Check out Janggi (Korean Chess), our featured variant for December, 2024.


[ Help | Earliest Comments | Latest Comments ]
[ List All Subjects of Discussion | Create New Subject of Discussion ]
[ List Latest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]

Comments/Ratings for a Single Item

Later Reverse Order Earlier
Latte Chess. Alternative starting position, with each player having 4 rooks/bishops/knights, 1 king and queen, and 8 pawns. (8x8, Cells: 64) [All Comments] [Add Comment or Rating]
💡📝Michael wrote on Sat, Sep 26, 2015 01:34 AM UTC:
Thank you H. G. Muller for your input, I am always happy to receive feedback and ideas. The king double move to c2/g2 is interesting, however I feel that the king may not always be too safe on those squares, mostly due to the number and power of the bishops. e.g say the king double moves to g2, and then later the g2 and b7 bishops trade off. Now the king is rather vulnerable on g2, and whites queen does not want to get lured onto the a8/h1 diagonal, as blacks other white squared bishop might immediately spring into action.

Also I don't want to give up on Luft Castling just yet! I agree with what you said about the location of the rooks, but I have found through testing (without special castling) that the luft manoeuvre sometimes tends to happen naturally (manually). It seems to have some advantages, that would be even better if sped up. This is sheer luck, as I can now see that there was a distinct lack of sophistication in how I decided to include it as an option (i.e. "okay so let me find something that is like castling, but with an extra rook"!). It has the advantage that it immediately accelerates the development of 2 rooks, while moving the king to a safer place behind some wing pawns and the luft rook. It seems counter-intuitive, but somehow it seems to work, at least from my very limited initial tests.

I'm not discounting the double king move idea either though (I would prefer your first option - i.e. 2 steps to g2 subject to f1 OR f2 not being under attack). I feel it definitely needs to be tested along with the other ideas. And it could even be included along with luft castling and normal castling. I still have a lot of testing to do, which will be much easier when I can I find a way to modify an open source chess engine to include these features.

H. G. Muller wrote on Fri, Sep 25, 2015 07:46 AM UTC:
I think both proposed methods of 3-piece castling miss the most important point of this array: the natural places to hide the King are not b1/g1, but b2/g2. The King not protecting the Pawn shield tremendously weakens the King fortress, and on b1/g1 it would furthermore be susceptile to diagonal attacks through the center.

Castling in orthodox Chess was invented to solve the problem of the Rook getting trapped by King and Pawns in the shield. This array (which is almost Makruk-like) does not have that problem at all. The King could move to b2/g2 without a worry, and leave the Rooks to roam the 1st rank.

So if an extra King double-move would be needed here to speed up development, it would be much more logical to make that a one-time sideway Knight jump, as in Chaturanga or Cambodian Chess (Ouk). That would put the King directly on g2. Note that a Rook on the f-file is not necessarily better than on the g-file; usually it has to be moved after that. Conventional castling only served to prevent its trapping, it doesn't really develop it.

You could make a King double move to c2/g2 subject to the squares d1 or f1 not being under attack, or perhaps f1 and f2 (or c1 and c2) not being both under attack, apart from the usual conditions of not being in check and not having moved before.

2 comments displayed

Later Reverse Order Earlier

Permalink to the exact comments currently displayed.