How the Vikings Really Played Hnefatafl
A game woven into the culture
In Völuspá, the great prophecy poem of the Poetic Edda, the gods of the golden age play tafl with golden pieces — tǫflur — and after Ragnarök destroys the world, the survivors find those golden pieces again in the grass. A board game, written into the beginning and the end of the world. One caveat before we get carried away: tafl is a generic Old Norse word for board games, so the poem doesn’t name hnefatafl specifically — but it tells you how seriously the Norse took play.
The aristocracy took it just as seriously. Rögnvald Kali Kolsson, jarl of Orkney from 1137 to 1158, left a verse in the Orkneyinga saga boasting of his nine accomplishments — and first on the list, ahead of poetry and runes, comes “Tafl emk ǫrr at efla”: I am swift at playing board-games. Board-game skill was aristocratic identity, worth bragging about before your verses.
The real boards
This isn’t just saga material — the boards survive.
The finest is the Ballinderry board: a 7×7 gaming board of yew, excavated in 1932 from a crannog at Ballinderry, Co. Westmeath, Ireland, from a site dated to the late 9th–11th centuries. Its interlace carving is in the Scandinavian Borre style, though it was likely made in Ireland, and it now sits in the National Museum of Ireland in Dublin. At 7×7 it is most likely a board for brandub, the small Irish member of the tafl family — the same layout we offer as Brandubh.
Larger boards existed too. A wooden board fragment from Trondheim, dated to roughly 1100–1150, is incised with an 11×11 grid, professionally finished with mitered trim and a tric-trac board on the reverse — evidence that the big 11×11 format was real and that tafl games persisted into the early 12th century. And the famous Gokstad ship burial (c. 890–900) included a board fragment with Nine Men’s Morris on one side and a latticed game of about 13×13 on the other, along with horn playing pieces.
The graves
The most striking find is a grave. Birka Bj 581, a richly furnished burial from about 940–950 excavated in 1878, contained 28 bone gaming pieces, a king piece, three dice, and the iron fittings of a game board. The set has long been read as a marker of strategic, commanding status — and in 2017, DNA analysis confirmed something that upended a century of assumption: the person buried with that command-and-strategy kit was female. (The identification of the set as hnefatafl is the scholarly consensus, though it carries the usual caveat.) Gaming pieces turn up as elite grave goods elsewhere too — the 9th-century Skamby boat grave in Östergötland held 23 amber pieces.
The game that names itself
How do we know these boards were for an asymmetric king-escape game at all? Because a saga tells us — in a riddle. In the Hervarar saga, a riddler calling himself Gestumblindi asks: Who are those girls who fight weaponless around their lord? The darker protect him, and the fair attack. The answer given: that is hnefatafl — the dark pieces defend the hnefi, the king, and the white pieces attack. It is the most direct Old Norse description of the game we have: two unequal sides, a king in the middle, defenders and attackers. Friðþjófs saga corroborates the picture, describing a game where “the king’s men are red and the attackers white,” with hnefi again meaning the king piece.
The twist: nobody actually knows the rules
Here is the part most “Viking chess” marketing quietly skips. No complete medieval ruleset for hnefatafl survives. Not one.
The closest thing we have to a rulebook comes from an unlikely witness: Carl Linnaeus, the Swedish botanist, who during his 1732 Lapland journey watched the Lule Sámi play a game called Tablut — a 9×9 board of reindeer hide, 8 defenders and a king (the “Swedes”) against 16 attackers (the “Muscovites”), with rook-style movement and capture by flanking. He wrote the rules down, and his notes are the only period eyewitness ruleset for any tafl game. Even that thread got tangled: James Smith’s 1811 English translation mangled a key passage — turning “likewise the king” into “except the king” — and the error skewed reconstructions of the game for a century.
So every modern hnefatafl ruleset, including the ones we use, is a reconstruction: Linnaeus’s Tablut, scaled and interpreted, then play-tested for balance. That’s not a weakness of the game — it’s an ongoing, genuinely interesting design problem, and it’s why we ran the numbers on our own rules in our rebalancing write-up.
Clearing up the myths
A few claims you’ll see repeated that don’t survive contact with the evidence:
- “The Lewis chessmen are Viking chess pieces.” No — the Lewis chessmen are chess: kings, queens, bishops, knights, warders, and pawns, carved in walrus ivory around 1150–1200. If anything they mark the arrival of the game that eventually displaced hnefatafl.
- “Hnefatafl dates back to 400 AD.” Shaky. The boards from Vimose (c. AD 400) are almost certainly the Roman game Ludus Latrunculorum, not tafl. The earliest unambiguous tafl finds are Viking-Age.
- “The Vikings invented it.” Viking chess is a modern nickname. The tafl family likely evolved from Roman latrunculi and was adapted by Germanic peoples before the Viking Age even began.
- “The Ockelbo runestone shows people playing hnefatafl.” The original stone was destroyed in a church fire in 1904; today’s stone is a 1932 replica based on 1880s drawings. It shows a board-game scene, probably tafl — but the variant can’t be pinned down with the confidence popular accounts claim.
- “Tafl in the sagas always means hnefatafl.” Tafl is generic — it could mean chess, tables, or fox-and-geese. Only the passages that say hnefatafl, or clearly describe the asymmetric king game, actually count.
Play the reconstructed game
What’s left when the myths are cleared away is better than the myths: a game embedded in Norse cosmology and aristocratic identity, known from real boards in Dublin and Trondheim and a warrior’s grave in Birka — whose rules had to be rescued from a botanist’s travel notes. You can play the reconstruction in your browser, or start with the rules we use and see how the pieces of evidence fit together on the board.