Robert S. Brumbaugh, Yale University

"Archaeological Note", American Journal of Archaeology

Volume 79, 1975, Pages 135-137.

Knossos Game Board

The graphic above is a detail from an artist's conception of a game being played in the Queen's chamber in the Palace at Knossos, Crete, about 1500 bc.

This drawing by Rudolph F. Zallinger appeared on pages 146-147 in the book The Epic of Man, Courtrand Carby, editor, Time, Inc. , Book Division, New York, 1961.

The text accompanying the drawing did not offer any information about the origins of this Minoan game or how it was played. Brumbaugh offers a theory about the origins of this game and how it was played.

(Page 135) In his marvelous History of Board Games Other than Chess, H.J.R. Murray has collected a wealth of evidence bearing on ancient Near Eastern game boards and the games played on them.1 But when he comes to the so-called Knossos Game Board, he concludes that it is not a game board at all, whatever may have been its function.2 On the contrary, I will argue that the similarities of the Knossos Board and the earlier Royal Game Board found at Ur are close enough to establish the identity of the Minoan artifact as a board game of the same family, and even to conjecture within fairly close limits how the game must have been played. For this purpose; I think it helps to give abstract structural pictures of the two boards, deliberately excluding 1) non-functional elements of design from the Ur game, and 2) extraordinary differences in scale between the two designs.

Board from Ur Ur Board Side Board from Knossos

Illustration 1

Game Board from the Royal Graves, Ur; figure 8, p. 20 - H.J.R. Murray, A History of Board Games Other than Chess, Oxford 1957. By permission of The Clarendon Press.

Side view of Game Board From Royal Graves, Ur.

Illustration 2

Game Board from Knossos, Crete; figure 72, p. 24 - H.J.R. Murray, A History of Board Games Other than Chess. Oxford 1957. By permission of The Clarendon Press.

The Ur board (Illustration 1) consists of a 3 x 2 upper block of squares, a 3 x 4 lower block, joined by a "bridge" of two single squares. Each square has its own inlaid design, but variations between boards of this type show that only those squares bearing rosettes are constant, so that only these designs are functional for the game. The other squares, with blue eyes and groups of five dots, may have some magical importance, but are not "standard." A diagram is given here:



The starred squares bear rosettes. The other designs are blue eyes or single or multiple fives, except for #19, which has a small game-board as its design. The board was found with pieces - seven for each player - and tetrahedral "dice," four of which give throws from 0 to 4. Murray suggests - surely correctly - that this was a "race game," pieces being entered, sent across the bridge, around the upper track, and off again. On the analogy of other Old World race games, (Page 136) he also suggests 1) that there was a capture move, by landing on the square occupied by an opposing piece, and 2) that the rosettes mark "safe" squares, where a piece is immune to capture. A lucky player, throwing consecutive fours, can make his tour completely via the safe squares. The unique game-board design of square 19 in this particular board suggests that this is the "entering square." The description of squares 7 and 8 as a "bridge" finds support in some Egyptian and Akkadian material3 that Murray does not cite at this point, mentioning a bridge, a river, and a canal in reference to board games of this general family.

Now, let us look at the pattern of the Knossos board (Illustration 2), disregarding differences of size among the elements.

1a1b 2a2b
4* 5*
7 8
9 101112
13 // 14
15 // 16


The starred spaces bear rosettes. The gap between 6* and 7, 8 is like the "river;' but this time the bridge crossing apparently must be supplemented by a jump. Slashes indicate squares used on the Ur board but lacking here in the lower block. Faired spirals ornament the upper corner rectangles, 1a1b, 2a2b.

The larger block seems designed for the same game as the Ur board, but with a different entering rule. Instead of entering on the Ur square 19, and having an option of paths to the bridge, one play. now enters his pieces on Knossos square 15, the other on 16. Thus no corner squares with rosettes, matching Ur 18* and 20*, are needed; there is no chance for capture, and so no point in "safe" squares, until the bridge is reached. Since the Akkadian verse about a game of this same family refers to "jumping across" the ditch, the break in the bridge in the Knossos game falls within the general idea of simulation. We may suppose, further, that in the Ur game a lucky player, throwing a four with a piece on 10*, could go directly to 4* or 6*. (There is precedent for such a shortcut jump move in the game boards for "Hounds and Jackals" race game.) Then the replacing of one bridge square by a GAP presents no special difficulty.

The upper block, however, is more troublesome. Even assuming that the paired spiral designs of squares (Page 137) 1a1b and 2a2b are functional, four of the six squares here are "safe." This surely makes for a dull game: all the "capture" action will concentrate on squares 7 through 12 - with some minor skirmishes possible in the 1a1b and 2a2b upper corners. But another way to make a race-game interesting is to require a specific throw before some moves can be made. (In "Hounds and Jackals" the tracks of the two players are practically independent; the point is simply winning an extended race, not highly complicated by captures.)

I admit that my next suggestion is pure conjecture; but, given the isomorphism established so far, it is relevant to show that there is any possibility of rules of play for the upper block. On the Ur board, the "lucky" or "good" moves proceed by fours from one "safe" square to another. Suppose that on the Knossos board, a four must be thrown to move from one rosette square to another, even if the squares are contiguous. To make this work, without so much delay that the game becomes dull, we must change the presumed rule of the Ur game concerning occupation of a safe square. I assume that in the latter game, only one piece can occupy a rosette-marked square at any time. For the suggested Knossos game, on the other hand, it seems we should allow any number of pieces to be on the safe squares simultaneously, waiting, passing, but advancing only on the "lucky" throws. In that case, the spaces capable of holding several pieces should be larger than those which can be occupied by only one piece. And this is a possible functional explanation of the far greater size of the rosette-circles of the Knossos game, compared to the squares of its lower block.

This is a simple modification of the older game. The rosettes, and jump moves between them on throws of four, have simply been transferred to the upper track. It is not easy to decide what the further details of the upper track course are; but there are many possibilities for circling its course, by way of the two corners, or simply around the rosette diamond.

I have played the Ur game, and find it not too bad. (It seems to have been standard to gamble for a large stake, which of course would add some further interest to the play.) The Knossos board, as a variant game of the same family, would be a simplified and rather less interesting version, because of decreased possibilities of capture. But it would still be a recognizable ancient race-game.


  1. H.J.R. Murray, A History of Board Games Other than Chess, Oxford, 1957, Chap. II. The two boards with which my present note is concerned are a Royal game-board from Ur, and the Knossos Board. I am indebted to my colleague, Professor William Hallo, for bibliographical references to articles on Near Eastern board games. The Ur game is described and illustrated in C.W. Wooley, Ur Excavations, 1934, I, 276-73; II, Plates 95-98. My illustration, below, is from Murray, 20. The board is from a Queen's grave; its date is between 3000 and 2500 B.C. The Knossos Board, about a millennium later, is described and illustrated in Sir Arthur Evans, "The Palace of Minos," sec. 25; BS,4, vii (1900) 77-82. My illustration, below, is from Murray, 24.
  2. Murray gives this opinion (supra n. I) 23.
  3. Here I would add to Murray's citations a translation of a "rhyme" about winning positions in a board game called "Pack of Hounds" (Benno Landsberger, "Eimge Unerkannt Gebliebene odor Verkannte Nomina des Akkadischen," Wiener Zeitschrift fur Kundes des Morgenlandes 56 [1960], one of the articles recommended to me by Professor Hallo). The rhyme ends, unhappily, with the news that "the rules of the game will not be given here"; but from what is given, it is clear that the board had two sections, joined by "a bridge" across a "ditch" or "canal." This could refer to a later copy of the Ur board itself. The same "bridge" and "river" terminology were used in an Egyptian game of somewhat similar design, and a capture rule is suggested by the reference to "taking a piece and throwing it into the river" (Wiedermann, "Das Brettspiel bei den Adten Aegyptern," Actes du roe Congres d'Orientalistes, 1894; cited by Murray but not used by him in his reconstructions). Another reference to a board with a "river" is discussed in C.J. Gadd, "Babylonian Chess" Iraq 8 (1944) 66-72.

Last update November 24, 2010