Appendix A: Mobility Maps
Unrestricted mobility (positional pressure) maps for the
king knight in the starting position:
1
move map (black knight represents potential move):
2 move map (moving a piece out of the way or doing nothing takes a
move):
┌───┬───┬───┬───┬───┬───┬───┬───┐
│
│ │ │ │ │ │ │
│
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ │ │ . │ │ . │ │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ │ . │ │ . │ │ . │ │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ . │ │ . │ o │ . │ o │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ │ . │ o │ . │ o │ . │ o │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ . │ . │ . │ │ ▒ │ . │ ▒ │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ │ . │ o │ ▒ │ o │ . │ o │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ │ . │ │ o │ │ N │ │
└───┴───┴───┴───┴───┴───┴───┴───┘
influence
diagram, w knight on g1
If a piece is on our 3-move map for the knight, then it is
possible to attack it or defend it in 3 moves, which include waiting moves or moves which move a piece out of the way.
Keep in mind that we need to take into account the location
of the other pieces on the chessboard when we generate our mobility maps for each piece. If we trace mobility through a friendly
piece, we must consider whether or not we can move this piece out of the way before we can continue to trace mobility in that
particular direction. In Example Number 1 in Appendix E we can see this process in more detail.
So looking at this 3-move map and a diagram of the starting position, We
can determine that the White knight can attack 3 enemy pieces in 3 moves. We can defend 8 of our own pieces
in 3 moves (the knight cannot defend itself).
Here are the mobility (positional pressure) maps for the king
bishop in the starting position:
1 move map (Black bishop represents a potential move):
2 move map (need to count waiting moves or moving pieces out of the way):
┌───┬───┬───┬───┬───┬───┬───┬───┐
│
│ │ │ │ │ │ │
│
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ . │ │ . │ │ . │ │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
o │ │ . │ │ . │ │ . │ │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ o │ │ . │ │ . │ │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
. │ │ o │ │ . │ │ . │ │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ . │ │ o │ │ . │ │ o │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
. │ │ . │ │ ▒ │ │ ▒ │
│
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ │ │ . │ │ B │ │ . │
└───┴───┴───┴───┴───┴───┴───┴───┘
influence
diagram, w bishop on f1.
So (looking at a diagram of the starting position) in 3 moves we
can *potentially* attack 4 enemy pieces and defend 6 of our own pieces.
Here are the mobility maps for white's queen and king rook
in the starting position - keep in mind that we are taking in to account the presence of other pieces on the board and the
fact that we must spend one move to move a friendly piece out of the way, if necessary:
|
1st
order queen mobility map
|
|
2nd
order queen mobility map
|
|
3rd
order queen mobility map
|
┌───┬───┬───┬───┬───┬───┬───┬───┐
│
│ │ │ │ │ │ │
│
├───┼───┼───┼───┼───┼───┼───┼───┤
│
. │ . │ │ . │ │ . │ . │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
. │ . │ . │ │ . │ . │ . │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
. │ . │ . │ . │ . │ . │ . │ o │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
o │ . │ . │ o │ . │ . │ o │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
. │ o │ . │ o │ . │ o │ . │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
. │ . │ ▒ │ ▒ │ ▒ │ . │ . │ . │
├───┼───┼───┼───┼───┼───┼───┼───┤
│
│ . │ ▒ │ Q │ ▒ │ . │ │ │
└───┴───┴───┴───┴───┴───┴───┴───┘
influence
diagram, w queen on d1
|
1st
order rook mobility map
|
|
2nd
order rook mobility map
|
|
3rd
order rook mobility map ┌───┬───┬───┬───┬───┬───┬───┬───┐ │
│ │ │ │ │ │ │
│ ├───┼───┼───┼───┼───┼───┼───┼───┤ │
│ │ │ │ │ │ │
│ ├───┼───┼───┼───┼───┼───┼───┼───┤ │
│ │ │ │ │ │ │
│ ├───┼───┼───┼───┼───┼───┼───┼───┤ │
│ │ │ │ │ │ │
. │ ├───┼───┼───┼───┼───┼───┼───┼───┤ │
│ │ │ │ │ │ │
o │ ├───┼───┼───┼───┼───┼───┼───┼───┤ │
. │ . │ . │ . │ . │ . │ . │ o │ ├───┼───┼───┼───┼───┼───┼───┼───┤ │
│ │ │ │ │ │ . │ ▒
│ ├───┼───┼───┼───┼───┼───┼───┼───┤ │
│ │ │ │ │ o │ ▒ │ R │ └───┴───┴───┴───┴───┴───┴───┴───┘ influence
diagram, w rook on h1. |
Now things get a little complicated when we generate these maps because
we must consider capturing enemy pieces or moving friendly pieces out of the way. Our mobility tables must keep track of all
this, in order to accurately determine the piece mobility (positional pressure).
The goal of this exercise is to demonstrate how a machine can be encouraged
to form attacking and defensive plans by *accurately* rewarding the placement of a piece based on its attacking and defensive
potential, in the form of positional pressure.
The generation of these "maps" is one step the machine could take.
The assignment of points based on these maps is another issue, but even a rough guess would play a tolerable game of chess.