Are there any inherent problems with point-symmetric random chess?

Question

The regular setup of chess is mirror-symmetric: If you mirror the positions of all figures on the equator of the board (between the fourth and fifth rank) and swap the colours, nothing changes. It is also almost point-symmetric or rotational-symmetric: If you reflect the positions on the centre of the board (or equivalently rotate the board by 180°) and swap the colours, only kings and queens will swap their positions.

By contrast, in Chess 960 a.k.a. Fisher random chess, opening positions are mirror-symmetric by construction, but in general far from point-symmetric. Now, suppose I want to introduce a Chess 096, where white’s setup is determined like in Chess 960, but black’s setup point-mirrors white’s. Castling rules would be point-symmetric as well. A possible starting position would be this:

rnnkbqrb/pppppppp/8/8/8/8/PPPPPPPP/BRQBKNNR w KQkq - 0 1

Like in regular Chess or Chess 960, the only imbalance is that white has the first move – unlike for example, if white and black’s starting positions were randomised independently. My question is: Are there any inherent gameplay problems with these point-symmetric starting conditions? For example, does this tilt the balance of the game?

Answer 1

Consider the following initial position:

[FEN "rkqnnbbr/pppppppp/8/8/8/8/PPPPPPPP/RBBNNQKR w - - 0 1"]

By putting the kings on opposite sides of the board, you've more or less introduced an opposite-side-castling situation, without anyone having castled.

Let's say White starts throwing his queenside pawns forward towards Black's king. In this position, this can be done without compromising his own king's safety, as it's on the other side of the board. What are Black's responses?

• Launch his own attack by throwing pawns towards White's king. But given the symmetrical nature of the position, it seems inevitable that White's attack will hit first.

• Castle long, to the side White's king is on. This runs into several difficulties. There are a lot of pieces in the way - there's no real way to do so before move nine. Black will have to move two pawns on the "long" side to get his bishops out, which may weaken his castled position, while White technically wouldn't have to move any pawns to castle over there (and White's king is already "there" even if he doesn't castle at all.) Also, if we're going to keep symmetry, Black's king should end up on the f-file (it's a long castle, after all) while White's will be on the g-file.

• Leave his king over there, and defend. This may work, although it doesn't seem ideal.

Also note that the queenside attack doesn't have to start on move 1; it can be delayed a few moves to establish a center first.

Overall, it seems to me that this variation may increase White's advantage, especially in scenarios like this example where the kings are very far apart.

Answer 2

Look at 'normal chess', but with rotational symmetry.

[FEN "rnbkqbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w - - 0 1"]

I haven't analyzed this opening much, so I may be missing some concrete lines that make this game degenerate. Barring that, I suspect this version of chess offers a few advantages over normal chess.

• Much less opening theory is available/known for this position, allowing for more over-the-board creative play.
• The most common castling combination in chess is that both players castle kingside, in part because there are fewer pieces in the way. If both players castle 'kingside' in this variant, they have castled on opposite sides, which often leads to more aggressive, 'asymmetric' middlegame play and more decisive results.
• For white, putting early pressure on c7 here is harder than putting pressure on f7 in normal chess. Normally, 1. e4 is considered aggressive because it releases both the light squared bishop (which threatens f7) and the queen. Here, we need to move both the d2 and e2 pawns to let both those pieces out. The same also applies to black for threatening f2. I suspect this leads to fewer 'sharp' openings (if the opening theory was worked out).

Lets try playing 'normal' openings and see how they compare...

• 1. e4 e5 2. Nf3: Usually black has to defend the e5 pawn, but here it is already defended by the queen, so black has more options. Further, 2..Nc6 is a common response, which is still valid here and has the added advantage of making progress toward castling 'kingside'.
• 1. d4 d5 c4 : The queen's gambit feels pretty strong. It seems to have the aggressive advantage of the king's gambit in normal chess, but without the danger of moving the f2 pawn.
• 1. c4 : For the same reason as above, the English opening seems strong.
• 1. g3 d5 2. Bg2 Nf6 seems stronger for white than it normally is, but by no means dominant.