Traditional Board Games

Royal Game of Ur

Royal Game of Ur publicity photograph
Royal Game of Ur publicity photograph

The Royal Game of Ur is a modern name for an ancient race game known only from archaeological excavations. Two players race around a figure-of-8-shaped board, with seven pieces, according to the throws of three binary lots. It is assumed that the pieces were entered onto the board at one place and exited at another; the shape of the board and its symmetrical markings also raise the possibility that each player started and finished on his own half of the board, the central row only being shared.

History of the Royal Game of Ur

In 1926-7 Sir Leonard Woolley was excavating at the royal tombs in Ur, in modern Iraq, and found among other things the oldest full set of gaming equipment known to exist. This dates from about 2500 BC, comprised a board of twenty squares, two sets of seven pieces, and six dice in the shape of pyramids. The game was afterwards given the name "the Royal Game of Ur".

Other similar sets were found in the same archaeological dig, some in a less well-preserved state. The games differed in materials and quality of workmanship, but all shared the same layout of squares, some decorated with rosettes. Decoration on squares without rosettes varied from board to board, but some had delightful animal scenes engraved on them.

No rules were found at the time of the excavation, so a number of different historians and archaeologists devised their own interpretations, and there are versions of this game for sale. In more recent years, an almost complete set of rules on cuneiform tablets has come to light, dating from the second century BC and telling us everything about the game apart from the direction the pieces travelled along the board.

Rules for the Royal Game of Ur

Rules for the Royal Game of Ur
Complete sets of rules were published by R. C. Bell in his book Board and Table Games from Many Civilizations, and by Irving Finkel in his game for the British Museum shop. The set of rules in this leaflet takes elements from these but introduces some ideas by H. J. R. Murray and David Parlett that make the game more logical and elegant.

1. The Royal Game of Ur is played on a board of 20 squares, arranged in 3 rows of 8 with four squares cut away (as shown in the diagram). Five of the squares are marked. The board can be thought of as in three distinctive sections: a large block, a small block and the bridge between them.

2. Each player starts the game with seven pieces in hand, the board being empty.

3. Each player has three binary lots in the form of pyramidal dice with four corners; two corners are marked. When throwing the dice the score is the number of marked corners pointing upwards, with none signifying a score of four.

4. Players decide at random who begins.

5. The path of a player's pieces starts on his side of the board, in the large block, at the square nearest the bridge. The piece moves toward the corner with the rosette, before moving to the adjacent square on the middle row and continuing till it crosses the bridge. Once across the bridge, it moves to the rosette on the opponent's side, then curves around the small block till reaching the rosette at the player's own side, from which it is borne off. The path is simpler than it sounds: see the diagram.

6. In his turn a player first throws his dice.

7. If none of his pieces are in play, then he must enter a piece on the first, second, third or fourth square on the board, according to the score of the dice.

8. If he has pieces already on the board, then instead he may move one of his pieces along its path by the number of squares indicated on the dice.

9. If the score of the dice was four, the player may after moving a piece, roll and move again.

10. A piece must bear off by an exact throw. For example, if a player's piece sits on the final rosette of its path, a 1 is required to bear off; if on the adjacent square, a 2, and so on.

11. Only one of a player's pieces may sit in a square at once; pieces cannot sit together in the same square.

12. If the roll of the dice gives no valid move, then the turn is lost and the opponent's turn begins. In this case no further roll is granted, even if the dice show four.

13. If landing on an opponent's piece, that piece is removed from the board and must begin its journey again.

14. A piece sitting on a special marked square is safe; the opponent cannot land on it.

15. The first four squares in a piece's path are also safe, as the opponent's pieces never land there.

16. The first player whose pieces are all borne off the board has won the game. Any form of binary lot may be used in place of tetrahedral dice; coins are the most readily available substitute.

Strategy in the Royal Game of Ur

Given that no complete and authentic rules exist for this game, the strategy is dependent on the version that you are playing. The rules used on this site are probably the ones giving most strategic interest to the game.

In all versions, one should pay attention to the probability of getting a particular throw on the dice. The chances of getting a 1 are 3 in 8, and similar for a 2. The chances of getting a 3 are 1 in 8, and the same for 4. Therefore a piece 3 or 4 squares ahead of an enemy is not as vulnerable as a piece 1 or 2 squares ahead.

Notice that with the paths given in these rules, the protected squares are exactly 4 squares apart throughout the board, so a roll of 4 not only gives another throw, but can move a piece from one protected square to the next.

Players will typically gather as many pieces as possible in the starting rows before setting out any further, as pieces there are safe from capture.

Once a piece enters the middle row, it should be progressed as quickly as possible either to the first protected square or, if that is missed, across the bridge to the smaller section of the board.

Pieces on the small section of the board are often safely left standing for a turn or two if no enemy is immediately behind them; this is a good time to attend to other priorities such as capturing enemies or moving other pieces out of danger.

The central protected square is the most advantageous on the board. A piece sitting here is safe, and is in a good position to capture enemy pieces as they move past. You will generally want to keep this square for as long as possible.


There is an interesting reconstruction of the game rules by Dmitriy Skyruk. There are lots of points which Dmitriy discusses, some of them are: 1. It is strange, that there is a special "safe" marking on a square 4, which is safe by design. It therefore should have different meaning 2. Not only "flower" squares have special rules, but more of them. 3. The rules by Bell do not make an interesting game. 4. Originally pieces were flat, so probably there was a possibility of reversing or stacking them (or both). There is a free Android app "Forgotten game of Ur" by Alexei Garbuzenko, which is build upon these rules, if you want to try them out. My e-mail is valid, so you may contact me for more details, if you wish.

Dmytro Polovinkin - 20:25, 05/03/2015

Thanks for this, Dmytro! Since I wrote the leaflet from which the rules above were taken, I've come to similar conclusions about the rosette squares. The simplest solution I've seen is that they give an extra throw.

I'll leave the leaflet as it is, and the rules above for now, not least because they give quite a good game (they're not Bell's, but take ideas from several sources). But if I revisit the subject in future, I'll be investigating that more closely, along with the work of Irving Finkel (Ancient Near East expert and board game enthusiast). In the mean time, I'll certainly be downloading that Android app.

Damian Walker - 09:25, 06/03/2015

hello i need to know how it looks like i mean can u please explain how it looks like??????????????????????????????????????

question - 15:30, 05/12/2016

The picture at the top of the page is of the ones I used to make. I don't reproduce museum photos here for copyright reasons, but if you do an image search on Google you'll find plenty of photos of the original in the British Museum.

Damian Walker - 06:12, 06/12/2016

I just ran across this.  Wonder if it's a descendent.

Karen Robinson - 17:03, 14/03/2017

I don't understand and i am wandering can you use 3 peices only for each side

Nogla72 - 18:17, 28/09/2017

Dmitriy Skiryuk's rules for the Royal Game of Ur are published on his blog in Russian here:

Eli Gurevich - 01:40, 11/10/2017

A variant of the rules that I have read gives a different interpretation to the rosette squares. This is a gambling variant, and any player landing on a rosette must pay one tally to his opponent. The winner of the game can claim a number of stakes from his opponent based on how many tallies he has. (I'd personally run with the rules that tallies can be paid off with the loser's own tallies, with only the difference between the two being paid in real money, and no money changing hands if the loser has equal or more tallies than the winner.) This interpretation is supported by the fact that some archaeological discoveries of the game included a number of small balls which may have been used as gambling chips.

I personally also wonder why all the squares have symbols on them rather than just the rosettes.

Jonathan Gwilliams - 10:44, 10/10/2019

Bell's reconstruction of the rules makes the rosette spaces 'penalty squares' on which a player must pay a certain stake into the pot, with the winner taking the pot. This seems like a good guess, since boards have been discovered with small round beads that may have been used like poker chips for placing bets.

The interpretation of the dice rolls for the Bell game are also interesting, as they are 1, 4 and 5. If you examine the antique boards, the symbols on the squares have one (rosette), four or five pips on them. I guess that the assumption was this was an aide memoire for novice players. I don't know if I agree with this or not, as it makes the game more complex.

On the whole, I lean towards the Irving Finkel interpretation of the game, as this was based directly on Mesopotamian descriptions of the game.

Jonathan Gwilliams - 06:20, 03/11/2019

In the Royal Game of Ur strategy section you write:
" The chances of getting a 1 are 3 in 8, and similar for a 2. The chances of getting a 3 are 1 in 8, and the same for 4. Therefore a piece 3 or 4 squares ahead of an enemy is not as vulnerable as a piece 1 or 2 squares ahead."

While the actual statictics when rolling 4 dice each with 2 out of 4 marked sides give the chance of getting a zero or a four are 1/16 each, of getting a one or three are 4/16 (i.e.1/4) and of getting a two is greatest at 6/16 (i.e. 3/8) [total sum 16/16].

So you should rewrite the vulnerability sentence to reflect that. E.g: "Therefore a piece 4 squares ahead of an enemy is least vulnerable while a piece 2 squares ahead is most vulnerable."

Gregor Shapiro - 06:01, 30/01/2020

The statistics you give, Gregor, are for four dice. In this reconstruction, though, there are three dice or casting sticks, matching the three pyramid dice (per player) found in some of the objects from Ur. For those, the stats I give are correct.

Damian Gareth Walker - 13:23, 19/02/2020

Ah you got there before me, Damian!

Worth mentioning that the newer Finkel interpretation does use four dice, but the older Finkel and Bell games both use three. Historic sets have been found with SIX dice, but whether this means there were three dice per player, which may suggest that three (per player) was the proper number, but it could also be that there were a couple of spares in the box.

Jonathan Gwilliams - 14:50, 19/02/2020

Did you know that apparently a version of the game survived among the Jews of Kochi until they moved to Israel.

Andrew J. Lucero - 05:17, 08/06/2021

As I understand it questionmark path must be the correct path and the path proposed by Finkel is wrong. Do you know if there exist any apps or online games implemented with questionmark path?

Olaf - 16:47, 13/02/2022

I'm sorry to be a nuisance guys, however, with four die each having two outcomes then the probability matrix is NOT of sixteen outcomes. Effectively these four sided Die are only two sided (like a coin) NOT Four sided numbered 1 to 4 Also, these die can score Zero. Yes, a four or a zero have the same probability (all die identical). (Making two Possible Outcomes) Then. A One has four outcomes, so therefore a three must also. (Giving eight more Possible Outcomes) (Making ten Outcomes). And finally. A score of two has EIGHT possibilities. Giving a total of Eighteen possible throws. Therefore: Zero = 1 / 18 One = 4 / 18 Two = 8 / 18 Three = 4 / 18 Four = 1 / 18

Dexx - 09:08, 12/09/2022

New Comment

Yes No