Chris Martin
England London

Introduction
Pylos, being a conceptually simple game with, nonetheless, many different play positions and possible game pathways, seems wellsuited to being notated, both for purposes of recording games and also for the purposes of discussing strategy. In this article I will seek to describe the system of notation which I use for describing my own games of Pylos, and give some examples of use. I will assume at all times that the reader is already familiar with the game of Pylos; this article on notation is not the right place to also attempt an indepth description of what is anyway a remarkably easy game to learn. If you are not familiar with Pylos, I recommend reading the reviews here [link] or, better yet, playing the game yourself: it is easy to learn, but very rewarding.
Terminology I shall refer to the colours as white and black, and the players as the white player and the black player (or, white and black), where the white player plays first. I will use masculine pronouns for the white player and feminine pronouns for the black player. [No racial or sexual connotations are implied  these conventions are adopted solely for the sake of clarity.]
I shall refer to levels, columns and rows. By levels, I mean the vertical dimension of the board; by columns, I mean the breadth of the board (from an orthogonal perspective); by rows, I mean the depth of the board. Thus, the columns are all equidistant from the player; the rows are progressively further from him; and the levels are progressively built up. I will be inconsistent in my use of the term 'bottom', which can variously mean lowest level or nearest row. I will use the term 'slot' to mean a space in which a ball can be placed within the game.
A Tridimensional Algebraic Notation Just as in Chess, where there are standard notations used for the purposes of notating games and also discussing strategy, I will be using a notation that I have created for Pylos. I will be using an algebraic notation based upon a board placed orthogonally to two players sat opposite to each other. (Although there is some merit, both strategically and diagnostically, to considering the board diagonally, it is easier to notate orthogonally.)
I have chosen to use a tridimensional notation: as such, the notation gives a numberletternumber, indicating levelcolumnrow. As in algebraic chess notation, this is done from the perspective of the white player. e.g. 1a1 for the bottom level, the leftmost column (as facing white) and the bottommost row (as facing white); or 1c4 for the bottom level, the centreright column (as facing white) and the topmost row (as facing white); or 2b2 for the central spot on the second level; or 4a1 for the top of the pyramid. The exact details of this scheme are entirely arbitrary, even within the premisses of tridimensionality and algebraicity: it would have been perfectly possible to refer to the levels with a Greek, Hebrew or other nonLatin letter, in order to differentiate it better from the other dimensions, but I felt that this would be too arduous to type, and that capital letters would be too confusing, especially since I wanted to retain letters for columns from chess notation. Also, one could have referred to the higher levels according to different standards for their rows and columns: I chose to take the bottomleft of each level as a new starting point, but it would not be irrational to refer to each level's columns and rows in absolute terms, that is, since the four slots of level 3 are directly above the central slots of the bottom level (level 1), it makes a degree of sense to refer to them as b2, b3, c2 and c3 respectively; however, this becomes problematic when combining even and odd levels; though the problems are not insuperable, I elected to adopt the simpler convention of restarting the notation of each level with a1 for its bottomleft slot.
It is also possible to use otherdimensional, or nonalgebraic notations. I toyed with the idea of a bidimensional notation, treating the levels as adjuncts of each other, and with the idea of a unidimensional notation, especially useful for inputting data into spreadsheet analyses of game positions. But ultimately I decided that a tridimensional notation would do the most justice to the game, and be the most easily comprehensible "ataglance". Similarly, one could use nonalgebraic notations. This would use the same kind of descriptions as I have used in describing the algebraic positions; for example, 1a1 in my system becomes something like BFL1, for Bottom level, Far Left column, 1st row  and these notations would be written perspectively. However, I reject this for the same reasons that it is rejected in modern chess notation, that algebraic notation is less ambiguous and more compact, while not being overly more difficult to read.
Here is a "map" of what all of the slots are called: Level 1 Level 2 Level 3 Level 4 Black Black Black Black 1a4 1b4 1c4 1d4 1a3 1b3 1c3 1d3 2a3 2b3 2c3 3a2 3b2 1a2 1b2 1c2 1d2 2a2 2b2 2c2 3a1 3b1 4a1 1a1 1b1 1c1 1d1 2a1 2b1 2c1 White White White White
Other Notation The remaining notation in Pylos is relatively simple. Unlike in chess, there are no different types of piece, and unlike chess there are only two "special" moves (or possibly three/four with the alignment rules, though really these simply give extra possibilities for using the same special move).
For the first special move, that of raising a ball to a higher level, I will use a standard method of describing from where the ball came, then adding a dash, and then describing, as normal, to where the ball went. Thus, 1a12b2 describes a ball being raised from the bottomleft slot on the bottom level to the central slot of the second level; 2b2 by itself describes a ball being placed into the central slot of the second level from the player's ball pool.
For the second special "move", that of removing two balls after the formation of a square of one's own colour (or of an orthogonal line on one of the bottom two levels in the advanced game), I will describe the removed balls in brackets after the placed ball. Thus, a move could be 1d3(1d3,1d2), to describe a ball which was placed, resulting in a square and the removal of the ball and its partner on 1d2. It is possible to distinguish between the three different types of reason for the second special move if necessary, though this information would be purely descriptive, and not necessary for a recreation of the game (much like the + notation for check in chess, which is superfluous but interesting). My concern in picking appropriate symbols is ease of typing on a standard QWERTY keyboard; though it would be possible to pick e.g. □, ║ and ═ from the Unicode set to represent squares, fourlong lines on level 1 and threelong lines on level 2 respectively, these symbols are not immediate to type. Therefore, I would suggest O, + and Y respectively, which hopefully bring clearly to mind the ideas of a square (or ring) of balls, of four balls and of three balls. the above described move would then become 1d3(1d3,1d2)O; I place the O at the end on the principle that necessary notation should come before merely interesting notation.
Improving the Notation This notation is still a work in progress; though I think it is quite sound, there are bound to be improvements that can be made. For example, I have considered dropping the initial 1 from moves on the bottom level: this is because the majority of moves in the game are made on the bottom level, and thus dropping that prefatory 1 could make our notation approximately 17% more concise across the course of a game, while, one may feel, not losing much by way of readability  much as the P is usefully abbreviated out of descriptions of pawn moves in chess. Alternately, it may be decided that my special move notation is misleading; for example, it might make more sense to describe the raising a ball move with the destination first, in line with the other moves.
Others may consider that the levelcolumnrow protocols which I adopted are misleading or insufficient; others yet may have entirely new, original ideas. However, until such time as a sound case is made for a superior notation system for Pylos, I will remain content with the system described in this article.
A Sample Game (standard rules) White Black 1b2 1c3 1c4 1c2 1d2 1b3 1d22b2 1d2 1d3 1c1 2c2 1d1(1c1,1d1) 1c1 1a4 1b1 1a42b1 1a2 1a1 1c42a1 1b4 1c4 2b3 1a4 1d4 1a42c3 1a4 1a3 2b13b2 1c12a2 1c1 1d1 1a42b1 1a4 2c1 2a23b1 2a3 2a2 3a1 3a2 4a1 01



I like the notation, seems precise, compact and easy to use.
I just replayed the sample game. If I read correctly, white would make a line with 2a2 on the second to las round, and should thus remove balls. Am I reading it wrong?

Pablo Schulman
Brazil Belo Horizonte Minas Gerais

it was a notation of a standard game (which doesn't allow removing balls after the formation of a line)

Chris Martin
England London

Yes, a standard game. Nice to see someone reading this all these years on!

Bennett Gardiner
Australia Brisbane Queensland

Hey Chris, I wanted to let you know I'm writing a short guide to Pylos (I've found a few interesting things to say about it) and will be utilising your notation. Just wanted to say thanks for this article.

Chris Martin
England London

dispatch134711 wrote: Hey Chris, I wanted to let you know I'm writing a short guide to Pylos (I've found a few interesting things to say about it) and will be utilising your notation. Just wanted to say thanks for this article. Thanks for letting me know! Post a link to the guide here if poss?



May I suggest: 116, L1L9, U1U4, 0 or "wins"

Michael Kandrac
United States Grand Prairie TX

Chris, thank you for posting this annotation system for Pylos. I enjoyed working through the moves of the sample game and here are some of my observations:
1. Of course, unlike Chess and many other abstracts, the first move is a decided disadvantage in Pylos. White must wrest black's inherent advantage or an inevitable win for black is the outcome.
2. In the sample game, white has the advantage for only two moves in a game that lasts 20 moves (turns 4 and 5). This was gained by a ball on the board played to the 2nd level.
3. On turn six, white blunders by not playing 6. 1c42c2. By white playing a reserve ball, rather than the ball at 1c4 to the 2nd level, the result of black completing a black square is a 1011 differential. This is a 2 "tempo" lead for black, and although white draws even with black on five of the remaining 14 turns (black retains the 1 tempo advantage), black will never relinquish its lead.
Your sample game helped to clarify the rules, and I hope to get around to making a DIY Pylos using golf balls, only because I have a bunch of golf balls that need a purpose other than taking up space in my shed.
Gg

Chris Martin
England London

Thanks Michael. To be clear, the sample game was not a game between experts. It was merely included as an example of how to use the notation. If a Pylos tournament is ever held, I would love to read the results of that in the notation!

Michael Kandrac
United States Grand Prairie TX

chrisjwmartin wrote: Thanks Michael. To be clear, the sample game was not a game between experts. It was merely included as an example of how to use the notation. If a Pylos tournament is ever held, I would love to read the results of that in the notation!
It would be interesting to see a record of two accomplished players! Thanks for the reply.
Gg


