Idea for updated Chess960 castling

In Chess960, short castling is always to the h-side of the board, and long castling is always to the a-side of the board.

In my version, short castling is always to the side of the board that the king is closer to, and long-castling is always to the side of the board the king is further from. I feel like this might make more sense and be more intuitive.

What do you guys think?

I think your castling method and the method actually used are of equal merit. Both are reasonable. Changing from one to the other now would not be desirable because of the confusion caused.

@Brian-E said in #2:
> I think your castling method and the method actually used are of equal merit. Both are reasonable. Changing from one to the other now would not be desirable because of the confusion caused.

Yeah, I'm split 50/50 on it. From a practical standpoint, I agree that 960 is probably too entrenched to change. But from a pure game mechanics standpoint, I think 960 castling feels a bit arbitrary/contrived. I feel like my castling might be simpler and more elegant.

But there are some arguments against my version too. There would only be half the positions (480) because the castling would be symmetrical. Also, when it comes to notation O-O and O-O-O wouldn't designate specific end squares for the king and rook, but would designate one of two possible end squares for the king and rook depending on the king's initial position.

It's hard to decide what's better lol

@Prophiscient said in #3:
> [...]There would only be half the positions (480) because the castling would be symmetrical.[...]

I don't think that's correct. Surely there will be 960 possible ways of setting up the position regardless of how the castling move is carried out.
Apart from that, your points are well taken.

@Brian-E said in #5:
> I don't think that's correct. Surely there will be 960 possible ways of setting up the position regardless of how the castling move is carried out.
> Apart from that, your points are well taken.

There would be 960 unique ways of arranging the pieces, but half of those arrangements would be a mirror of the other half, so theoretically, gameplay should be equivalent.

Imagine the traditional starting position except the king and queen are swapped. If there is no castling (or there is symmetrical castling), then theoretically, the game should be equivalent to the traditional starting position, just mirrored.

So with symmetrical castling, there would be 960 unique arrangements that are possible, but only half would have unique gameplay.

In terms of pure mechanics (popularity and confusion aside), do you think Chess960 or my version is better?

Ah yes, you're right, I was too entrenched in the mindset of kingside and queenside, but there would indeed be no such concept if the castling was symmetrical.

> In terms of pure mechanics (popularity and confusion aside), do you think Chess960 or my version is better?
I think they are of equal merit.

@Brian-E said in #7:
> Ah yes, you're right, I was too entrenched in the mindset of kingside and queenside, but there would indeed be no such concept if the castling was symmetrical.
>
>
> I think they are of equal merit.

Interesting. If they are of equal merit, then I'd probably prefer 960 castling cuz I generally prefer more positions. But I also feel like my castling is simpler and more intuitive which makes it tough to decide. I guess it's a stalemate for now lol

I think I prefer 960 castling because of more positions. Also, short castling is better than long castling the majority of times; So in your version, if the king is close to the edge there's little reason to long castle since it takes even longer to set up.

@Finnfinity said in #9:
> I think I prefer 960 castling because of more positions. Also, short castling is better than long castling the majority of times; So in your version, if the king is close to the edge there's little reason to long castle since it takes even longer to set up.

Interesting. So in Fischer Random, the king might have to move anywhere from 0 to 5 squares to short castle and anywhere from 0 to 4 squares to long castle.

In my version, the king might have to move anywhere from 0 to 2 squares to short castle and anywhere from 2 to 4 squares to long castle.

In terms of setting up castling, it will be easier to do it on whatever side the king is closer to in both my version and Fischer Random just because it requires less pieces to move. I guess that means long castling would generally be easier in many Fischer Random positions (basically when the king starts closer to the a-side). So I guess long-castling would have more utility in Fischer Random. That's a pretty good argument and the extra 480 positions helps too.