# Four-player Chess (need some help)

This topic is 4093 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

## Recommended Posts

I'm designing a flash based online 4-player chess playing game. I have most of the details of the RULES of the game figured out but there are a few more that I need. I'm looking for logical answers.. not just answers of taste. Like... for dealing with how pieces moved I tried to logically scale chess up to four playerability. I didn't make any judgement calls. (ok there had to be lots of judgement calls but the way they came out was very logical.. as if it was almost a perfect mathematical extension of two player chess into four player chess) So let me show you what I have so far rule-wise and maybe you guys can offer input on my choices so far and help me with some dilemnas I'm still stuck with. The board looks like this: i took this image from another implementation of four-player chess that i found on the web (http://www.pathguy.com/chess/FourChss.htm) Anyways he seems to have come to the same conclusion about king placement as me. Kings are not facing each other from across the board as in 2P chess. But this is logically equivelant because adjacent players have the same setup as 2P chess. Also this is the only symetrical way to do it. Otherwise you will have one player who is closer to his partner's queens, and his opposing player will be closer to his adjacent partner's kings. So that's settled. Knights/bishops/queens/king movement are simple, castling is not changed. En passant works the same way. If an enemy pawn is on your fourth rank and you move forward two moves, on their very next move they can en passant you. When a pawn enters another players camp, that pawn can only move in one direction: the direction towards that player's first rank. Just like normal chess. They can only move in three directions when in the 8x8 center field. Now for the issues: Question One: When pawns move into the center 8x8 grid, should they be able to move in three directions instead of one? I was thinking they shoudl be able to march towards any of the three opponents and capture in any of the four diagonal directions. I also figured they should be able to promote on the 1st ranks of ANY of the three opponent's starting areas. Question Two: If Stalemate for a player occurs should he simply be able to skip his move? This seems the best and simplest way to handle that situation. Or maybe the army would isntantly vanish. So this would be a punishment for stalemating someone.. instead of getting his army, you don't get it.. and that player escapes (score-wise) with a draw, as in regular chess. Question Three: This is the doozy.. What happens when someone gets checkmated? I see three options: -his entire army magically vanishes. -his army becomes neutral. capturable, but exerting no check or influence whatsoever, just sitting there.. other players cannot move through his pieces as though they aren't there, they must be captured or the player must take a detour to get around them. -his army is given to the checkmating player. These all seem to have some logic to them. The 3rd has gameplay incentive. Otherwise it might be wise to just sit back and let some tactical battles rage on. (though this could be unwise if the victor then gets to promote some pawns.) I'm leaning towards this solution: when a player CAPTURES an enemy king, that enemy's army instantly becomes his own army as well. That player thus becomes much stronger. In so doing, all pawns in the starting territory of the dead player now move outwards.. away from that first rank.. even if they were pawns FROM the victors original camp that had made it into this camp. now they must head back to center and can only promote in the reamaining two players camps. Question Four: (this is two questions) how to modify the 3-move repitition rule and the 50-move rule my ideas: 3 move rule becomes: 5 move rule. (this one is easy and logical 5 repititions means at least once we have the same position with the same player to move.. and it must end in an immediate draw) 50 move rule becomes: 100 or 200 move rule. These are my questions. I have all other issues figured out, including how to update the ratings of the players upon conclusion, and all programming issues are sorted out as well. I'm not even going to try to make a worthwhile AI yet. Pointless for this task unless I want to give the players some AI to practice against. The likelihood of making an AI strong enough to give humans real trouble is very low.

##### Share on other sites
This is a tough one. I can answer question one and question three.

For question one; Keep in mind all that could occur if a pawn is allowed to move in three directions. You could amass vicious blockades, amongst other things. Also, unless you put in some definition of which way a pawn is "facing", a pawn will be able to take pieces in four directions diagonally. Keep them all going in a straight line is probably the best advice I can give you. Make that rule only change if they enter another player's teritory by maybe taking a pawn on a side row (for example, if the bottom player's leftmost pawn took a piece to the right,) but even then, probably just keep them going in a straight line.

As for question three; Do not, under any circumstances, give one player another's army. It's a massive feedback loop just asking to happen. At the same time, don't remove the army, because if you remove it entirely, that might be a clear line of fire to get some pawns through. Leave the army there, but sterile. That's my advice.

##### Share on other sites
i also agree, dont give the army to another player. Sterile, remain.
once a pawn advances in 1 of the 3 directions from the 8x8 it must continue on that heading imo. You cant have pawns retreating, thats just cowardly.

##### Share on other sites
Quote:
 Question One:When pawns move into the center 8x8 grid, should they be able to move in three directions instead of one?

How about having the Pawns have a specific faceing. They could take a turn (instead of moving) to change their faceing by 90 degrees (left or right), but can never face backwards.

Quote:
 Question Two:If Stalemate for a player occurs should he simply be able to skip his move?

This is, I think the hardest question to answer. My best solution is, if they can't move anyother peice (without puting their king in check) and, while the king is not currently in check, any move it makes would put it in check and there are at least 3 players left (more than 2) then that plaeyr is as good as beaten.

You might impose a 3 (or 5 turn limit) which says that if they can't get out of the situation (say be another player forcing the player(s) who is stalemateing you to release the stalemate), then you are considdered beaten. This would be because only oen other piece would be needed to check mate you and the other players have many pieces.

If the whole game is stalemated and no one can break it, then the game should be a draw.

Quote:
 Question Three:This is the doozy.. What happens when someone gets checkmated?

Giving the conquerer the losers pieces should eb a big NO. As sated above, this will lead to a massive advantage for the conquerer in a positive feedback loop that will result in a runaway leader situation (imagine having 2 queens, 4 rooks, 4 knights, 4 bishops or 16 pawns).

I would say, wipe them out and remove them imediately from the board (this could be seen as the army, seeing their king dead, disbands and flees the battle field).

Quote:
 Question Four:(this is two questions)how to modify the 3-move repitition rule and the 50-move rule

I would increase the 3 move rule to a 5 move rule as with more players you are more likely to get repetitious movment.

As for the 50 move rule, I think that because of the greater number of players, this could actually be lowered. However I think that it could be left alone just the same. Increaseing it is not nessesary and just makes it less likey to be put into play (there by removing a valid strategy - to go for a tie instead of a loss). Because the game has more players it will be less likely that the players will go on for a long time without captureing a piece (there are more pieces to capture with and be captured, plus there are 3 moves between your moves - one for each of the other 3 players).

So my recomendation would be to keep the 50 move rule as it is.

##### Share on other sites
My initial reaction is that the board is too big. Have you considered putting all the armies right next to eachother?

111111112
4...........2
4...........2
4...........2
433333333

like so ^

I know that makes a LOT of peices immediately capturable, but might make for some interesting strategy and there wouldnt be any problems with pawn movement.

Regarding your setup, remove checkmating, if a player can make no move in which the king becomes vulnerable he forfeits his turn (kings MUST be captured.) This is because theoretically a player could be mated by all 3 other players at the same time. Once an enemy king is captured his pieces all flee the battlefield, OR

1. His peices are capturable and turned to pawns by the other players (by whatever methods you can come up with.) This would prevent instant road block removal.
2. A player must lose all peices to be considered out of the game (continues playing with some modified rules (eg. he cant win)

Pawn promotion only occurs in an active players first rank. If a pawn reaches an inactive first rank, it then reverses direction. If it gets all the way back to base then promote? I would still have the pawns move in only one direction and capture with the same rules. If you capture into an adjacent players fourth rank that pawn is considered to be advancing in that direction.

There was a chess game on a hexgrid called triple cross that ran into some of the same problems.

##### Share on other sites
Have you considered a higher dimensional board?

2D for 2 players, a cube for 4 players. That would be the locial extension of chess to 4 players in my mind.

Just a thought. To your questions:

1. I would make pawns move in any direction once they cross they move a total of 6 squares forward. Otherwise you have issues of the efficacy of your side pawns vs. the middle pawns.

2. I would have players be able to skip their moves. Sort of similar to scrabble.

3. I would have his army be either neutral or eratic. During any turn that the checkmated player would have had, one of his former pieces makes a random lawful move. Simulating the disarray of an army without a commander.

##### Share on other sites
i dont think players should be able to skip moves.

##### Share on other sites
Here is my feedback to the feedback so far. I appreciate all input.

Quote:
 As for question three; Do not, under any circumstances, give one player another's army. It's a massive feedback loop just asking to happen. At the same time, don't remove the army, because if you remove it entirely, that might be a clear line of fire to get some pawns through. Leave the army there, but sterile. That's my advice.

So he gains from doing so. That's what he gets for having outplayed the other two though right? And maybe the other two are engaging each other and one of them wins an army as well. And if they are playing a more slow game, it is their fault for letting him win the army. It is a game of 'perfect information' after all.

Quote:
 Also, unless you put in some definition of which way a pawn is "facing", a pawn will be able to take pieces in four directions diagonally.

Stated this in the original post already. If they can move in three directions, it would be logical for them to capture in the four diagonal directions.

Quote:
 once a pawn advances in 1 of the 3 directions from the 8x8 it must continue on that heading imo. You cant have pawns retreating, thats just cowardly.

Stated this in original post. Suggested three not four.

Quote:
 I would increase the 3 move rule to a 5 move rule as with more players you are more likely to get repetitious movment.

In original post I suggested five as the logical modification of 3-move repitition.

Quote:
 So my recomendation would be to keep the 50 move rule as it is.

I think you're right. Or maybe even diminish it to a 25 move rule. Four players reduces it. You're right on that. I was looking at it backwards.

Quote:
 I would say, wipe them out and remove them imediately from the board (this could be seen as the army, seeing their king dead, disbands and flees the battle field).

What if this puts me in check? I don't think player A should be safe from player B's checkmate just because player C happens to have a bishop that would put player B's king in check if player A's army was removed.

Quote:
 You might impose a 3 (or 5 turn limit) which says that if they can't get out of the situation (say be another player forcing the player(s) who is stalemateing you to release the stalemate), then you are considdered beaten. This would be because only oen other piece would be needed to check mate you and the other players have many pieces.If the whole game is stalemated and no one can break it, then the game should be a draw.

Imposing such a limit is a perfect example of adding an arbitrary rule to the game. I truly want this to be an extension of chess.

A game wouldn't end in a draw just because one player was in a stalemate. If one player gets a draw that is his score for the game. The game will still go on till all players have mated, been mated, or drawn.

Quote:
 How about having the Pawns have a specific faceing. They could take a turn (instead of moving) to change their faceing by 90 degrees (left or right), but can never face backwards.

This seems like another arbitrary rule addition. But I will definitely give it some thought. It might not be completely arbitrary. I'll have to give it some time to bounce around the board a bit.

Quote:
 1111111124...........24...........24...........2433333333

The initial setup is one of the few things in four-player chess that I'm certain about.

Quote:
 Regarding your setup, remove checkmating, if a player can make no move in which the king becomes vulnerable he forfeits his turn (kings MUST be captured.) This is because theoretically a player could be mated by all 3 other players at the same time. Once an enemy king is captured his pieces all flee the battlefield, OR1. His peices are capturable and turned to pawns by the other players (by whatever methods you can come up with.) This would prevent instant road block removal.2. A player must lose all peices to be considered out of the game (continues playing with some modified rules (eg. he cant win)

I listed a few options for what happens in case of a checkmate. It included 2/3 of those. Your option #3 (labeled '2.') isn't the way chess works logically. The game involves mating or getting mated. Capturing material is just a good way to reach that goal usually.

Quote:
 Pawn promotion only occurs in an active players first rank. If a pawn reaches an inactive first rank, it then reverses direction. If it gets all the way back to base then promote? I would still have the pawns move in only one direction and capture with the same rules. If you capture into an adjacent players fourth rank that pawn is considered to be advancing in that direction.

Stated in initial post that if a territory becomes vacant while a pawn is in it it will reverse direction.

I still disagree with the general concensus that pawns should only be able to move forward. The pawn marches towards its enemy in a game of chess. In 4p chess there are 3 enemies and they are all equally important and symetrical. I don't want to give special preference to players that are directly across.

RE: Numsgil

Very insightful points I think.

Except that 3d chess would not be the logical extension. 3d chess might have 6 players (one per face) or 12 (one per edge) or 12 (two per face) There is nothing symetrically four-esque about a cube. I have written a 3d-chess engine before. (Not a playing engine but an engine for human-human play.) It was difficult to figure out the logical way to extend chess into the 3rd dimension, but I feel I got all the details correct.

Six squares forward. Your efficacy point is strong. I don't think imposing a limit on forward movement before moving sideways solves it though. I think the side pawns will still have the proximity issue. I think it will work ok just letting them move sideways from the beginning.

Your responses to 2. and 3. are judgement calls.. arbitrary additions but #3 is really tempting. I'll have to not give into the temptation to implement random eratic moves. As for move skipping.. that is probably viable.. I'll have to think about that one a bit.

##### Share on other sites
Quote:
Original post by sharpnova
Here is my feedback to the feedback so far. I appreciate all input.

Quote:
 As for question three; Do not, under any circumstances, give one player another's army. It's a massive feedback loop just asking to happen. At the same time, don't remove the army, because if you remove it entirely, that might be a clear line of fire to get some pawns through. Leave the army there, but sterile. That's my advice.
So he gains from doing so. That's what he gets for having outplayed the other two though right? And maybe the other two are engaging each other and one of them wins an army as well. And if they are playing a more slow game, it is their fault for letting him win the army. It is a game of 'perfect information' after all.
Not neccesarily; he might be the sloppiest player on the board and simply get lucky enough to 'capture an army' based on someone else's work. I think I agree that it would be too great a feedback loop; anyone who defeated another player would likely be too difficult to defeat.

Leaving the pieces on the board in an inactive state seems the most logical progression from normal chess rules in that there isn't really anything in chess which can cause such a drastic change to the playing area as the instant removal of an entire army or the switching of sides. I tend to feel that it would be far too difficult for players to actually calculate the effects of such a massive change to the board.

##### Share on other sites
The stalemate question is a stumper. But here's my opinions regarding the thread:

Pawns: Perhaps the player could be allowed to change the direction of a pawn, to the left or right, ONCE per game?

I also think the simplest solution to a beaten player is to leave all of the peices on the field, but not allow them to move/capture. That way some of the roadblocks will still be in place, allowing for more intricate strategic play. I do think it's too much to give a player someone else's army; having two queens, or 3-4 of the other backrow peices, is just too much of an advantage for any player. I also think removing all of a player's peices from the board injects too much chaos into the game for long-term planning.

Also, if you take my suggestion about pawns being able to turn once, then leaving the armies on the fields would work well, since a turned pawn could make a mad dash in that direction, only able to be stopped by a still active player.

It'd also be interesting if the condition for losing wasn't checkmate, but the actual capturing of a king. Other players might take a player out of checkmate before the next turn of the player who initiated the checkmate. In that case, as long as a player was in checkmate, he couldn't move, but it'd still be posible for another player to thwart the checkmate.

Edit: I think it'd be great if the players could select Fischer Random Chess starting positions as well as the standard, especially since this is an unorthodox way to play the game. If you're unfamiliar with Fischer Random Chess, basically the back row is randomized, with these rules:

1. The King is between the rooks.
2. Bishops start on different colored squares.
3. The peices of the opposing side are placed in opposite order.

Now, you could additionally let the player choose if opposite sides follow rule 3, all sides are random but following 1&2, or if all sides have the same random backrow placement.

I think it'd be a great addition.

[Edited by - thundergunslinger on October 1, 2006 10:27:14 PM]

##### Share on other sites
If you do allow a player to capture another players army, he has two big disadvantages to offset his new advantage in army size.

1. The large player will become the instant target of a logical alliance between the other two players (due to both his percieved threat, and the gain of the reward of two armies for beating him). This becomes very bad for the large player, because:

2. The large player still only gets one move per turn. So he will be facing two moves from his opponents for every one move he gets. I'm assuming he will still only have one king. Wouldn't it be very easy to checkmate someone who gets one move for every two of yours, even if you do have a large (but possibly split into both corners of that board) army?

This isn't an rts, it's not like you could win through force of numbers. A huge army isn't that big an advantage if you can only move with one piece a turn.
But I'm not a chess buff, this is all supposition.

##### Share on other sites
get 4 people and play the game with realy chess - just make a board. Change rules after couple of games to see what is realy FUN. I dont think we can come up with idias for FUN based on logic becouse FUN is not logical. Just play the game with freinds 1st!

##### Share on other sites
CIJolly - That's kind of how I'm seeing it as well.

eiforall - Yeah I'm considering allowing the players on the site to specify the rules from a few pre-selectable options. And after the site's popularity grows and the community grows larger, I'll take some polls and start to remove some options slowly.

thundergunslinger - As for capturing the king outright instead of checkmate I think I agree there. As for fischer random (yes I'm quite familiar with it and have played it quite often), that would probably not happen (at least not for awhile). There is a website www.schemingmind.com that implements many chess variants. I'd be interested only in the logical extension of real chess into the four-player dimension.

Kazgoroth - inactive state is seeming to work it's way up to #1 in which option I choose.. but I still don't think it matters if a sloppy player happened to checkmate. You could win a game of chess by being a sloppy player and just lucking out against an incautious opponent. Why should it be any different in four-player chess? See CIJOLLY's post.

##### Share on other sites
Quote:
 Original post by CIJolly1. The large player will become the instant target of a logical alliance between the other two players (due to both his percieved threat, and the gain of the reward of two armies for beating him). This becomes very bad for the large player....

The reason I don't like this is because it does essentially force a 2 on 1 situation. Players might refrain from finishing off an opponent simply because the don't want to force the other two players against them. Keeping a neutral army on the field will give players freedom to develop their own strategies to eliminate either of the two remaining players. I would definitely opposed to a game where this 2 on 1 alliance is almost necessarily formed in every match--you're taking choices away from the player in how he wants to play the game.

Upon reflection, I think having pawns be able to move forward, right, or left is fine. At first I just thought it would be confusing, but then I realized (head slap) because of the different colors, it'd be easy to tell when a pawn is making an illegal (backwards) move.

##### Share on other sites
although kungfuchess.com sports a real time version of chess, it is definitely worth checking out how they handled the details of 4 player chess. I also believe there are various board games to that effect that you could seek out and reffer to.

##### Share on other sites
Quote:
 Original post by sharpnovaRE: NumsgilVery insightful points I think.Except that 3d chess would not be the logical extension. 3d chess might have 6 players (one per face) or 12 (one per edge) or 12 (two per face) There is nothing symetrically four-esque about a cube. I have written a 3d-chess engine before. (Not a playing engine but an engine for human-human play.) It was difficult to figure out the logical way to extend chess into the 3rd dimension, but I feel I got all the details correct.

I was thinking, a 2D chess board has two sides that belong to two players, and two "neutral" sides. If you add 2 more players, that would give you four sides that belong to players. Now that I think of it, though, you'd want to add another 2 "neutral" edges as well. Maybe some sort of octahedron...

Probably too complex. The idea was to give each player a "side" that is surrounded by neutral "sides" (sides could mean faces in 3D). If you have 4 players playing on an octahedron, I think you could do this without too much effort. Wouldn't really be chess anymore though.

##### Share on other sites
Here is a different setup:

....AP....PB........AP....PB........AP....PB........AP....PB....AAAA________BBBBPPPP________PPPP....________........________........________........________....PPPP________PPPPDDDD________CCCC....DP....PC........DP....PC........DP....PC........DP....PC....

"." are blank spaces
"_" are impassable
"P" are pawns
"A" through "D" are the pieces of players A through D.

The board is now a circular loop -- each player has half of his or her army facing two other players.

The middle of the board is a no-mans land.

Behind each player's troops is a bunch of empty space -- this could be made impassable.

I will note that 4 player chess would probably work out better if it was two teams of 2 players -- there is a reason why there are very few 3 way wars. The players would naturally form alliances -- and unless there are predetermined alliances, it is quite possible that a 3 way alliance will form.

Once alliances are fluid, the game becomes more about the forming of alliances than it does become about the actual gameplay.

Yet another varient:
    AP....PB        AP....PB        AP....PB        AP....PB    AAAA////\\\\BBBBPPPP///  \\\PPPP....//    \\......../      \........\      /........\\    //....PPPP\\\  ///PPPPDDDD\\\\////CCCC    DP....PC        DP....PC        DP....PC        DP....PC

On this map, the topology of the map is strange. When you enter the center of the board, the diagonal lines indicate where you end up.

The above layout has the advantage that it allows N player chess -- any number of players can be connected together in the above way. It does get rather confusing -- like I said, the board topology is seriously wierd! :)

[Edited by - NotAYakk on October 3, 2006 2:25:53 PM]

##### Share on other sites
For the pawn rotation question, consider this:
Draw two diagonal lines on the center square of the board, from topright square to bottomleft and topleft to bottomright. When a pawn is on a square with a diagonal line through it, the player may choose to rotate it to face that direction(as if the line was a mirror or something.) This would prevent players from putting a bunch of pawns in the same row or column.

##### Share on other sites
Funnily enough, when chess started, it WAS 4 player (2 player teams, if memory serves). It was also played with dice until gambling devices were outlawed in Persia, at which point it turned into a deterministic 2 player game.

##### Share on other sites

This topic is 4093 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

## Create an account or sign in to comment

You need to be a member in order to leave a comment

## Create an account

Sign up for a new account in our community. It's easy!

Register a new account