Comments/Ratings for a Single Item

Mike, sounds interesting.  Care to name this variant?  I think I'll make a
zrf and try it out.  I think I'll also experiment with some similar games
putting all of the pawns on the third rank and dropping the pawn-2 and
en-passant rules.

H. E. Bird invented a precursor to Capa Chess, the difference being the new
pieces next to the King & Queen.

The page on Bird's Chess is
http://www.chessvariants.com/large.dir/bird.html

Capablanca's Chess design analysis
# squares: 80
# piece types: 8
Piece-type density: 10%
Initial piece density: 50%
Long diagonal: a1-h8
Est. piece values: P1, K2, N3, B3, R5, A7, C8, Q9
Power density: 1.40
Exchange Gradient: 0.469 (1 - G = 0.531)
Ave. Game Length Projected: M = 3.5T/P(1-G) = (3.5*8)/(1.4*(0.531))
                                 =    38  Moves
Features: Includes all three two-fold R-N-B compounds, low G means very  
          good exchange potential
Comment: Around 80 years now since the Grandmaster's advocacy of larger 
         board to confront draw problem, Capablanca's Chess 
         practically mimics Carrera's idea from about 400 years ago.

Michael Howe: Larry Smith in 21-3-04 Game Design comment: 'The advantage
in the exchange: No matter the number of the various pieces, a game might
have a significant difference between the weakest and the strongest. This
allows for the potential of advantage in the game, even if the exchanges
are equal. Of course this value would be quite difficult to quantify and
would vary from one game to the next, being dependent upon field and
goal.'
Exchange Gradient now quantifies this, and used for Moves, it closely
predicts game lengths, looking at Courier games and elsewhere.  I
repeatedly called attention to Mark Thompson's article 'Defining
Abstract'(Depth, Clarity, Drama) until someone took note. Now I call
attention to Smith's Exchange Gradient as useful predictor. Here
Capablanca's Chess should show longer games systematically than Orthodox,
its low EG not overcoming higher board size.

There is a new program Smirf (still beta) playing not only 8x8 classic
chess and Fischer Random Chess but also 10x8 Chess variants based on the
Capablanca piece set. It supports also Janus chess and the Capablanca
Random Chess (or FullChess) variant proposed by myself since several
months:

CAPABLANCA RANDOM CHESS (2004-Nov-26) Proposal 

This definition of CRC should cover the following goals:

a) creating an interesting drosophila for chess programmers 
b) using Capablancas 10x8 Chess board geometry 
c) using Capablancas piece set (incl. archbishop and chancellor) 
d) applying rules aligned to Fischer Random Chess 
e) avoiding conflicts to any claimed patents 

The CRC rules are: 

a) creating a starting position (one of 48.000): 
 1) the bishops have to be placed upon different colored
  squares; same rule applies to the implicite bishop pieces:
  queen and archbishop (aligned to FRC)
 2) the king always has to be placed somewhere between the 
  rooks to enable castlings (aligned to FRC)
 3) use only such positions without unprotected pawns (Chess)

b) describing a method of generating starting positions on 
   free squares by using a dice or random number generator: 
 1) select queen or the archbishop to be placed first (2x)
 2) place the selected 1st piece upon a bright square (5x)
 3) place the selected 2nd piece upon a dark square (5x)
 4) one bishop has to be placed upon a bright square (4x) 
 5) one bishop has to be placed upon a dark square (4x) 
 6) one chancellor has to be placed upon a free square (6x) 
 7) one knight has to be placed upon a free square (5x) 
 8) one knight has to be placed upon a free square (4x)/2 
 9) set the king upon the center of three free squares left
11) set the rooks upon the both last free squares left 
12) this establishes White's first row, the Black side 
    has to be built up symmetrically to this 
13) place ten pawns similar to traditional chess in a row 
14) skip this position if it has unprotected pawns or not
    at least three positions in line 1 differently filled
    compared to Gothic Chess, this finally gives about
    21.259 distinct starting arrays.
   
c) nature of (asymmetric Fischer-) castlings:
 1) castlings are (like in traditional chess) only valid
  if neither the affected king or rook has been moved, or
  there would be a need to jump over any third piece, or
  the king would be in chess somewhere from his starting
  position to his target field (both included). Therefore
  all squares between king and its target square (included) 
  have to be free from third pieces, same applies to the
  way the rook has to go to its target square.
 2) the alpha-castling (O-O-O, White's left side):
  like in FRC the king will be placed two rows distant
  from the border (here c-file) and the rook at the next 
  inner neighboured square.
 3) the omega-castling (O-O, White's right side):
  like in FRC the king will be placed one row distant
  from the border (here i-file) and the rook at the next 
  inner neighboured square.

d) performing castlings:
 within a GUI try to move the king upon the related rook
 or at least two squares into that direction; manually:
 1) move the king outside of the board
 2) move the rook to its end position (if need to)
 3) move the king to his end position

e) extended FEN encoding:
 1) the extended FRC-FEN could be used as a base
 2) 'a'/'A' are used to identify archbishops
 3) 'c'/'C' are used to identify chancellors
 4) '9' is used to mark nine empty fields
 5) '0' is used to mark ten empty fields
 6) if a castling enabled rook is not the most outer one
  at that side, the letter of his file has to be placed
  immediately following his castling marker symbol, where
  'q'/'Q' are used for the alpha-, 'k'/'K' for omega-side.
 
f) engine notation rules for castling moves:
 According to UCI convention the castling moves should be
 written by using both coordinates (source and target field)
 of the involved king. But there are castlings, where the 
 king does only one or none simple step. In that cases the 
 castling should be distinguishable by appending a 'k', like
 already practized in promotion moves to make them unique.
 Overmore an engine should accept O-O or O-O-O (no zeroes),
 but only use them, when the GUI would demand for such a
 less precise notation.

However give my best best best auguries for your proposal, that surely
you've explained in more complete and exhausting way than how much I
wanted to make (to only transmit an idea, the same one that you have
already had :-), but that many seem not had spoken up to now, in spite of
the lot of number of chess variants).
This isn't the first time that i crossover with your ideas, and i've
talked with you about all possible pawns structres with same results with
same resulting number (do you rememerber? after that I haven't no more
working about calculating the exact number of all possible chess
positions
because I'm not a serious one-question-at-time solver, in fact I
preferred
work on my personal 'classical chess opening book'...).
However a little difference in my proposal were on the possibility to
incluce in the randomize also the 'Archbishop or Chancellor' chosen, so
to include in the random possibilities the well-known/well-playable Janus
Chess variant (pratically Capablanca Chess with 2 Archbishops instead of
Archbishop+Chancellor). I also propose to ponder the random possibilities
in the chosen of disposition of the pieces, in the randomize process. In
fact as Ed Trice says that his disposition is more well-playable than the
original Carrera/Capablanca disposition, and as Anand says that some
dispositions in Fischer Random Chess are bad-playable, we can start to
ponder the dispositions do it's more probable in the random-chosen a
disposition like the Gothic Chess or Janus Chess or Grotesque Chess,
instead of a disposition with more unprotected pawns (not well-playable
as
Ed Trice says).
Such new specialty could very well be Wild 30 between
the 'Wild' Chess Variants of ICC (the Internet Chess Club)
(http://www.chessclub.com/helpcenter/tips/wild.html), as perhaps Random
Thematic Chess, another chess variant that I've proposed in another
discussion
(http://www.chessninja.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=9;t=000674),
both to obtain the same result to diminish the advantage of the theorical
preparation (not being able to choose the opening from a personal
mnemonic
repertory, like Bobby Fischer says).
My best regards, Reinhard
I will follow you
Giulio

Hi! So see me astonished, I have not thought on such a congruency. If you
would have inspected my homsite www.chessbox.de, you would have detected
that proposal much earlier. But I think, that it should be a better place
here to introduce and discuss that proposal than at a private hompage.

You try to motivate also to include Janus chess somehow within that
proposal. If you would have noticed in my current Smirf program beta,
Janus chess is included in that fine 10x8 and 8x8 aware program. But I
have to learn, that there are some incompatibilities, which force to
exclude Janus from that ramdom idea: a) the notation for castling is
reversed, b) the usage of 'J' instead of 'A' in FEN and encoded moves,
c) the need to also encode the inverted castling by preceeding the castling
block within FEN with an 's' for 'symmetric'. 

If you would spend some time in watching the Smirf program approach you
would notice such interesting things like that it supports all capablanca
based positions (and even CRC or Janus chess) and a PGN load and save of
played games.

Regards, Reinhard.

Hi Reinhard
I had given a glance on your site many months ago and I don't remembered
this proposal (perhaps I wasn't interested in it). Up to now I don't
know
the castling rules but I think that it isn't a problem to change 'J'
in
'A' in the notation, because Janus and Archbishop are practically the
same
piece (I think). I hope you increase the strength of your program that I
haven't yet downloaded but I don't think it's yet more strong than
Shredder in regular chess.
I want only know from you what do you think about to ponder the best
dispositions in randomize process (for example excluding the dispositions
with undefended pawns). 
Regards
Giulio

Well, I do not want to talk that much on Janus Chess here. But the difference is more subtile, e.g. the a-side castling brings the king to the b-file instead of the c-file.

When I have understood your intentions right (I am not sure) you want to know something on the filtering of those 48.000 basic positions. Well, the idea is, that having such a big bool of randomized targets, it would be a good idea to kick off all of them which could have the potential to be an argument that the CRC would produce unfair or unstable positions. So reducing to about 21.000 positions without undefended pawns will nevertheless leave back a huge number of possible starting arrays, which might be sufficient for the current century.

Reinhard.

Question on encoding Capablanca FEN strings:

Smirf actually encodes positions like:
rnbqckabnr/pppppppppp/0/0/0/0/PPPPPPPPPP/RNBQCKABNR w KQkq - 0 1

do you think it would be better to use completed numbers like:
rnbqckabnr/pppppppppp/10/10/10/10/PPPPPPPPPP/RNBQCKABNR w KQkq - 0 1

I am not sure, what to do. So I am gathering arguments.

Reinhard.

There has been stated that CRC (Capablanca Random Chess / FullChess) would
not yet be playable e.g. as a Zillons emulation. But that is not quite
correct. Of course there already a beta version is existing of the soon to
be finished Smirf program. You could find it downloadable at
http://www.chessbox.de/beta.html (see Project Chronicle at 2004-Sep-29).
But it has to be remarked that this is a version finally to be released as
shareware. Thus a lot of functions are shrinked and the user sometimes will
be 'invited' to watch the licensing screen.

Before I was aware of the existence of Capablanca Random Chess(CRC), I had
designed my own hybrid of Fischer Random Chess(FRC)(sometimes known as
Chess960) and Capablanca Chess. My hybrid, Capablanca84000, includes 84000
set-ups as opposed to the 21259 for CRC.
The rule differences are:
1. CRC states that the queen and archbishop must be placed on opposite
coloured squares. Since neither piece is colour-bound (unlike the bishops)
I had not chosen to include this rule. Indeed, a common and logical first
move for the archbishop is that of the knight-style jump, thus landing it
on a different coloured square. If it can be proven that the jump is the
more common first move for the archbishop, it would be equally logical to
place the queen and archbishop on same coloured squares.
2. CRC states that each pawn must be covered. FRC does not and neither
does Capablanca84000. If FRC did include this rule, it would no longer
contain 960 set-ups since some contain uncovered pawns. For example,
set-ups which begin with knight-knight-rook starting from either the a- or
h-file contain 2 uncovered pawns on either the a and b files or g and h
files.

Why filter random positions based on Capablanca' extended board?

Let's talk first on FRC (I have written a small book on that in German language). One main intention to create FRC (or Chess960) has been to make it impossible to provide a complete opening theory for each position. Thus the number of 960 distinct starting positions is helpful to reach that goal. Uncovered pawns are not that problematic because any situation will have to be set up randomly very short before a game starts.

Looking at the Shogi game there are indeed three uncovered pawns in the beginning and the game still does exist today.

Capablanca's chess is somehow different to that because of the huge number of possible starting arrays viewing all shuffled combinations. But during the history from Carrera to Bird, Capablanca [through to contemporary versions] it has been a point of critic and missing acceptance of that extended board. So it could not be counter productive to select special starting arrays which seem to be positionally better constructed, without reducing the huge number of possible initial positions too much.

That leads to the both new rules: a) placing Queen and Archbishop (Archangel) at different colored squares, and b) avoiding unprotected pawns. I cannot see any negative payload connected with this two additional demands. More then 20.000 possibilities should be sufficient.

Also see a nice SMIRF (providing both: FRC and CRC) preview at: http://www.chessbox.de/_tmp/SmirfPrototyp.png

Reply to Jeanette:  I am one of Capablanca's granddaughter.  I don't
think Sergio        
Capablanca is related to us.  I'll ask  some other family members anyway.

To Jeanette and Alina with inquiries dated 2005-05-14 and 2004-12-23
respectively:
There is a good probability that whenever you see an individual with
Capablanca as his or her last name, there is going to be a relationship
with José Raúl Capablanca (1888-1942).
Sergio Gustavo Capablanca (1918-1997) was the son of Bernardo Salvador
Tadeo Capablanca Graupera (1885-1940) and Maria de la Gloria Graupera
Capablanca (1890-1975). Bernardo Salvador, was one of José Raúl's
brothers, the others were Aquiles, Ramiro and Carlos and six sisters,
Aida, Hilda, Graciela, Alicia, Zenaida and Clemencia.
I hope this helps. I am Sergio M. Capablanca, son of Sergio Gustavo
Capablanca and grandnephew of José Raúl.

'Uncovered pawns are not that problematic because any situation will
have to be set up randomly very short before a game starts. 

Looking at the Shogi game there are indeed three uncovered pawns in the
beginning and the game still does exist today.

Capablanca's chess is somehow different to that because of the huge
number of possible starting arrays viewing all shuffled combinations.'

I think the problem is more a matter of the piece set and shape of the
board. Even if a pawn is undefended in a Fischerandom setup, it can't be
attacked instantly, unless it's an a/b/g/h pawn and the piece on its
diagonal is a bishop or queen. But an archbishop or chancellor has a
pretty good chance of being able to make an instant attack on that pawn by
jumping over its own pawn row (as the chancellor can indeed do to the
i-pawn in Capablanca's setup), and the diagonal discovered attack can
affect 80% of the pawns instead of half.

Upon further review, we're discussing opposite ends of the issue. The
points I just made are why the no-undefended-pawn rule is desirable; the
large number of positions is what makes it practical (i. e. you still have
a huge pool of positions to choose from).