Well, it has taken me some three years to work this out, but here is my idea of a notation system to record games of Kensington.

You will need to sketch out a pattern of triangles and make one pattern for the starting player and another (slightly different) for the second player. Draw a pattern of triangles, photocopy it a few times, then write on the photocopies.

The principle idea is this; if you look at a Kensington board, all the points are the corners of squares and triangles. If you consider the squares only, you will notice that they are joined at the corners; that is, many point can be shared by two squares.

However, if you consider only the triangles, they are all separate from each other; that is, the corners of triangles are unique to each triangle!

So, to identify any point on the board, you need only identify its triangle, and its place on the triangle, 1, 2 or 3.

BINGO!

I just drew up a system geared for the starting player (Red or Blue, it doesn't matter) and turned it round for the second player.

Draw up one pattern as above for the starting player. Now turn it round, so A B are at the bottom and W X are at the top. This is the view the second player gets. Draw up a new pattern (I suggest you do a blank pattern of triangles and photocopy it a few times) and label it from the second players view, e.g., the first triangle is X, going X2, X3, then X1 (reading clockwise); W2, W3 then W1. So on to the last triangle, A, numbered A3, A1 then A2 (remember to follow the numbers on the starter's sheet).

Now you have the patterns for the starter and second player. Have a look at the sample game to see a good way of using the notation system. Here's a example (right);

As each player takes its turn, write the point number (from the diagram) under the Player's column. When you start moving your pieces at move 16, write the point (in your Player column) moved from and to, e.g.; G1-G2 means a piece moved from G1 to G2. If in either phase, you make a Triangle, Square or House shape (the triangle/square combination), add a T, or S or H accordingly. Then, in the mover column, write the point or points (referring to your opponent's pieces) moved from and to, e.g.; R2-M1 means the piece on R2 was picked up and moved to M1. I think it might be worth going back up the opponent's list of moves (on your sheet) and cross out the R2, to show it was moved.

If this works, you should be able to look up the last fifteen moves from the bottom of the lists (since no pieces are removed from the board) and work out where all the pieces are. Your pieces will be on your Player List and your opponent's Mover List, and vice versa.