Move notation. Dots and Boxes
5 replies. Last post: 20040721
Reply to this topic Return to forum
Knox (Computer) at 20040621
To be able to discuss games and alternative lines of play,
we need a notation for referring to specific moves. Since
a move is a line connecting two dots, an obvious method is to label the dots and specify a move by listing the two dots
being connected.
My first idea was to specify a dot by (r,c) where r is the
row number and c is the column number. In this notation,
the move connecting the dot in the lowerleft corner with
the dot to its right would be specified as (6,1)–(6,2) or
more simply as 61–62 [I like this better]. The one
drawback to this notation is that it is easy to confuse
the row with the column and hence, the notation is prone
to error. To remedy this problem we can instead use
letters for the columns and numbers for the rows
(or viceversa) similar to describing a chess board
with “algebraic notation.”
I propose we adopt following move notation. Label the
columns from left to right as a,b,c,d,e,f and label the
rows from *bottom* to top as 1,2,3,4,5,6. In this proposed notation, we would represent the move that connects the lowerleft corner with the dot to its right as a1b1. If
a turn consists of multiple moves because we have completed some boxes, then we can simply separate the moves by commas.
We would record the start of a hypothetical dots and boxes
game in the following manner.
1. b5c5 2. c4c5
3. b4c4 4. b4b5, d3e3
5. f1f2
The board would look then like the following.
6 + + + + + +
5 + +–+ + + +
\Bl\
4 + +–+ + + +
3 + + + +–+ +
2 + + + + + +
\
1 + + + + + +
a b c d e f
This notation is consistent with the most commonly used
chess notation which is called algebraic notation.
People who are used to this notation would hate it if
you used letters for the rows and numbers for the columns,
or if you reversed the row numbering so that 1 was at the
top and 6 at the bottom (trust me on this).
I believe the proposed notation is simple, clear, and unambiguous. Comments? Alternative proposals?

Knox (Computer) at 20040621
The carefully constructed spacing from my ascii dotsandboxes diagram got removed. Is there any way to prevent this?
Let me try again with html tags and see if that works
6 + + + + + +5 + ++ + + + Bl4 + ++ + + +3 + + + ++ +2 + + + + + + 1 + + + + + + a b c d e f

Nick Wedd at 20040711
A Google search for '“dots and boxes” notation' finds several other methods:
http://homepages.cae.wisc.edu/~dwilson/boxes/explain.shtml
http://www.mathpuzzle.com/dotsandboxes.htm
http://216.239.59.104/search?q=cache:axwyz50beZgJ:www.msri.org/publications/books/Book29/files/westboxes.pdf+%22dots+and+boxes%22+notation&hl=en

Knox (Computer) at 20040712
But which notation do you like best?
In regards to the notations described in the three links
provided:
The third one requires a diagram of the board which misses
the point. The purpose of having a notation is to provide
a good *textual* method of describing the moves without
having to draw a complicated ascii diagram for each
variation.
I really don't like the first one. When recording moves,
you could easily reverse the row and column labels.
Perhaps more significantly, it seems next to impossible
to visualize a move in this notation without having a labeled board in front of you.
The second one seems like a good notation. This one
gives names to the squares instead of to the dots as
mine does. Both notations allow easy visualization of
the move on the board (once you get used to the notations).
I won't vote between mine and the one on the second link
because of any bias I might have towards my notation.
This issue is probably more significant for me because I
can freely discuss variations as Knox is playing the game
 the discussion can't possibly effect Knox's moves. If
it could, Knox would have beaten Taral 1510 whereas it
is now about to win 1411.