Modifying chess conflict rules

I set myself the task of modifying chess in just one of the nine fundamental sub-systems of games: interaction of assets/conflict. Chess has a very simple conflict system: the attacker always wins via “displacement capture”. What happens if we change that?

The obvious change, to me, is to use a randomizer such as dice so that the attacker doesn’t always win. When one piece attacks another, dice are rolled and a comparison determines which piece survives.

In the first method a different multi-sided die is assigned to each kind of piece. A pawn is a D4, a bishop or knight is a D6, a rook is a D8, and the Queen is a D12. Or if you really want a game where the Queen runs amok (I don’t) make it a D20. The attacker adds one to his die roll. So if a pawn attacks a bishop we have an even fight, D4 +1 versus D6. If a rook attacks a pawn it’s D8 +1 versus D4. What happens if it’s a tie? If we give too much advantage to the defense, such as ties are won by the defense, the game could become very static. So I’m going to choose the more exciting alternative which is to reroll ties. (Another possibility: both are eliminated.) I’ll get to the King in a moment.

The drawback of this method is that you need a collection of special polyhedral dice. So in the second method we just use regular D6s. A pawn rolls one D6, a Knight or Bishop rolls two D6s, a Rook rolls three and a Queen rolls four (or again if you want the Queen to run amok have her roll five). The attacker gets an extra die. Reroll ties.

Now in either case, what about the King? The first alternative is that the King is an exception to the rules, with infinite attacking strength and no defending strength. So the king always wins when he attacks, and the usual chess check and checkmate rules apply. In other words we haven’t changed the standard chess rules for the King.

The second alternative is to treat the King as another part of the combat. In this case the King has a dice strength and participates in combat like any other piece, and a player is not obliged to call check when he threatens the King. The question is what dice to give the King. My first thought is the same as the bishop and the knight, enough so that the King can take chances and might actually survive an attack but not so much that he becomes impregnable.

Now we can go further by introducing “support” from nearby units into the conflict. For example, one way to do this would be to add one point to the die result for each piece of the rolling player that is in one of the eight squares adjacent to the conflict square. Another way would be to add one point for each piece of the rolling player that can move to the conflict square (although they don’t actually move). One point is going to mean less in the dice method that uses several D6s, and some players might want to add an entire die for each supporting piece, but then the support becomes more important than the identity of pieces in actual conflict. Nonetheless this support rule should help prevent a raid by one strong piece that just wades into the massed ranks of the other player and hopes for decent luck. Imagine a rook or Queen just attacking one piece after another and accepting counterattacks while hoping that its superior dice carry the day. The support rule is going to discourage that.

Now as I write this I have not tried to play any version of this game with an actual board and pieces. I am trying to anticipate what might happen, that is, I am playing the game “in my mind’s eye”. I think this is much more efficient than not anticipating what might happen and simply playing the game. But that depends on your situation and on how you do things.

Alternative randomization
Another way to use a randomizer is to give each player some cards from a standard deck of cards. When there’s a conflict, each player chooses a card from his hand and places it face down on the table, both then reveal their cards, and the strength of the cards is added to the strength of the pieces. A card once used cannot be reused. Each player gets an identical selection of cards, say four aces (1s), four 2s, and two 3s for a total of 10. When these 10 cards are used up they become the player’s new hand.

What is the strength of the pieces? A pawn would be a 1, Bishop and Knight 2, Rook 3, and Queen 4 or 5. If the King were used as no ordinary piece rather than an exception that it might have a strength of 2.

The problem I see in my mind’s eye is that players might simply play the largest cards first in order when the initial battles and hope that the consequent material superiority would carry them through even though the other player might have better cards later.

This card method could result in “card counting”, that is, memorizing what cards have been used in order to know what’s left in your opponent’s hand. To avoid that memorization problem - memorization has nothing to do with chess and is generally not desirable in modern games - I’d say that all cards that have been used can be viewed by the opponent so he’ll know what you have left.

This can be further varied by increasing the size of the deck but allowing the player only three or four cards in hand at a time. This introduces uncertainty and further requirement for hand management. If you happen to get really good cards to start with you had better use them well because you’re going to get worse cards later on.

If you use the point values normally assigned to chess pieces and varied the cards much more, with higher numbered cards, that might work better.

Some people will prefer this method strongly to dice because they feel they can manage their cards, and there’s an element of bluff and Yomi (reading the other player’s intentions) that doesn’t exist when you simply roll dice.

I would be unsurprised to find that other people have devised fairly similar chess variants over the years. These methods are original to *me*, but that doesn’t mean they’re original to the world. That’s often the case with game mechanics.

Alternatives to displacement capture
Chess is a game of perfect information, with the only uncertainty coming from the other player’s intentions, but a minimax player can maximize his minimum gain regardless of the intentions of the other player. What we’re doing with the methods above is adding uncertainty to chess conflict.

Can we alter the chess combat system without changing its perfect information nature? In other words, can we avoid uncertainty in combat but change how combat takes place, change from displacement capture to something else. This has been done, more or less, in many variations of chess collectively called “fantasy chess”. New kinds of pieces are devised with new movement and capture methods. For example some may capture by moving away from an opposing piece. Others may capture by forming a pattern around an opposing piece, for example the piece moves so that it and another piece of the same player are on opposite sides of the piece to be captured. There are all kinds of jumps and hops and oddball captures from a distance. I haven’t looked into this for many years but there are chess variant web sites and there are certainly sites about fantasy chess that you can look at.

But how else can we alter the conflict system for all the pieces without changing the movement of the pieces and the perfect information nature of chess? A simple change would be, when a piece moves it attacks a piece in an adjacent square, perhaps only orthogonally (up down and sideways but not diagonally). So a piece cannot move into a square occupied by another piece, opposing or otherwise. An attacker could often have more than one potential target, but he would only be able to attack/take one of them.

Some variation on the support I mentioned above might work, but I’m afraid it would result in two great defensive masses of pieces and regular stalemates.

Another method would be, you move your piece to a square, then it attacks a piece that it could move to if it could move again. That might make the attack quite strong, and would certainly make for a chaotic game. Somehow I feel it might be better if combined with one of the dice methods.

Helping balance the game?
Could we find a conflict system that gives an advantage to the black player, who has much less chance to win than the white player in standard chess? That would need to be a very small advantage or we might find that black wins more than white.

Perhaps we could introduce an input randomizer as opposed to the output randomizers suggested in the methods above. (These are Geoff Englestein’s terms: an input randomizer occurs and is known before players act and (frequently) all players are affected by it as they play; an output randomizer occurs after players have acted and then they get a result as modified by the randomizer. Most conflict methods in wargames are output randomizers.) Because the black player moves second perhaps we could find a way to use the input randomizer to give him the last choice.

Unfortunately, nothing has come to mind that would be sufficiently small to balance rather than overbalance the game. Perhaps someone else will think of something.