The origin of the Viking board game Hnefatafl and the Berserk rules.

Hnefatafl origin.

The historical period 1-400 A.D. is in northern Europe named the Roman Iron Age because of the vast influence from the great Roman Empire just south of the area. Many northern European warriors served in the Roman army and there became acquainted with the widespread Roman war game Ludus Latrunculorum ("Latrunculi, soldiers"). Returning to their home lands, talented gamers modified the game into tafl, replacing the model of Roman warfare with the homely, iron age way of warfare.

Latrunculi is based on the custodian capture rule which is also the basis of the tafl game.

Robert Charles Bell's book "Board and Table Games from Many Civilizations", published 1960, republished 1969 and 1979, tells about "The Latrunculorum Group" of board games that descendants of this game are found in Somalia and in northern Europe, both areas previously bordering to the Roman Empire.

In Somalia still exist the games Seega and High Jump, both based on custodian capture. The rules of both games having very much in common with Latrunculi and tafl, one notices specifically two interesting rule details: In both games the player may continue to move the same stone as long as he makes captures with it; and High Jump has an alternative way of capturing by jumping over an enemy piece.

R.C. Bell notes that Seega and High Jump may be survivals of the Roman game, each preserving one of its forms of movement and then suggests these rules for Ludus Latrunculorum:
1. Using an 8x7 board each player has 16 pieces. They are placed two at a time by alternate turns of play anywhere on the board. During this first phase no captures are made.
2. When the 32 pieces are in position each player adds a blue piece to the board. This is the DUX. The second phase the begins. The pieces can move one square orthogonally in any direction.
3. A capture is made by trapping an enemy piece between two of one's own pieces on rank or file.
4. When a piece makes a capture it has another turn, and from this it follows that an isolated piece may endanger itself and several of its fellows by starting a sequence of captures.
5. The dux may move in the ordinary way, or it may make a short orthogonal jump over an enemy piece, landing on an empty square beyond. It does not capture by this maneuvre, unless it traps another enemy piece between itself and one of its own men. The power of jumping enables it to penetrate a defensive position which may start a whole series of captures from within a walled stronghold. It is captured in the same way as any other piece.
6. A piece may move between two enemy pieces without being captured.
7. The game continues until one player has lost all his pieces, or a blockade has developed which neither player can break. The player with most pieces left on the board then wins.
8. If no captures have been made in 30 moves a blockade has been established and the game is over.

Latrunculi being the role model for Hnefatafl, one notices that two interesting Latrunculi game rules, 4 and 5, are missing from the Hnefatafl reconstructions.

The Storhaug Hnefatafl set.

September 2011 Adam Bartley sent me an interesting link to the Nordvegen History Centre, which showed a photo of Hnefatafl glass and amber playing pieces from Bergen Museum.

Bergen Museum Hnefatafl set
(Photo Bergen Museum.)

Norvegen Historiesenter tells more about these pieces (Text: Marit Synnøve Vea):

In the summer of 779, a powerful king was buried in a large ship at Avaldsnes.
Two sets of board games.
Amongst the burial gifts were two sets of costly board games, one with counters made of amber and the other with counters of blue and yellow glass.

The counters were probably of the type used in the board game called Hnefatafl (Old Norse for "the king's board"). Hnefatafl was certainly played in Northern Europe and in England from the 8th century and into the Middle Ages. The game is similar to chess, whereby guards aim to defend the king against attackers by bringing him safely into one of the castles on the board.

The skald of Harald relates that Harald's "quick warriors" play a board game in the yard of the royal court. The Edda poem Voluspå also tells of the gods playing a board game at Idavollen.

Photo text:
Two sets of board games were found in Great Mound (Storhaug). The counters made of amber were originally 20 pieces. The counters made of glass were distributed as follows: 12 blue, 4 yellow and 1 black. (Photo Bergen Museum)

My comments on the Storhaug photo.

The photo shows one black glass piece with decorated top and a large pin hole suggesting a substantial, missing top decoration [the king].
12 dark blue glass pieces [the king's men, defenders]. A closer look on the dark defenders reveals that one more of them has a distinct pin hole on the top, suggesting that also this game piece originally had a mounted, decorated top.
18 larger, yellowish / reddish amber pieces (originally 20) [attackers]. And 4 yellow glass pieces with decorated tops [attackers].

Now, what if these pieces are one set, and not two sets as the museum presumes? The total collection is 1 black, 12 blue and 24 yellow pieces, which is precisely what it takes for one 11x11 Hnefatafl board game set. However, this game set is unusual in that it has four types of pieces instead of the usual two. Usually there are men and one king; but here there is a special piece more on the king's side, and a special piece more on the attackers' side (four of them).

Why is this game set so unusual? Perhaps it was prestige for a top ranked viking king to own a hnefatafl game set differing from and more sophisticated than the common vikings'?

After productive discussions with Adam Bartley on Viking war ethics (se insert below), to me these playing pieces tell a story of drama:
Four Viking long boats, each with a ship commander and his crew of warriors, has landed on the coast of a foreign country, near a castle where a king has barricaded himself with his warriors. The Vikings have surrounded the castle from four sides, each side grouping a commander and his men. The large, yellowish / reddish pieces are the blonde Vikings, represented by a light color, and to that the color of blood and burning fire. The smaller, dark blue pieces are the opponent, represented by a dark color, the color of the cold ocean.
   Being a Viking game, the viking hotheads would probably identify themselves more with the orange fire-pieces, that is the attackers!

Bergen Museum Hnefatafl set
(Archeologist's drawing of the Bergen Storhaug pieces.)

Bergen Museum Hnefatafl set

The Berserk Hnefatafl version.

With a look at the reconstructed Ludus Latrunculorum with the two extra rules, and related ancient games and the Storhaug set of playing pieces with some of them distinctively decorated, Aage Nielsen, Denmark, proposed a Hnefatafl version for this set of pieces, which could be named the Berserk Hnefatafl version.

Explanation of the rules derivation.

The Berserk Hnefatafl rules are deduced from combining the Fetlar Hnefatafl, the R.C. Bell's reconstruction of the Roman soldiers' game Ludus Latrunculorum, the Somali related games Seega and High Jump, the Saami Tablut, and the Bergen Museum set of Hnefatafl playing pieces from Storhaug.

The logics behind the rules are this:

2-side king capture, or a weaponless king, results in a cautious playing style where the king is very immobile. On the other hand, 4-side king capture and the king beweaponed results in a much more vivid game. This is also the rule in Fetlar Hnefatafl.

When the king is this strong (4-side capture and beweaponed), he must go to a corner to win, like in Fetlar Hnefatafl. If the king can be captured on the edge, then it takes only two enemies to capture him next to a corner square. As the Fetlar Hnefatafl Panel found out, this would be a game spoiler. Therefore the king cannot be captured on the board edge.

The Bergen set of playing pieces include a king, 12 defenders and 24 attackers. This number of pieces matches the 11x11 game board. Four of the attackers are distinctively decorated, suggesting they should have a distinct position on the board, fx. in front of the ordinary attackers. So the initial ordering is that of Fetlar Hnefatafl.

This far, the rules are identical to Fetlar Hnefatafl.

Now to the additions:

The Bell's Latrunculorum reconstruction allows any piece to continue moving as long as it makes captures. So do the Latrunculorum Somali descendants Seega and High Jump. This rule is introduced to Berserk Hnefatafl as the Berserk rule (or re-introduced, as it has perhaps been there but was lost in time).

Four of the attackers in the Storhaug set are distinctively decorated (all four alike) and must have a special function, fx. being four ship commanders. And to that, one of the defenders has a pin hole in the top, suggesting some missing decoration here also. This piece could likewise have a special function, fx. being the king's knight or his personal guard.

The game now has three types of officers, the king, his knight and the ship commanders. These must have some kind of special abilities.

Those abilities might be:

The Bell's Latrunculorum reconstruction allows the Latrunculi officer, the Dux, to jump over enemies. Introduced (or re-introduced?) to Berserk Hnefatafl, the three types of Hnefatafl officers, the king, the knight and the commanders, should be able to jump over enemies.

As in Bell's reconstruction the Dux'es are captured the same way as any other piece, so are the knight and the commanders.

There is good use for a piece which can block the way between the powerful king and a corner, especially with the king's new ability to jump over an enemy and right into a corner square. This would be an obvious task for the attacker commanders, with introducing of a rule that the king cannot jump over a commander. With such a rule in effect, it is natural to generalize, so that the rule applies to all officers, meaning that the king, the knight and the commanders cannot jump over each other but only over ordinary pieces.

Another flavour of extra power to the attacker commanders can naturally be that two commanders can capture the king (with custodian capture) instead of the normally required four attackers. And just like the hostile king's squares can function as an enemy piece in ordinary custodian captures, it is logical that the king is captured by one commander against a corner square, the corner square functioning as a second commander.

If the king is always allowed to jump over enemies, it will be nearly impossible to bring four attackers in place to capture him. Therefore this ability of the king's must be limited in some way. A natural bid would be that he can jump only to and from the five king's squares but not jump enemies in the open field.

The king has only one special helper, the knight. To give the knight some extra ability, it was tried to allow the knight to enter the five king's squares just like the king. But test games revealed that such a rule made the knight of very little use beyond that of an ordinary defender.
   In the search for some other ability for the knight to have, one notices that the Somali descendant of Latrunculi, the game High Jump has an alternative way of capturing by jumping over an enemy piece, in addition to the custodian capture. This ability given to the knight makes him a very powerful piece, which fits with the observation that there is only one distinctively decorated defender besides the king in the Storhaug game.
   The suggested move pattern for the knight makes him the only piece on board killing enemies all alone, with no help from comrades or king's squares. The move pattern looks like the behaviour of a warrior on horseback - the knight armed with horse and sword, jumping over enemies with his horse and killing them off.
   The attackers arrived in long boats and cannot bring horses.


The Berserk Hnefatafl uses the Storhaug game set and merges rules from Fetlar, Tablut, Ludus Latrunculorum, Seega and High Jump. Noone rule was introduced which is not already known from one or more of those games.
The three types of "officer" and their color and number, come from the Storhaug game set.
The berserk run comes from Ludus Latrunculorum and its Somali descendants Seega and High Jump.
The knight jump capture comes from High Jump.
The commander jump and the king jump come from Ludus Latrunculorum.
The commander king capture comes from Saami Tablut.
The rest is Fetlar.

The Berserk Hnefatafl 11x11 rules were ready October 9th 2011.

Berserk Hnefatafl board
Berserk Hnefatafl 11x11 board.

More about the Berserk Hnefatafl in our forum.
Berserk test tournament.
The measured game balance is +1.18 (118 white wins per 100 black wins), which is a fine balance.

Aage Nielsen, Denmark, October 9th 2011.

I received this from Adam Bartley, Tønsberg, Norway:

I have had an idea, which I will be getting some more evidence of as my idea is based on one viking historian's answers to some questions I had about the vikings concept of honour.

The viking war exhibition I'm working on has the concept of honour and ettermæle (how one is remembered) featuring heavily. The book we are basing the exhibition on doesn't clarify it enough, so I asked our historian about it, here is his response:

Hei Adam
Hvorvidt boken forklarer begrepet ære i denne sammenhengen kan jeg ikke akkurat nå si med sikkerhet. Begrepet en "ærerik død" er det som best beskriver forholdet, og er således helt knyttet opp mot ettermælet. Om man tapte kampen, død eller levende, betyr ikke nødvendigvis noe for det ærerike i hvordan kampen ble utkjempet. Å tape et slag ved å kjempe til siste mann er derfor en ærerik sak. Uavgjorte kamper - som på fotballbanen - var neppe en opsjon. Selvfølgelig kunne man bevare æren selv om man kom levende ut av et tap, men absolutt IKKE ved å rømme fra slagmarken! I alle fall ikke hvis man var hærleder.
   Farlig å sammenligne med ærebegreper i vårt vestlige samfunn i dag, og kanskje like farlig å sammenligne med ærebegreper i andre deler av dagens verden. Samfunnet var den gang skrudd sammen på en annen måte. Jeg pleier vanligvis å betegne dette som et før-religiøst samfunn. Dette er selvfølgelig kontroversielt, men henger nøye sammen med definisjonen av begrepet "religion".

Whether the book clarifies the notion of honour in this context (war/death in battle) is something i can't say with certainty at this point. The term 'an honourable death' is that which best describes the relationship [between death and battle], and is thus tied up with legacy [how one is remembered after death]. Whether one lost the battle, dead or alive, didn't necessarily effect the 'glory' in terms of how the battle was fought. To lose a battle fighting to the last man is therefore an honourable case. Undecided battles, as in football, were an unlikely an option. One could of course retain honour even if one lived through a losing battle, but absolutely NOT by running from the battlefield! At least if one was the captain(leader/head of fighting force.
   Its dangerous to compare notions of honour in our western society today, and perhaps equally dangerous to compare it to notions of honour elsewhere in todays world. Viking society was composed in another way. I usually denote this as a pre religious society. This is of course controversial, but hangs closely together with the definition of the notion of 'religion'.

So here is my idea: At first it seems, if this historian is right about the vikings, that draw forts are emphatically not an option. BUT, the very concept of an escaping king is also ruled out by exactly the same argument! Now, we know the kings job is to escape, its the same in the two written sources. So, this suggests to me either: 1. The attackers are vikings, the defenders are NOT vikings, they are indeed considered cowards by the vikings. 2. The attackers are good, odin fearing vikings, the defenders are cowards who don't deserve to be called vikings, maybe they are christian converts or something.

The upshot of all this, is that the kings side is positively encouraged to play dirty, to build draw forts, to force perpetual check, to play as cowardly and dishonourably as possible in order to infuriate the vikings. Then the players swap, and the entertainment continues, one playing the dastardly cowards, the other the noble vikings.

This fits in well for me with the idea that hnefatafl is a 'hunt' game, as opposed to a pure battle game. It also fits well with the moscovites vs swedes culture clash. It doesn't confirm a weaponless king, but it does make a weaponless king acceptable if he is not a proper viking. And if my historian is right, an escaping king is NOT a viking. I can't quite believe nobody has spotted this glaring point before!

I think its worth cross checking. Importantly my inquiry did NOT mention hnefatafl at all. It was purely a question about defining viking honour on the battle field.

If other historians agree, then this would rather put to bed many of the arguments we keep running into.

Adam Bartley, Tønsberg, Norway, Oct. 4th, 2011.

Updated 13.7.2017
Copyright © 1998-2017 Aage Nielsen, All Rights Reserved;