Friday, 2 March 2018

Oblong chess

Oblong chess is a chess variety that was also played during medieval times, however not in Europe, but in the Arab world. The game is known under the names Shatranj al-Mustatîla, al-Tawîla (long) or al-Mamdûda (lengthened). This chess variant is mentioned in 947 AD by the Arab historian Al-Masudi his book Muruj adh-dhahab (the meadows of gold). However, the rules therein say that they were derived from another source dating from the 9th century. 

 The most used set-up of  oblong chess: elephant - king - consellor - elephant; 2x knight; 2x rook; 4x pawn; 4x pawn.

Playing the game

Oblong chess is played on a board of 4 by 16 squares, with help of dice, making it the earliest chess game using dice (also Alphonso the Wise (1282 AD) mentions that dice can be used to play chess). Several setups of the game exist, the one shown below the one most used. Oblong chess uses the same rules as Shatranj, i.e. that of medieval chess. A win is by checkmate or by bare king. The game can also be played without dice.

Using dice with oblong chess

A player must move the chess piece that is shown by the dice roll. A roll of 6 moves the King, a 5 the Vizier (queen/counsellor), a 4 the Elephant, a 3 the Knight, a 2 the Rook (chariot) and 1 the Pawn. If the player is unable to move the designated chess piece his turn is lost. A player can also choose not to move a chess piece (after the dice roll), thereby also losing his turn. When the king is checked, the player may only move the king and no other pieces. Hence, he must roll a 6, otherwise (i.e. on a roll of 1-5) his turn is lost and he remains checked.

Six variants of the set-up. The difference between 'a' and 'd' is that in the latter the king faces the opponents counsellor instead of the king.

Making the board

The game board was made from a leftover piece of poplar (from the Daldosa game described in the previous post) in the same manner as the medieval chess board, but without accenting the lines with black. The lines of the squares were carved, the board edges made from oak, and finally the compete board finished with linseed oil.

  • H.J.R. Murray, 1913. A history of chess.  (2012 Reprint Skyhorse Publishing).
  • J.L. Cazaux and R. Knowlton, 2017. A world of chess: its development and variations through centuries and civilizations. McFarland and Company, Jefferson, NC, USA. ISBN 978-0-7864-9427-9.

Monday, 26 February 2018

A medieval 'ship' game

14th century games-board made from spruce. The other side holds a nine men morris board. Novgorod, Russia. Image from the book Wood use in medieval Novgorod by M. Brisbane. 

The games board above dating from the 14th century has already been shown in a blogpost several years ago, but at that time it was unclear (for me) which game this should represent. It has two rows of 14 columns, or 15 lines crossed by 3. Our visit to the re-enactors of Aisling 1167 with their huge collection and knowledge of medieval games, revealed what this game probably was: a tâb-like game, of which some ship-formed variants still are played in parts of Norway, Denmark and Finland (as Daldøs, Daldøsa and Sáhkku, respectively). The tâb game has its origin in the near east and Africa. It was already played in Egypt during the thirteenth century BC. In medieval times (1267) the game was mentioned in his poem Dıwan by the Muhammad Ibn Dâniyâl, a Persian poet and physicist. Tâb is a game of which the purpose is to eliminate the (pieces of the) opponent, while they are moved along a linear track depending on the roll of dice. The number of lines or points along the track for varies; most commonly they are 12, but examples as large as 17 exist as well.

A 13th century century games-board from Novgorod, Russia.

The barrel lid from the Mary Rose has two game boards, a nine-men morris game and a daldøsa board. Photo Mary Rose Trust. The measurements of the daldøsa board are from Daniel Diehl and Mark Donnelly in their book 'Medieval furniture. Plans and instructions for historical reproductions.'  Below: the board found on a 15th century ship from Newport

In medieval Europe several game boards have been found that could be boards for Daldøsa or tâb-like games. Two game boards in ship-form were found in medieval Novgorod, Russia, dating from the 13th and 14th century, which have 17 and 15 lines inscribed, respectively. Also a presumed board was found in Newport, Wales, in an excavated 15th century ship (1446-1468) with 11 lines. More game boards were found inscribed in stone at some British church sites, such as Lincoln Cathedral. A slightly later dated board was found on a barrel lid on board the Mary Rose, a Tudor warship that sank in 1545. This board has 13 lines, with an extra point along the central line. Finally, a mysterious games-board representing such a game was drawn in a manuscript dating from the 13th century. The manuscript belonged to a monastery in Cerne, UK but is now in the Trinity College Library in Cambridge, UK  (MS O 2.45; folio 2v). Likely these European tâb games were imported by Viking merchants or mercenaries from the Middle East and further spread during their conquests to the England. Interestingly, also the Libre de los Juegos of Alphonso X the Wise (1283) does contain an elimination game along a line. In the game of astronomical tables, seven players move their pieces along a circular line by dice throw, until one is left.

A (not very visible) daldøsa game board inscribed in stone at Lincoln cathedral, UK.
Photo by Mark Hall in Histoire et Images Medievales 28.

Folio 2v and 3r of manuscript MS O 2.45 (after 1248 AD) containing two chess boards, an alquerque, a nine-mens-morris and a daldøsa game board. The first moves of the game are already played. A part of folio 2v has been cut of.  Wren digital library, Trinity College Library, Cambridge, UK.

Making the games board

The games board is made from poplar, like most of our games boards with an walnut edge. The board was sawn in a ships form using a band saw. On the board twelve lines crossed by three were carved, with an additional line to a point in the prow of the ship, like the May Rose board. I also decided to drill holes at each point, like in the modern daldøsa games. Such a system of holes holding the game pieces is also known for early medieval boardgames, such as the Ballinderry and Waterford tafl game boards found in Ireland. Holes were drilled using a small Forstner drill and smoothed with a cross-hole countersink cutter.

The board is carved and the holes drilled for the game pieces, of which some unfinished ones are shown.

As the board resembles a ship, it became necessary to bend the decorative walnut edge. In order to do this, a small steam cage was built consisting of pvc pipes connected to a steaming kettle (the outer glue pot) on an electrical heater. Walnut proved to be hard to bend, and remained at least an hour in the steam cage. I had to directly glue the warm, wet and malleable walnut strip to the ship and clamp it with help of a counter-mould.Afterwards the board was sanded and oiled with linseed oil (including all the holes).

The steam-pipe consisted of several PVC elements connected to a wooden lid closing the glue pot. Having the steam-pipe at an angle ensured that the condensed steam flowed back into the glue pot.

Left: The gameboard and the walnut edge pressed into the counter-mould. Right: Some extra wedges were used for extra pressure in the counter-mould.

The completed game board.

Making the game pieces

The different steps of making the game pieces. 

The game pieces were made from hornbeam and walnut. First, long strips of wood were sawn which were cut into small pieces. Then a square pin was sawn, which was rounded using a file. A scrap piece of poplar with a 6 mm hole was used to test the roundness and fitting of the pin. Then the top was rounded using a stationary belt sander, and the head was formed using a triangular file. After this, the belt sander was again applied to create a man shaped game piece. Finally, all edges were rounded with files and smoothed with sandpaper, after which a coat of linseed oil was applied.

Left: The game piece was put into a vise when the pin was sawn and filed. 
Right: Testing the pin with a scrap piece of poplar with a 6 mm hole.

Left: the oiled game pieces. A few extra pieces were made in case some go missing or get broken.  
Right: The game pieces in the board.

An early medieval pin style game piece: the 'kingpin' of a hnefatafl game, 9th century. Image from Waterford Museum, Waterford, Ireland. The pin is large compared to the gameboard that was found as well at Waterford, and does not seem to fit it. Perhaps, it had some other unknown use.

Making the dice

Daldøsa and daldøs use oblong dice with values of 2-4, and the fourth side carved with an X or an A on it (which also represents the number 1). The sum of two opposite sides is different, i.e. 4 or 6, not 5. The two small edges are slightly pyramidal, so that they cannot stand on these sides and much fall to one of the oblong sides. The oblong dice are modelled after Scandinavian archaeological finds. They were made from scrap pieces of antler that were sawn into square long strips. These were equally flattened with a stationary belt sander; also the pyramidal tips were made by sanding, but with a circular sanding disc with a 90 degrees platform. The pips were drilled using a home-made pip-drill, while the X was sawn and filed. The pips and X were inked and the complete dice was smoothed using 400 grid sandpaper.
Different sides of the two oblong dice.

Dice on the daldøsa board.

Historic oblong dice from Vimose, Denmark, around 400-500 AD. 
Note that the number of pips on the sides are not 1 to 4.

Playing the game

 The game board and its pieces.

The gaming pieces are placed with the flat sides facing in the direction of movement. Two dice are used in turn by the players. ‘X’ stands for the number one, and is also called ‘dal’. If the result is two ‘dal’s’, the player has an extra throw. To start the game the players throw the two dice in turn, and the higher throw decides who starts.

First, to be able to move a man, it has to be activated. This is done by a 'dal' throw, after which the man is turned (so one can see it is activated) and moved one step forward.  All men must be activated in this way (i.e. turned and moved one hole), one by one from one end to the other before they can enter the game. As soon as a man is activated, it can start to move. All throws of the dice are invalid until a player has thrown his first X and thus activated a man, and an activated man can move as far into the middle row as the results of the two dice allow. If the first player’s initial throw does not result in at least one ‘X’, the dice are passed on to the opponent. For the next throws, one can use the X either to move an activated man one step, or activate an inactivated game piece. A throw can be used to move either one man or two. When only one man is moved, the results on the dice must be used separately (as in backgammon).

The pieces move in the direction of the arrow. The activated men first move to the middle row, after which they enter the 'enemy line' from behind. Then they continue until they again reach the middle row. They never return to their starting row. (white = movement white pieces, black = movement black pieces).

When a man has travelled to the end of the middle row it moves into the opponent’s home row and through it. Then, it re-enters the middle row and from then on continues to move in the middle row and the opponent’s home row, never to return to his own home row. A man goes on moving in these two rows until it is killed or the game is over. A man moving through through the opponent’s home row can ‘kill’ as many men, activated or not, as the dice allow for and cannot be ‘killed’, unless it is placed in the hole right in front of the man whose turn it is to be activated, or in any hole in front of any already activated man.

A player is not allowed to jump over his own men, but he can jump over the men belonging to the opponent. If a throw enables a player to land his man in a hole occupied by one of the opponent’s men, the latter is dead and taken off the board and cannot be brought back into play. The aim of the game is to kill all the opponents game pieces.

A strategy can be to combine his moves such that he always stands behind any activated men belonging to the opponent. It is an advantage to have an activated man ready at the beginning of the home row in order to chase the opponent’s men as soon as they move into the middle row. Also, quickly having an active man behind the enemy line slaughtering their men can be a strategy. The end game, with only a few pieces left, can take long as the the opposing game pieces have to land on top of each other.

Historic and alternative play


Above is the 'daldøsa' game from the 12th century manuscript. The game has already started here. Black has started. According to our daldøsa rules, black must have rolled an 'X' (activate and move one place forward) and a '2' (two extra pieces forward). Green has the next move. As it has moved, it also must have rolled an 'X'. The move forward by the X, places the green piece at the spot it is drawn. Thus, it must still have one dice move left.

All medieval daldøsa game boards have lines connecting points of both outer rows with points of the inner row. One can speculate whether the lines might have a function in the game. For instance, an active man in an outer row (e.g. both rows or only the opponents row) could directly move into the middle row, creating more challenging situations. Or the game board could shorten itself: if the three lines, as from the stern, all are empty, these spaces may not be used any more.  Thus the travelling circle becomes smaller. Both 'options' also explain why the middle row must be one point longer (i.e. 13 instead of 12), as otherwise minimal circular movement is not possible. Both 'options' also result in a faster and less boring end game.

Sources used:

  • Peter Michaelsen (2012) Un jeu medieval arabe en Scandinavie. Histoire et Images Medievales Thematique 28, page 25-29.
  • Alf Næsheim (2001) Daldøsa, an old dice game with an obscure origin. Board Game Studies 4, page 9-14.
  • Erik Østergaard and Anne Gaston (2001). Daldøs – the Rules. Board Game Studies 4, page 15-18.
  • Peter Michaelsen (2001). Daldøs, an almost forgotten dice board game. Board Game Studies 4, page 19-32
  • Alan Borvo (2001). Sahkku, the “Devil’s game”. Board Game Studies 4, page 33-52.
  • Thierry Depaulis (2001). Jeux de parcours du monde arabo-musulman (Afrique du Nord et Proche-Orient). Board Game Studies 4, page 53-76.
  • Thierry Depaulis (2001) An Arab game in the North Pole? Board Game Studies 4, page 77-84.
  • Wikipedia. Daldos.

Tuesday, 30 January 2018

More on sevensided dice

Several years ago I made some seven-sided dice for some medieval board games mentioned in the Libre de los Juegos (1283) by Alphonso X 'The Wise'. These dice were not entirely fair, as they showed more often the 6 and 7 on the pentagon sides. As I now had some spare antler (from the reconstructed Noyon chess pieces), it was time to create some better dice.

The old seven-sided dice.

First, a large pentagon column was made having equal sides at a 72 degree angle. Then, the idea was that the pentagon and the rectangles should have the same surface area. If you have the length of one side of the pentagon, this is easily calculated using the next formula. Or even easier, you can use a website to do this for you. From the surface area you then can calculate the length of the rectangle, as a of the pentagon and a of the rectangle are the same, as well as the surface A.

The pentagon and the formula for the surface area (A).

 The rectangle and the formula for the surface area (A).

Funny enough this did not yield a fair dice. On the contrary, here, most of times the rectangles showed. If you reconsider this, it is logical, as the edges of the dice are not the same: 72 versus 90 degrees. What followed next was a careful shortening of the rectangles with a belt sander, and checking whether the dice throw (>200 throws) became fair. In the end, after several removal steps, I now have reasonably fair seven-sided dice.

The new seven-sided dice.

Monday, 18 December 2017

Medieval chess pieces from antler: individual chess pieces

This post will show the individual chess pieces of the medieval Noyon-like chess pieces made from antler. Some more information on the original pieces and the construction of this set can be found in the previous blogpost.

The pawns

The two surviving pawns from the original Noyon chess set. The number of ribs is different due to the size of the pawn (9 vs 10). Both pawns belong to the same side. They are 30-31 mm high and have a diameter of 16-17 mm. Adjusted scan from the 'Revue archeologique de Picardie'.

Two pawns from each side. Also our (left side) pawns do have different numbers of ribs because of their diameters. The right side had to be newly designed by us and consists of triangles.

The rooks

The two original rooks from the Noyon chess set. Both are from the same side. Height 54/55 mm, width 37/34 mm and depth 20/21 mm, respectively. Adjusted scan from the 'Revue archeologique de Picardie'.

The four sides of a rook from the 'ribbed' side. This one was made from roe antler.

The four sides of a rook from the 'triangular' side. This one was made from deer antler.

All four rooks next to each other.

The knights

The two knights from opposing sides, as can be seen by the different front and backside. They measure 58/57 mm height, 36/30 mm width, and 32/27 mm depth, respectively.  Adjusted scan from the 'Revue archeologique de Picardie'.

All four sides from a knight from the 'ribbed' side. Our knights were adjusted to either the ribbed or triangular pattern, which is slightly dissimilar to the original patterns. The head of the knight has two lines.

All four sides from a knight of the 'triangular' side.

All four knights together. All knights were made from deer antler.

The elephants/bishops

Only one elephant survived from the Noyon chess set. The back looks like the 'ribbed' side. Height 56 mm, width 31 mm and depth 29 mm. Adjusted scan from the 'Revue archeologique de Picardie'.

All four side from an elephant from the 'ribbed' side.

All four sides from an elephant of the 'triangular side'.

All four elephants together. All elephants were made from deer antler.

The queens

Only one queen survived from the Noyon chess set, which has a 'ribbed' backside. Height 57 mm, width 42 mm, depth 34 mm.  Adjusted scan from the 'Revue archeologique de Picardie'.
The four sides of the 'ribbed' queen. Our queens are different from the Noyon one in such that the 'ribbed' variant also has a ribbed front.

The four sides of the 'triangular' queen.

The kings

Both original kings from the Noyon chess set. The differences between the two kings are the wider spacing on the back and the slashes on the triangles of King 2, while King 1 has an extra ring on his crown. Height 78/72 mm, width  46/38 mm and depth 34/37 mm, respectively. Adjusted scans from the 'Revue archeologique de Picardie'.

The four sides of the king from the 'ribbed' side. There is no distinction in crowns of the kings, as I noticed too late the extra ring, when there was not enough space left to add it.

The four sides of the king from the 'triangular' side. 

The kings and queens from the new chess set.

Sources used:

  • J-F. Goret, M. Talon and J-H. Ivinec, 2009. Le jeu d'echecs de Noyon dans sons contexte archeologique et historique. Revue Archeologique de Picardie 2009: 79-119.