SwapMove Strategy

A pawn buys you one teleport. Here is how to know whether you got a good price.

If you come to SwapChess from classical chess, your instincts will mislead you in a specific, predictable way: you will value pawns at one point and conclude that spending one for tempo is usually fine. It is not. In this game a pawn is also a portal, and every pawn you spend is one fewer destination you will ever have again.

This page assumes you already know the rules. If not, start with How to Play.

1. Pawns are a depleting resource, not a currency

Count your pawns. That number is an upper bound on how many SwapMoves you will make for the rest of the game — and it only ever goes down, because swaps consume pawns and captures consume pawns and promotion consumes pawns.

This produces an asymmetry that has no equivalent in classical chess. Early in the game, with eight pawns, a swap costs you one-eighth of your mobility budget. In an endgame with two pawns left, the same swap costs half of it. The price of a teleport rises through the game even though the material cost stays at one pawn.

Practical consequence: be more willing to swap in the opening and middlegame than your material instincts suggest, and much less willing in the endgame. A player who hoards swaps for a decisive late attack usually finds that by the time the attack arrives, the pawns funding it have been traded off.

2. Position matters more than the piece

Beginners ask "which piece should I teleport?" That is the wrong question. Ask "which square do I want to occupy?" — because the set of squares available to you is exactly the set of squares your pawns are standing on, and that set is small and shrinking.

Your pawn structure is therefore also your teleport network. A pawn on g6 deep in enemy territory is not just an advanced pawn; it is a permanent invitation for any of your pieces to arrive there. A pawn chain locked on its home squares gives you eight destinations, all of them useless.

3. The colour-changing bishop

This is the strongest single justification for spending a pawn, and it is unique to this game.

In chess, a bishop is condemned to one colour complex for the entire game. If all the action is on light squares and you hold the dark-squared bishop, that piece is close to worthless and there is nothing you can do about it. In SwapChess there is: teleport it onto any pawn standing on a light square and it is a light-squared bishop forever.

A bad bishop is worth roughly two points. A good bishop is worth roughly three and a half. If a single pawn converts one into the other, you have paid one point for two and a half, and the conversion is permanent. Take that trade nearly every time it is available.

The same logic applies weakly to knights (relocating a rim knight to the centre) and barely at all to rooks and queens, which can already reach most of the board.

4. Swaps ignore blockers — so nothing is ever really defended

The instinct that a position is safe because the lines into it are closed does not survive contact with this rule. Your opponent's queen does not need an open file. It needs a pawn.

Before every move, run a two-second check: which of my opponent's pawns are near my king, and what could land on them? That is the entire threat surface. It is a much shorter list than "all possible attacking sequences," which makes SwapChess defence oddly tractable once you build the habit.

The corollary is offensive. Getting a pawn near the enemy king is worth more here than the space it gains, because it converts every piece you own into a potential attacker of that square. Pawn storms are stronger in SwapChess than in chess, and they are stronger even when they stall.

5. King safety is a trap in both directions

The king can teleport. This feels like a defensive superpower and is mostly a liability.

The seduction: your king is under attack, you have a pawn on the far queenside, you teleport the king there and the attack evaporates. Sometimes this genuinely works and it is spectacular when it does.

The problem: a king that teleports arrives alone. Your defenders are still where you left them, and they cannot follow without spending more pawns you no longer have. You have not escaped an attack; you have relocated your king to an undefended square and given your opponent a fresh target with a full tempo to redeploy.

The mirror-image mistake is defensive and much more common: swapping a piece onto a pawn in your own castled shelter. The f2 or g2 pawn you just removed was doing real work. Placing a rook there in exchange for opening the diagonal into your king is frequently a losing trade even though it looks like you added a defender. Count the attackers on the resulting position, not the intention behind the move.

6. Tempo is real but overrated

A swap can move a piece across the board in one move where a normal route would take four. That is a three-tempo gain for one pawn, which sounds excellent. It usually is not, for two reasons.

1. The piece arrives with no support. Tempo gained by teleporting a single piece into enemy territory tends to be given straight back when that piece is attacked and has to retreat — on foot, this time.
2. You still only made one move. A swap consumes your turn. You have not gained a move; you have made one very long move. The three-tempo figure counts a journey you were never obliged to make.

Swaps that pay off are the ones that produce an immediate concrete threat — a fork, a mate threat, a defended outpost — not the ones that merely improve a piece's address.

7. Common blunders

8. A checklist before every swap

1. What concrete threat does this create this move? If the answer is "none, but the piece is better placed," think again.
2. What was the pawn I am spending doing? Shelter, passed pawn, chain base, landing pad?
3. How many pawns will I have left, and does my plan need more swaps than that?
4. Does removing this pawn open a line toward my own king?
5. Can the arriving piece be attacked immediately, and can it be defended if it is?

Five questions, most of which resolve in a second once you have the habit. The players who beat the Master bot consistently are not the ones who swap the most — they are the ones who can say what each swap bought.