Hnefatafl: Authenticity and Good Game Design

Saturday, 28th May 2016

There are two quests among those of us trying to promote hnefatafl: authenticity and good game design.  The first is to recreate the ancient game played by the Vikings and other historic people.  The second is to have a game that plays well.  Sometimes these aims seem to work against one another.

To achieve the aim of authenticity you would need a game that satisfies the historical evidence.  It is one thing to speculate on evidence not yet found, but you can't have authenticity by contradicting the evidence we have.  For example, when the botanist Linnaeus saw Tablut played in the eighteenth century, he gave specific examples of the king winning at the board edge.  Any reconstruction calling itself "tablut" that has the king aiming for a corner square is not authentic.

But to achieve good design, you need a game that works well: both sides must have an approximately equal chance of victory or the game becomes pointless.  Many games from past decades had a strong king who could capture attackers, who needed to be surrounded on four sides to be captured, and had to reach the board's edge as Linnaeus told us.  These matched the evidence available at the time and could have supported the claim of authenticity.  But they were badly designed: the king had an easy time would always win.

The latter problem has been solved as more information has become widely available; old works such as Linnaeus' full account of the game have been digitised by libraries and universities and put on-line for anyone to consult.  But there are still areas where the quests for authenticity clash with quests for good game design.

A decent reconstruction of the Irish game Brandub has been found that matches historical evidence and plays well.  It uses thirteen pieces on a board of seven rows of seven squares.  The rules are similar to tablut with the small change that the king does reach a corner to win, as attested by corner markings on some boards and a strong implication in some Irish poetry.  But similar Scottish boards lack corner markings, making it unlikely that the Scottish games were played in the same way.

Simply using the rules for tablut gives the king too easy a task, and so such a game would fail the design criteria.  Crowding 25 pieces onto the board (as in the spurious historical game of Ard Ri) has the opposite effect: the attackers easily win.  Adding in rules such as single-step movement or a weaponless king (who cannot capture) have no historical evidence so while they may result in a well-designed game, they depart from authenticity.

A similar issue affects the thirteen-row board.  Putting 49 pieces on this board (a king with sixteen defenders against 32 attackers) and using the otherwise authentic tablut rules creates a well-designed and playable game.  But very little evidence of such a large set of pieces exists (the 47 pieces from Nes being a disputed and sole example).  Putting the more authentic 37 pieces on this board creates a badly-designed game where the king will always win.

Aage Nielsen, who runs an excellent on-line play site, adopts a scientific approach to the problem.  The large number of skilled players who regularly use his web site are a great help in determining which rules work well and which don't.  These players willingly submit themselves to experiments in game-play, and Aage can analyse the results of many games.

Starting with the historical assumptions, Aage tests any uncertanties with friendly tournaments.  Where history provides no help, modern ideas are introduced.  Sometimes this results in departures from authenticity, such as the adoption of R. C. Bell's improbable 1969 layout for tawlbwrdd or my own equally improbable layout for a 15x15 board.  But in these cases authenticity takes us as far as it can, and departures are at least justified by good game design.  Inauthenticity is kept to a minimum.

Aage has published his conclusions on the site. This page gives a unified set of rules for an historic hnefatafl game.  It's very useful for someone who wants to use hnefatafl as a re-enactment or educational activity.  Not only will participants be playing a game close to that played by the Vikings, Anglo-Saxons and Celts, but the game will be well-designed and free of fundamental flaws.  Perhaps then the participants will want to continue playing the game when the themed activity is over.