1 above, Problem 2 has seen many steps that have been possible thanks to the freedom that the composer has in A to B proof games. Try to get the content of Problem 2 into a normal proof game, and I wish you luck!


Computer testing

Popeye a->b is the only program known to me that can test this type of proof games. The big amount of pieces in Problem 2 above makes the testing slow. In cases where testing the whole move sequence is practically impossible, I test the beginning, the end and sometimes parts in between. This does not give a 100 % certainty, but helps the composer to detect flaws. Problem 2 has been tested as follows: separately moves 1.-8. (from 1.Se2 to 8. - Sc6) and moves 4 . - 10 (from 4.Sh8 to 10. - Sa2), both sequences C+.

As mentioned in post 10, computer testing with Popeye is not the easiest of tasks. But, I think we must be grateful for what we have and use it until we have something better. The limitations in computer testing is one more reason to have light positions! - For more details concerning computer testing, please, contact me through the Notes service of MatPlus Forum.


Summing up

Economy is fundamental, artistic freedom is desirable.
 
   
(Read Only)pid=12755
(13) Posted by Dupont Nicolas [Thursday, Oct 2, 2014 17:39]

Yes, I also use Popeye's trick "C+ by parts" while composing a fairy proof game. But Euclide with implemented a->b possibility should solve your (beautiful) problem 2 in less than a second...

It makes me think that, a couple of years ago, I asked to Pascal Wassong if he could program a->b inside Natch. He answered me that he doesn't want to work anymore on his program, but that Natch's source code is open, and hence anyone who wish is free to develop it further. It would be very useful to find some brave guy able to do this a->b job inside Natch...
 
 
(Read Only)pid=12758

No more posts


MatPlus.Net Forum General all things being equal, in a retro problem...