The History of Backgammon

Backgammon is one of the oldest known board games in human history, with origins stretching back roughly 5,000 years to ancient Mesopotamia. It is older than chess, older than Go by most estimates, and has been played continuously across cultures and centuries in forms remarkably similar to the modern game. From royal tombs in Ur to internet servers handling millions of games, backgammon's story is one of remarkable endurance and adaptation.

The Royal Game of Ur (3000 BC)

In the 1920s, archaeologist Sir Leonard Woolley excavated the ancient Sumerian city of Ur (in modern-day Iraq) and discovered game boards dating to approximately 2600 BC. The Royal Game of Ur is one of the earliest known board games — a race game played with dice and pieces on a board of 20 squares. While not backgammon itself, it shares the core DNA: two players, dice-driven movement, and a race to bear pieces off the board.

Similar dice-and-race games were found across the ancient Near East, Egypt, and the Mediterranean. The Egyptian game Senet, dating to around 3100 BC, and the game of Twenty Squares, played throughout Mesopotamia for over 2,000 years, are part of the same family of race games from which backgammon eventually evolved.

Roman Tabula (1st Century AD)

The Roman game Tabula (also called Ludus Duodecim Scriptorum, the “game of twelve lines”) is the first game that is recognizably backgammon. Played on a board of 24 points with 15 pieces per player and three dice, Tabula had the same basic mechanics as modern backgammon: pieces move in opposite directions, landing on a single opponent piece sends it back to the start, and the goal is to bear off first.

Tabula was enormously popular in Roman society. Emperor Claudius (41-54 AD) was reportedly obsessed with the game and had a board mounted on his chariot so he could play while traveling. Emperor Nero allegedly played for stakes equivalent to $10,000 per game in modern currency. The Roman Senate building had Tabula boards available for senators during breaks.

When the Roman Empire fell, Tabula survived. It spread throughout the Byzantine Empire and into the Islamic world, where it became known by various names. The game continued essentially unchanged for centuries.

Medieval Europe and the Eastern World

By the Middle Ages, variations of backgammon were being played across Europe, the Middle East, and Asia. In Persia, the game was called Nard or Nardshir. Legend attributes its invention to the Persian sage Bozorgmehr in the 6th century, though the game clearly existed earlier. The Persian rules were nearly identical to modern backgammon.

The Moors brought the game to Spain, where it became Tablas Reales (“Royal Tables”). Alfonso X of Castile included it in his famous 1283 Libro de los Juegos (Book of Games), one of the most important medieval documents on games. In France, it was called Tric-Trac; in Germany, Puff; in England, simply Tables.

The game was universally popular among the aristocracy and was considered one of the essential skills of a gentleman. It also had a controversial side — because it involved dice and was frequently played for money, religious authorities periodically banned it. Cardinal Wolsey ordered all backgammon boards in England burned in 1526. The game survived every prohibition.

The Name “Backgammon”

The word “backgammon” first appears in English in 1645. Its etymology is debated: the most common theory is that it comes from Middle English baec (back) + gamen (game), referring to checkers being sent “back” to the bar. Other scholars suggest a Welsh origin from bach (small) + cammaun (battle). By the late 17th century, the game was called backgammon throughout the English-speaking world, and Edmond Hoyle's 1743 treatise standardized the rules that are essentially the same today.

The Doubling Cube Revolution (1920s)

The most important innovation in modern backgammon was the invention of the doubling cube in the 1920s, likely originating in New York City gaming clubs. The cube transformed backgammon from a simple race game into a deep strategic contest with elements of poker-like bluffing and risk assessment.

The doubling cube added an entirely new dimension: players had to evaluate not just the best move but whether the position justified raising the stakes. Knowing when to double, when to accept, and when to concede became as important as checker play itself. The cube also shortened games — players could resign unfavorable positions rather than grinding out hopeless endings.

The cube helped fuel a backgammon boom in the 1960s and 1970s, particularly in affluent social circles. Prince Alexis Obolensky organized the first major backgammon tournament in the Bahamas in 1964, and the World Backgammon Championship began in 1967 in Las Vegas. By the 1970s, backgammon was a mainstream sensation — nightclubs hosted tournaments, celebrities played publicly, and the game appeared regularly in magazines and television.

Computer Backgammon and TD-Gammon

Backgammon has played a fascinating role in the history of artificial intelligence. In 1979, Hans Berliner's BKG 9.8 became the first computer program to defeat a reigning world champion in any board game, beating Luigi Villa 7-1 in a short exhibition match. However, BKG relied heavily on hand-coded rules and was not consistently at world-class level.

The real breakthrough came in 1992 when Gerald Tesauro of IBM created TD-Gammon, a neural network that learned to play backgammon through self-play using a technique called temporal difference learning. Starting with no knowledge of strategy, TD-Gammon played over 1.5 million games against itself and reached a level comparable to the world's best human players.

TD-Gammon was a landmark in AI — it demonstrated that a neural network could learn complex strategy from scratch through reinforcement learning, predating DeepMind's AlphaGo by over two decades. TD-Gammon also changed human backgammon strategy: the program discovered that certain moves considered wrong by human experts were actually superior, particularly aggressive early slotting plays.

Modern backgammon programs like GNU Backgammon and eXtreme Gammon build on TD-Gammon's approach and play at a superhuman level. They serve as analysis tools for top players, much like chess engines do for chess players.

Backgammon in the Digital Age

The internet has transformed backgammon in the same way it has transformed chess and other classic games. Online platforms allow players to find opponents instantly at any skill level, analyze games with neural network engines, and compete in tournaments from home. The game that was once limited to coffeehouses and private clubs is now accessible to anyone with a web browser.

Backgammon remains one of the most widely played board games in the world, with particular popularity in Turkey, Greece, the Middle East, and the Nordic countries. Major tournaments continue to draw international fields, and the combination of strategic depth, dice excitement, and quick game times makes it one of the most engaging two-player games ever created — a 5,000-year track record that shows no sign of ending.