lichess.org
Donate

Modern Free Chess

Hey chess players,

this post may eventually become a bit longish, so here is the short version:

I am proposing a new chess version, that includes only those Chess960 positions, where the kings are placed on E (204 positions) and call it 'Modern Free Chess'

But why???

Its probably best to start with why Bobby Fisher proposed the chess variant that became Chess960. In 1992 he argued, that random starting positions could make chess more interesting again, '(...) because chess is becoming more and more simply memorization (...)'. With 'simple memorization' he is of course referring to opening theory, whose impact rose substantially over the last century, especially with the emergence of computers.

To those, who are not familiar with Chess960 - in Chess960, the starting positions of the pieces in the 1st row are shuffled with only two constraints:

* There must be a black and a white bishop

* The king must be placed somewhere between the two rooks

This leads to 960 different starting positions (black and white always have the same (mirrored) starting positions). A random position can easily be chosen with a six-sided die. However, the castling rules are quite counter-intuitive, which makes the game less friendly for beginners and less aesthetic: the position of the king and the rook AFTER castling are the same as in classic chess. There is thus an a-side castling (queenside for classic chess) and a h-side castling (kingside for classic chess). These rules can result to the king and the rook moving in the same direction while castling.
(For further information see (en.wikipedia.org/wiki/Chess960))

Because of this counter-intuitive castling, John Kipling Lewis proposed a different variant in "Castling in Chess960: An appeal for simplicity". The new castling rules are more intuitive: the king always moves two squares towards the rook, then the rook is transferred to the 1st square, the king has crossed (if it is not already there). Since the castling rules do no longer introduce an asymmetry (as in Chess960), there are only 480 positions left, that are effectively different. (The other 480 are just mirrored and lead to mirrored games) Thus, the new version was called Chess480.
This new version however has another downside: quite often, castling results in bringing the king to the centre rather than protecting it in the corner of the board, which contradicts the original intent of castling.

The version I introduced above (2nd paragraph) is in fact the intersection of possible Chess960 and Chess480 games. It only includes the positions, where both give the same rules for castling, thus avoiding the flaws of counter-intuitive castling and the possibility of castling towards the centre. It can be constructed from Chess960 with the additional constraint

* The king is placed on E

which leads to 204 different positions.
In order to popularize this new variant, I gave it a catchier name than Chess204:

Modern Free Chess

A few additional notes:

* as in Chess960, it is recommended to always play 2 games with a given starting position, the second being played with swapped colours.

* a six-sided die can easily be used to construct a starting position (the same way as for Chess960, with just one new constraint):
> 1st: Place the white king on E1
> 2nd: Roll the die to find a position for the black bishop: 1-3 are the free black squares from left to right. Roll again if > 3.
> 3rd: Roll the die to find a position for the white bishop: 1-4 are the free white squares from left to right. Roll again if > 4.
> ATTENTION: From here on, if there is only one free square left either on the right or on the left side of the king, place a rook on it.
> 4th - 6th: Roll a dice three times to find a position for the queen and the two knights. (looking for all available squares from left to right).
> 7th: Place the rook(s) on the free square(s).
> 8th: Mirror for black.

* I wrote a short python script to generate starting positions, which can be found and run here: (repl.it/LMdD/0). It also creates a Lichess html, which can easily be used to play online: open the html, click on 'Continue from here'.

Hope you enjoy the game!

Cheers

Libavius
This is actually pretty cool. I hope they make it a variant in lichess.

The castling has bothered me in Fischer random because like you said, it's pretty ugly and messy sometimes.

Thanks for the improvement, I'll see if I can get some friends to try this with me.

The "counter-intuitive" argument is completely arbitrary. You yourself even contradicted it when you said about 480 that it will often place the king in the center and thus defeat the purpose of castling, so what on earth makes it more "intuitive" then? Why is this more intuitive than the rook and king ending up on the same square they always do when they castle? I understand you are suggesting getting rid of this ambiguity anyway, but the fact that it is ambiguous shows that it is not in fact intuitive, thus undermining the reason you say this variant is necessary.
my fav would be a chess where both players start with just the pawns and then they place the remaining eight pieces behind the pawns, one piece per move.

Factors which make it more interesting:

a) It is not allowed to place an identical piece on the same column, except there is no other possibility. For example if opp places a rook on e8 i am not allowed to place one on e1, except this is the only square left for the rook. This creates dysbalances.

b) _After_ one has placed the king, he does not have to place his remaining pieces immediately, he can also instead move pawns or pieces if he wants. So for example he can wait until move 30 until he places his queen if he desires to do so. This would introduce an interesting new problem, that of missing information.

The pawns should stay where they are, and the pieces should be placed on the first rank, because this keeps the development problem.

0-0 and 0-0-0 does not exist any more, as one will place the king at the border anyway.

Everything else stays the same. Especially endgame theory, pawn strategy and the development problem stay the same. On the other hand opening theory is anihilated.
The reason why i prefer this over fischer random or the above mentioned variation is that there is no randomicity, the players decide on their own.
@Nezvanova Haa!!1!

@NoLuckOnlySkill Thanks, enjoy the games!

@Snuffington
Thanks for your reply and putting thoughts into this!
With counter-intuitive I wasn't referring to the result of castling but to the act of castling. You are right, regarding the result, Chess960 is probably more intuitive than Chess480, but the castling itself looks weird (especially if both pieces move in the same direction, or the king does not move etc.)

This topic has been archived and can no longer be replied to.