Tafl games

• 1. Introduction

  ›

  Tafl games were the dominant board game across Northern Europe for centuries, played from Norway to Ireland, from Iceland to what is now Russia. Two armies faced each other on a checkered board, but in a twist that set these games apart from almost every other strategy game, the armies were never equal. One side fielded roughly twice the number of pieces as the other. And in the centre of the board sat a single figure: the king.

  The games went by many names. Hnefatafl in Scandinavia. Brandubh in Ireland. Tawlbwrdd in Wales. Tablut among the Sámi people of the far north. Each was a variation on the same ancient family, a family whose origins stretch back to Roman influence during the medieval period. Then, in the 12th century, chess arrived in Northern Europe and tafl games slowly vanished. Almost completely. The Sámi kept playing tablut for another six centuries. When a Swedish naturalist named Carl Linnaeus joined their communities in 1732 and wrote down the rules, he may not have known he was preserving the last living thread of a game that had once spread across an entire civilization.

• 2. Across the Viking World

  ›

  Hnefatafl thrived during the Viking era, roughly from the end of the 8th century through around 1000 CE. It was a turbulent period, and the game followed the people who played it. Viking expansion carried hnefatafl not just to the British Isles but as far as Garðaríki, the Viking territory in what is now part of Russia. Archaeologists have turned up editions of the game in places as far apart as Ireland and Ukraine.

  The word hnefatafl literally translates from Old Norse as "fist table", with hnefi meaning "fist" and tafl meaning "table". The "fist" referred not to a clenched hand but to the central king-piece, the hnefi, which began every game in the middle of the board. That piece was so central to the game's identity that when Jarl Rögnvald Kali Kolsson composed a list of his nine boasts in the Orkneyinga saga, he placed skill at tafl at the very top.

  The physical remnants of the game are scattered across burial sites and religious locations alike. A wooden board and a single horn gaming piece were found in the ship burial at Gokstad in southeastern Norway. Twenty-two whalebone gaming pieces turned up in Orkney. Gaming boards made of walrus ivory, bone, amber, and wood have all been recovered. In 2018, a board dated to the 8th century or earlier was excavated at the site of what later became the Scottish Monastery of Deer. The following year, a small blue glass piece with swirled etchings and white glass droplets was found on the Holy Island of Lindisfarne, the target of a famous Viking raid in 793 CE. The trench it came from dated to the eighth or ninth centuries.

• 3. The Rules No One Wrote Down

  ›

  No complete, unambiguous description of how medieval tafl was played has survived. What remains are fragments: playing pieces, partial boards, incidental references in sagas, a handful of riddles. The king's objective was to escape to the edge or corners of the board, while the larger force worked to capture him. Beyond that, almost everything is contested.

  Whether dice were ever used in hnefatafl has been debated for generations. The sagas offer tantalizing but inconclusive evidence. A riddle posed by a character identified as Odin in disguise in the Saga of Hervör and Heidrek asks: "What is that beast all girded with iron, which kills the flocks? He has eight horns but no head, and runs as he pleases." The answer involves something called a húnn in hnefatafl. Some translators read that as a die, with the "eight horns" as the corners of a six-sided die and the "flocks" as the stakes the players lose. Others read húnn as the king-piece, with the eight horns being the eight defenders surrounding him. The riddle was designed to be ambiguous, and it still is.

  In Friðþjófs Saga, a conversation during a game of hnefatafl reveals that the king's men were red and the attackers white, consistent with a riddle in the Saga of Hervör and Heidrek that describes brown or red defenders and fair or white attackers. The same exchange confirms that hnefi refers to the king-piece. Archaeological evidence suggests the board may have been 13×13 or 11×11, but nothing settles that question with certainty either.

• 4. Tablut and a Flawed Translation

  ›

  Carl Linnaeus traveled along the Lule River into Sámi territory during his 1732 expedition to Lapland, crossing between the Swedish side of the border and Salten on the Dano-Norwegian side. Among the customs he observed was a board game played on a 9×9 mat of embroidered reindeer hide. The players called the defending pieces "Swedes" and the attacking pieces "Muscovites", a reference to the Grand Duchy of Moscow, a regional rival of Sweden. Linnaeus recorded the rules, a drawing of the board, and the piece designs in his journal, known as Lachesis Lapponica.

  Linnaeus likely misunderstood the word tablut, which in the Sámi language simply means "to play board games", as the name of the game itself. But no other name was known, and tablut stuck. The journal was translated into English in 1811 by James Edward Smith. The rules section was handled not by Smith but by a Swedish merchant in London named Carl Troilius, and Troilius made a critical error. His translation stated that four attackers were always required to capture the king. The original rules required only two, except in special cases. That single mistake turned a balanced game into one that heavily favored the defending side.

  Every attempt to reconstruct other tafl variants in the 19th and 20th centuries used the Troilius translation as a foundation, spreading the imbalance across the entire modern family of tafl games. Modern retranslations by John C. Ashton in 2007, Nicolas Cartier in 2011, and Olli Salmi in 2013 corrected the record. With the original two-attacker capture rule restored, statistics show the attackers win marginally more often, by around 9% on average.

• 5. The Variants

  ›

  Each corner of Northern Europe developed its own version of the game, and the differences are striking. Brandubh, the Irish form, was played with five pieces against eight on a 7×7 board, with one of the five designated as the chief, called a Branán. The most famous artifact of the Irish game is an elaborate wooden 7×7 board found at Ballinderry in 1932, featuring holes for pegged pieces, possibly so the game could be carried and resumed during travel. Two poems preserve what little is known of the original rules. Working from those poems, the 7×7 boards, and the tablut rules as a base, the World Tafl Federation reconstructed a playable balanced version.

  Alea evangelii, meaning "game of the gospels", was described with a drawing in a 12th-century manuscript held at Corpus Christi College, Oxford. It was played on a 19×19 board of intersections, and the manuscript frames the layout as a religious allegory, though the game structure is clearly in the tafl family. Ard Rí, a Scottish variant whose name means "High King" in Gaelic, was played on a 7×7 board with a king and eight defenders against sixteen attackers. It is the least documented of the known variants, and one of its generally accepted rules gives the defending side such a strong advantage that the game is considered solved: the king can always escape.

  Tawlbwrdd, played in Wales, is documented through an incomplete account by Robert ap Ifan, with a drawing in a manuscript dated 1587. Ifan's version used an 11×11 board with twelve pieces on the king's side and twenty-four on the attacker's side. His text includes a warning rule: if you say "Watch your king" before the king moves into a trap and he cannot escape, you capture him.

• 6. Modern Revival

  ›

  The first significant effort to bring tafl back to a general audience came in 1981 with the publication of "The Viking Game". It was based on the Troilius mistranslation and introduced the modern innovation of limiting the king's escape to the corners of the board, intended to offset the defender advantage the flawed rules created. The game booklet did not tell players it was actually rooted in the Sámi game tablut, and it claimed that hnefatafl had last been played "in Lapland in 1732". Sámi terminology including "raichi", "tuichu", and "konokis" was left out entirely. Despite those omissions, the publication did much to ignite serious interest in the game.

  In 2008, Peter Kelly revived hnefatafl on the island of Fetlar in Shetland. The annual World Quickplay Hnefatafl Championships have been held there each summer since, under the World Tafl Federation, with a gong marking the ten-second-per-move time limit. The World Tafl Federation itself was formed in August 2011, with Tim Millar as Chairman and Adam Bartley as Vice Chairman. Millar also won the first two world championships, in 2011 and 2012, under Fetlar Hnefatafl rules on an 11×11 board.

  Online play expanded the community further. Aage Nielsen launched his tafl website in 1998, which now hosts the World Tafl Federation Hnefatafl Championship Tournament. A second major site was launched in 2014 by Jacob Teal and John Carlyle. By the 2020 world championship, participation had grown to 34 players, and the championship included both Copenhagen Hnefatafl and a historical 9×9 format based on Sámi tablut. Mario Aluizo of Los Angeles has won the championship five times, most recently in 2025.