Team-Mate Chess

The guiding idea behind the design of Team-Mate Chess is that no single piece should have the capability to force checkmate against a bare King, but that you would always need at least two pieces (i.e. a team) to do that, in end-games similar to a Knight + Bishop mate in orthodox Chess. To provide as large a variety of such team checkmates as possible, all pieces on the back rank are different. This way there are already 21 different possible pairs of pieces.

Promotion can in theory give rise to an additional 7 homogeneous pairs, but as you are allowed (and in fact likely) to promote to an 8th piece type that is not initially available, it also adds 7 new heterogeneous pairs. Of these 28 heterogeneous pairs, only one combination (Knight + Mortar) is not able to checkmate a bare King. There are two non-mating homogeneous pairs (Knights and Mortars, and of course all pairs of color-bound pieces that are on the same color), but they are composed of pieces that you would virtually never consider as a promotion choice in the first place. (Although a Mortar in theory could deliver a check that no stronger piece can, so M+M would be as common as 3 Knights in orthoChess.)

So for afficionados of the B+N end-game, this should be the variant of their dreams!

Play Team-Mate Chess against the computer through Jocly

Setup

Pieces

Kings (K) move as in orthodox Chess, but can castle with either corner piece as if these were Rooks.

The Acromantula (A) is a bent slider, which moves the first step of its trajectory orthogonally, and then turns 45 degrees in either direction to continue as a Bishop. So it can make Knight moves, but only if the corresponding W square is not occupied. It is a color alternator, worth approximately 8 Pawns.

The Adjutant (J) is not present in the initial setup, but is available as a promotion choice. It moves as Bishop or Dabbaba-rider (BDD), i.e. as Rook but jumping over all odd squares. It is color bound, and worth about 7 Pawns.

The Cobra (C) is a sliding version of the Gnu (Betza NC), which can reach the Camel squares only if the corresponding Knight square is not occupied. (It can also be viewed as a limited-range version of the Griffon.) Its value is 5.5.

The Unicorn (U) moves as Knight, or one step orthogonally (WN). It is a color alternator. Its value is about 5.

The Phoenix (F) moves one step orthogonally, or jumps two squares diagonally (WA).

The Knight (N) moves as in orthodox Chess, and is a color alternator. Like the Phoenix and Elephant it has a value of around 3.

The Elephant (E) moves one step diagonally, or jumps two squares diagonally (FA). That makes it color bound.

The Mortar (M) jumps two or three squares diagonally (AG). That makes it color bound. It is the weakest piece, value 2-2.5, but its deep forward leap gives it some dangerous forking power, offering good prospects to trade it for something stronger in the middle game.

Pawns (P) move as in orthodox Chess: they have the regular initial double push and e.p. capture. They can promote to any of the non-royal pieces in the initial setup except Acromantula, or promote to Adjutant.

The Brute (B) is a special-purpose piece for attacking a King; it cannot capture any other piece. It moves (and checks) one step diagonally, or straight backwards.

Rules

Apart from the piece movement, the rules are the same as for orthodox Chess: moving into check is not allowed, stalemate is a draw, Pawns promote on the last rank, without mating potential the game is a draw. A draw can be claimed on a 3-fold repetition. The 50-move rule is stretched up to a more generous 64 moves, however, because some of the 3-vs-1 checkmates take longer than Knight + Bishop mates take in orthodox Chess. And although they can in theory all be forced within 50 moves, it seems unreasonable to expect perfect play.

Pair promotion

In the original design of Team-Mate Chess king + pawn vs king would automatically be a draw, due to insufficient mating potential. I saw this as a flaw, so several years later (in 2023) I added the following rule: apart from the usual promotion to a piece from the initial setup or Adjutant, a player can choose to promote a Pawn that reaches last rank to a pair of Brutes, a newly introduced piece. One Brute will appear on the destination of the pawn, as usual. The other will appear on the square the pawn came from.

The Brute can only capture a king (i.e. check). Never any other piece. When the opponent is reduced to a bare king, this peculiarity is irrelevant. But it makes the Brute a very weak in other situations, less valuable than a Pawn. So an 'unsupported' promotion to an attacked square, in the certainty that the opponent can only capture one of the Brutes at the time, and the other will survive, will in general not be a good idea.

On a non-capturing promotion the Brutes will appear in a constellation such that the rear one is protected by the foremost one against capture by a King, and together the Brutes will attack all squares through which a King could approach them. This makes the rules for whether a KPK ending is won or a draw exactly the same as in orthodox Chess: if the defending king cannot reach the promotion square on the half-move following the promotion, the Brutes will survive until their king comes to their rescue, and together they can force checkmate.

This pair promotion can also be an attractive option in cases where both sides have a single non-royal piece, and one has a Pawn. When the weak player would defend against promotion by sacrifycing its remaining piece on the promotion square for the promoted Pawn, the single piece that is left could not force checkmate. But a pair promotion in this case would leave one Brute, which can force checkmate together with any of the other pieces.

Notes

Because the Acromantula is excluded, the strongest promotion choice would be the Adjutant. But as the latter is color-bound, and the promotion square might be of the same color as that of the only other piece you have, choosing it might not give you a team with mating potential. You would then be forced to under-promote to a non-color-bound piece. (In which case you would probably prefer Cobra or Unicorn, unless a necessity to check dictates the choice for a Phoenix.)

The pieces were chosen to make even mating with pairs of them non-trivial. When one of the pieces is color bound, the board corners are not equivalent, and mate might only be possible in corners of a specific color (This happens for Elephant + Knight, Mortar + Elephant, Mortar + Phoenix, Mortar + Unicorn, Mortar + Adjutant).

A difficulty with color alternators is that they cannot tri-angulate, so that non-trivial manouevring might be needed to lose a turn.

The Unicorn is upward compatible with a Knight, but this can also be a curse, because the extra moves would stalemate a bare King. So you sometimes need complex manouevring to achieve what would have been simple for a Knight.

Some mates need the 'Ferz + Zebra' pattern, where the winning King is on c3, one piece traps the King on a2-b1 by attacking a1, a3 and c1, while the second piece can then position itself at leasure to attack a2 and b1 simultaneously.

The Fairy-Max version included with WinBoard 4.8.0 will be able to play this variant.

Last revised by H. G. Muller.