Chess on a Go (18x18) board

Hi, I'm thinking of making a chess varient to play with friends which would be played on a Go (18×18) board.

However, I'm not very good at chess yet so I don't know whether my varient will actually end up fair and balanced, and I'm mulling between different board arrangements, so I'd like some advice.

The basic rule changes would be:

-The board is 18x18

-Pawns can now go forward 3,4,5,6, or 7 spaces first move as well as 1 or 2.

-When they do this, en passant can capture anywhere along the line they just moved.

-There's a 100-move rule instead of a 50-move rule.

In my head, the advantages of this varient are:

-It's easy to set up with a Go board and chess pieces.

-Go players talk about the added complexity of their game due to the board size, it might be interesting to see this in chess.

-The additional pieces would have a greater feel of controlling an "army".

-For a bad player like me, the increased piece count would mean games are less decided by a single blunder of a piece.

-Having >3 knights means you can get knight-only checkmates.

I'm deciding between 4 different piece arrangements:

1A:

RBBNNNNBQKBNNNNBBR

1B:

RBNNNNNBQKBNNNNNBR

2A:

BBNNNRNBQKBNRNNNBB

2B:

BNNNNRNBQKBNRNNNNB

The idea of all of these is to have extra knights closer to the middle since they'll take longer to get there.

The 1 postions let the rook move further in casting, giving it more of an advantage but it's harder to do. Also the edge pawn is protected by the rook like in normal chess.

The B postions make more of the extra pieces knights to compensate for their weaker power due to a bigger board. However, this means you need more chess sets to set up the board, and I might be overcompensating a bit.

Which setup would be the most fun and balanced? Would any of them be?

What do people think?

If you want to stick with orthodox pieces, I would try to keep their ratio the same as in FIDE. (As it is close to ideal there). Too many pieces of the same type will make it boring. This would also reduce the number of sets you need.

With 2 sets you could use one of the Kings as 3rd Queen by wrapping a rubber band around its head. If you want to experiment with unorthodox pieces you could also give this 'marked King' the role of an Amazon (moving as Q or N).

I would recommend to start the pieces a bit more forward, e.g. Pawns on 3rd rank, King and Rooks on 1st rank, the other pieces on 2nd rank. Then you don't have to wait excessively long before you can castle.

When I was making the pieces, I was worried about adding extra major pieces because

a) adding extra rooks would complicate casting

b) as mentioned, I'd expect knights to be weaker on a bigger board so you'd need more of them to compensate and

c) I was worried adding extra major pieces would make checkmating too easy, since you can now make powerful batteries and such you can't in normal chess.

I'd expect a) is an issue regardless, but are you saying b) and c) aren't the problems I thought they were?

I don't want to complicate things by adding extra piece types or 3 rows of pieces, I think a bigger board with more pieces is enough already.

Although it sounds like you prefer the 2 varients with a conventional setup in the centre for ease of casting?

In that case, I guess this setup might work:

3:

QBBNNRNBQKBNRNNBBQ

That would have equal knights and bishops with 2 extra queens so there are more major pieces too.

I don't see (a) as a problem. Castling should be considered a property of the location of a piece, not its type. So you can castle with the piece in the corner, no matter what type it is. This is what most variants do. (Especially those without Rooks...) You could also start the pieces such that there are only two Rooks on the same rank as the King.

I think (c) is inherent to making a larger variant with more pieces. Batteries are not the only way you can aim multiple attackers to the same square. It is even more unhealty to be under attack by 4 Knights than by a battery of 4 Rooks.

As to (b): Knights are indeed nearly useless pieces on such a large board, but not every shortcoming can be remedied by quantity. The problem with the Knight is that it is slow, and having more of them doesn't make them any faster. On the contrary; it means it will take ever so much more moves to get the Knights to where they can threaten the enemy King by breaking through the swarm of enemy Knights that is defending it because they were close by to start with.

This is actually the main problem with variants on large boards: the slow pieces start in defensive locations, so that the defender has a huge local majority until you bring your own slow pieces to the attack. But that takes a very long time.

This 'ease of castling' idea seems a bust. Castling is not a goal in itself; it is a method of bringing your King to a safe location without trapping the corner Rook. Moving it two steps out of the central file does next to nothing for the King safety, and would also not trap any of the Rooks outside of it, which would still have plenty of room to move laterally. Actually moving all Rooks to the same side of such a centralized King would dangerously deplete your powers on the other wing .

Okay based on your feedback here's a corrected version (with rules amendments); altjoughbit is a shame it's falling away from the simplistic "chess with a Go board idea"

4:

-Pawns can now move up to 12 places turn 1. This is to mimic the distance from the frontrow in normal chess, as pawns are also slow.

-No castling as it's to convoluted to make it work.

-The Backrow is:

RRRRBBBBQKBBBBRRRR

-However in addition, the centre of the board looks like this:

This is so the Knights can get in on the action earlier.

Would that setup work better?

Interesting concept, I have never seen a setup like that. I am a bit worried that the Knights would quickly trade each other away, if you start them in such a dense cluster. It might be better to spread them out a bit, so that it is not so easy to protect each other. If they are unprotected they cannot attack enemy Knights. Perhaps you should start them on the 9th rank (and black on 10th), with 2 empty squares between them. Then you can only place 6, but that should be enough. (Or you put 2 in a more backward location, for defensive purposes.)