The site has moved to a new server, and there are now some issues to fix. Please report anything needing fixing with a comment to the homepage.

The Chess Variant Pages

DISCLAIMER: The ideas expressed on this page are the creative work of its author, Daniil Frolov, and they might not be shared by others responsible for this website.

The most generalized piece classification ever

We usually think that there are five groups of pieces: leapers (standart knight, shatranj elephant, Tamerlane chess camel and others), riders (standart rook and bishop, gryphon, boyscout, etc), sliders (Chinese and Korean knight, Korean elephant, non-jumping Jetan pieces...), hoppers (Chinese cannon, grasshopper) and another one, wich usually is not mentoided but implied, combined pieces (for example, Tamerlane chess giraffe starts it's move as slider, then may continue moving as rider). But actually these five groups are just groups, wich imitates moves of certain real live things, pleasant for human eyes. For example, how you'll classify piece, wich jumps as camel, but needs from 3 to 6 pieces knight's leap away from starting square and 2-5 pieces zebra's leap away from destination square? And it's still quite simple, even more complicated pieces are possible. Wich classification will include ALL possible pieces? Here is it (might have some mistakes)! p.s.: this classification have no any point, i have written it just for fun.


First, let's classify movement of pieces. There are only two (!) groups: 1. Leapers. They moves to certain positions away from oneselfs and nothing can stop this move. 2. All other pieces. Thier movement depends on certain conditions*. If piece is compound of leaper and something from second group, it belongs to second group, as not all it's moves are free. Actually, there is a third group, wich have only one piece: dummy, wich don't move at all. Note that some moves may be used only for capturing or non-capturing moves, and piece with these restrictions can still be considered as leaper. Second, clssification of capturing methods. There is also not many of them: only three (or four). 1. Replacement. 2. Coordinating: realization of certain conditions* when certain piece is captured (most of popular capturing methods are coordinating: checkers overtaking, custodianship, withdrawiing, rithmomachia captures, en-passant or even shoting - all belongs to this group). In second-most generalized piece classification this group can be divided to these subgroups: A. To capture, player must move piece. B. When capture happens, position is not changed, expect for removing captured piece from board (shoting). 3. Coordinating, combined with replacement: piece captures by replacement, but it must be made when certain conditions* apply. For example, dev from Giant Chess can be seen as four simultaneously moving dababas (see below), one of them can be captured only by replacement, combined with coordinating, other three only by coordinating (on same turn as first and two others). ...Can BE CAPTURED... Does it mean that besides classification of capturing methods, wich piece uses, there also must be classification of capturing methods, with wich piece can be captured? No. You can just say that all (maybe, with expections) pieces of this game captures that piece only with this method. Now, three more ways of pieces division. Simultaneously moving pieces. 1. Pieces, wich can't move simultaneously with other pieces. 2. Pieces, wich can move simultaneously with other pieces. All pieces, occupying more than one square (dev above or wall from Ganymede Chess) also belongs to this group. So, where are pieces, occupying less than one square? It's simple: crowd of pieces on same square can be considered as different kind of piece. Moving one piece of crowd is like making new piece. By the way: Making new pieces. 1. Pieces, wich can't add new piece on board. 2. Pieces, wich can add new piece on board. There several games with these pieces. Growing large pieces also belongs to this group. Actually, it would be enough to classify any piece, but it makes some pieces very compliated and sometimes pieces, wich feels very leaper-like are not leapers (see below)! So, here is the last devision: Changing other pieces. 1. Pieces, wich can't change other pieces (or oneselfs). 2. Pieces, wich can change other pieces (or oneselfs). For example, standart chess pawn changes itself to another standart piece permanently on last rank. Immobilizer changes adjecent hostile pieces to dummies as long as they are adjecent. Bodyguard from hiashatar changes queens, rooks and bishops into thier limited range versions, depending on thier positions in relation to it. * - i marked words "certain conditions" with this symbol. So, what are these certain conditions? 1. Positions of pieces in relation of board and each other. 2. Previously made move and thier directions (for example, can pawn capture en-passant other pawn or can't, depends on this condition). 3. Total number of moves or turns, wich was made before by all pieces, by one piece or by group of pieces. For example, when existentialist and bobber (from EXISTENTIALIST CHESS) moves, they changes thier own way of movement. 4. Statements, made by players (they may be known for other players or only for referee). Statements of disinterested person belongs to 6th category. 5. Random (dice). 6. Conditions out of board (for example, two games of Bughouse don't depent on each other. Or, say, if player coughs, he loses one of his pawns :). Actually, "random" also belongs to this category (hm... Interesting variant - bughouse, where one of games is played randomly and checkmate or stalemate is impossible in random game). Hope, i did'nt miss anything. But wait: for example, there is game, where only pieces, rolled with dice, can be moved. And knight also can be moved only when, say, 2 is rolled. Thus, knight's moves depends on certain conditions. Does it mean that in this game knight is not leaper?! And many other examples are possible! There is one solution: it counts as game rules (same as board shape and opening position). As you see, there is even no strict border between game rules pieces movement. Originally i wanted to add here classification of royality: royal, non-royal and royal or not, depending on certain conditions. But there are problems: in shatranj alternative winning condition is to capture all pieces, expect king. So, are all pieces royal here?! Maybe, piece can be concidered as royal only when capturing it will win the game (thus, in shatranj last non-king piece is royal unless king can capture opponent's last piece on next turn)? No. Even more complicated royal piece rules are possible. For example: pawn promotes only to lost pieces. When king is captured, it's owner will not lose if pawn will promote to king within next 20 turns. So, where is royal piece? There is only one way: royality is rule of game, not piece's property.


I hope, this articlkle did'nt spoiled romanticism of inventing new pieces :).

This 'user submitted' page is a collaboration between the posting user and the Chess Variant Pages. Registered contributors to the Chess Variant Pages have the ability to post their own works, subject to review and editing by the Chess Variant Pages Editorial Staff.

By Daniil Frolov.
Web page created: 2010-09-24. Web page last updated: 2010-09-24