The board is 4x4, 8 gladiators are generated from Queen, King, Pawn, Rook, Bishop and Knight (this decision came at a later time after I posted this). Gladiators do not move, however they will instantly kill any gladiator in their range (A gladiator is in the range of another if that other gladiator could take him if they were playing chess). If they have several gladiators in range, they will kill them all. For instance:

123|..B|45Q|---+

I own B and Q, the others are my opponent's pieces. B(ishop) can kill 2, but Q(ueen) can't. B can also kill 5. Q can kill 1 and 5. It can't kill 3 because B is in the way. It can't kill 4 because 5 is in the way. However, if B plays first (the pieces with highest bids on them play first), 5 will die and Q will be able to hit both 1 and 4, and also 3 if B dies.

So this is not a deathmatch, stale mate can exist.

Since the players cannot see where the opponent is placing the pieces on the board, what kind of strategies do you project the players would use when placing the bids on the tiles?

Suppose you have K B N R and the enemy has Q N N R
(bid values not shown, assume that all the bid values are the same and player 1 will attack first. arrange your pieces intelligently to get the most kills, without knowing exactly how the enemy will the pieces.)

+---------+| . . . . || . . . . || . . . . || . . . . |+---------+

How would you go about placing your pieces?


Some thoughts:

It seems undesirable that the board is symmetrical: not only you have to guess the pattern, but also the orientation of the enemy's placement. How about the players take turn to bid on the tiles, and after the tiles are bid, the players can then place the pieces?

Actually, instead of bidding, just have them take turn claim a tile. While doing this, you would have an idea of what strategy your opponent is attempting, and you can defend against the strategies. There seems to be more interactions this way.

For example, after the two had already claimed the tiles, but before placing the pieces:

 +---------+4| . 1 . . |  Player 1 is you. You have K, B, N, R3| . 1 . 2 |  Player 2 is the enemy with q, n, n, r2| . 1 2 . |1| . 2 2 1 | +---------+   a b c d

Now it looks possible that the enemy is going to put the Q at c2, and R at b1. The enemy's Q can target three of your positions, the R can target all four of your pieces, the knights can target 2 pieces and 1 piece respectively.

For your pieces, d1 can only be attacked by a queen d2 and a rook at b1. However, as long as the knight at C1 does not die, the rook cannot attack your d1. Your most vulnerable positions are b2 and and b3, where both are targeted by three enemies. However, if you can kill the enemy queen and rook, your rook at d1 will guarantee victory. But is that likely to happen?

Since you are player 1, you will attack first. Thoughts:

 +---------+   +---------+4| . N . . |  4| . N . . |  3| . B . n |  3| . R . n |  2| . K q . |  2| . K q . |  1| . r n R |  1| . r n B | +---------+   +---------+   a b c d       a b c d placement 1   placement 2

which configuraion is better?

For placement 1: K x r     K x q     K x n1    N x n3q x R     r x K     q x R     r x KK x n1    R x n3    K x r     B x qn x N     r x B     q x K     r x BK x q     R x n1    N x n     R x nn x K     r x R     r x B     r x RDraw      r x N     r x N     r x N          Lose      Lose      Loseplacement 2:K x n1    K x q     K x r     N x n3q x K     r x K     q x B     r x KR x r     R x r     N x n     R x rq x R     n x N     q x K     q x BN x n3    R x n     N x q     N x qq x B     n x R     n x R     n x RN x q     Draw      Draw      DrawWin

Do the rules seem manageable for a strategy game?

The problem is precisely that I would want to avoid as much as possible the need for multiple intervention. Turns in this game will never happen in a simultaneous way so it might take days or weeks for the opponent to respond, and you'd just lose track of what was going on or what you were thinking.

However, you have an excellent point that the initial board layout is too white. I guess randomly allocating tiles to players might give the players more data to base themselves on, for instance giving each player 8 tiles out of the 16.

To stick to the low transaction version, you can also have this variation:

1) Initially, the game randomly assigns the two players 4 to 6 tiles EACH on the board, but not 7 or 8. It is because a complete partition seems to introduce too many possibilities.


2) 8 pieces are randomly generated. Possible pieces inlcude K, Q, B, N, R. Pawn is not availble. There are no properties for the pieces except there range of attacks. The pieces can be classfied conceptually like this as follow:

[max # of attacks, max # of unblockable attacks, max # of blockable attacks, pieces that you are invulnerable to if they are your target, pieces that you might be invulnerable to if they are your target]Q = [11,8,6,----N,K]K = [ 8,8,0,----N,-]R = [ 6,4,4,---BN,K]B = [ 5,4,2,--R-N,K]N = [ 4,4,0,QKRB-,-]

3) The two player will analyze what pieces are most valuable for the given tile assignments. The player will consider what pieces the enemy is most likely to bid on, and devices a bidding strategy so that you get the pieces you want or that the your enemy don't get their strongest pieces. The result of the bidding will give each player 4 pieces.

An alternative way is to get rid of bidding and allow the game to assgin a set of pieces for each players to select. However this eliminates the entire original concept of having bids. If you want to keep the biddings I would suggest a different combat system. Currently there are already sufficient strategies so that bidding is not necessary from the gameplay perspective.


4) After the players got the pieces, they will pieces them on the board and submit the placement.


5) Due to the limitation of transactions, the players will not get to play through their strategies. The AI will duke it out. And the outcome of the game is declared.