Tafl games

Tafl games are a family of ancient Germanic board games played on a checkered board with two teams of uneven strength. Versions were played across much of Northern Europe from at least 400 CE (AD) until it was supplanted by chess during the Renaissance. Tafl games are possibly descended from the Roman game ludus latrunculorum.

The term "tafl" (pronounced "tavl") comes from the Old Norse word for "table," which was often used to refer to a game board.


Versions
Hnefatafl is the name of a game frequently referred to in the Norse sagas. Vikings considered skill at hnefatafl to be a valuable attribute, and in one story one player kills the other due to a disagreement over the game. However, the rules of this game were never recorded, and only playing pieces and fragmentary boards are extant, so it is not known how the game was really played.

Tawl-bwrdd was played in Wales. Robert ap Ifan documented it in a manuscript dated 1587. It was played on an 11×11 board. The rules are somewhat vague, but it is unquestionable that this is a tafl game.

Alea evangelii, named after the first two words of the manuscript in which it is described, was a version played in Saxon England. It is played on a 19×19 board (actually, a 18×18 grid with pieces on the intersections.) The manuscript gives the rules of the game as a religious allegory, so it is difficult to infer exactly what was meant, but it seems to strongly indicate that this was a tafl game.

Brandub is an Irish game likely part of the tafl family, but little is known about it. We know from two poems that it was played with five men against eight, and that one of the five was a "Brannan", or chief.

Tablut, from Lapland, is the best documented version. In 1732, Carolus Linnaeus recorded the rules and a drawing of the board in his journal while travelling through the area. It is played on a 9×9 board. Linnaeus referred to the light (defending) pieces as "Swedes" and the dark (attacking) pieces as "Muscovites."

Halatafl dates from at least the 14th century and is still known and played in Europe, see Fox games.

Reconstruction
As mentioned above, no complete, unambiguous descriptions of the rules of a Tafl game exist. However, there is enough information to make a reasonable guess as to how the game might have been played. The following is one possible reconstruction of the rules of Tablut, staying as close to the original source as possible.

The game is played on a 9×9 board. Initial set-up is as shown in the diagram. 
The object of the game for the defenders is to have the king reach any square on the periphery of the board. 
The object of the game for the attackers is to capture the king. 
All pieces have the move of a chess rook; that is, they can move any number of spaces in any orthogonal direction until they encounter an obstruction. 
All pieces, with the exception of the king, are captured by surrounding them on two opposite sides with enemy pieces. A capture can only result from the movement of an enemy piece; a piece moving between two existing enemy pieces is not captured. 
After the king has left the center square (the castle) of the board, no pieces can be moved onto it (however, other pieces can move through it). 
The king is captured by surrounding him on all four sides, or three sides and the throne. 

Controversy
The above represent the rules of Tablut as described by Linnaeus' diary from 1732, and this is the typical way in which Hnefatafl and its derivatives are played by reconstructionists. However, Neil Peterson has researched into the historical sources and even carried out a scientific study suggesting that Linneaus' special capture of the king is incorrect.

Firstly, snippets within the description of Tawlbrydd by Robert ap Ifan (written earlier than Linneaus' account, in 1597) imply that the king could be captured by two men. Secondly, it is generally believed that Norsemen considered themselves to be equal, while the notion of "the divine right of kings" was a Christian concept. So based on literary sources, there is some reason to believe that the king could be captured by only two men.

Lastly, Peterson's attempts to determine the balance of the game with 4-man capture of the king provided somewhat more ambiguous results. While his study itself found that the king's side wins 100% of the time when 4-man capture of the king is required, there is some reason to believe that Peterson's study failed to accurately replicate play, because his study used novice players who tended to finish their games rapidly. Considering the performance of Black and White listed at www.brainking.com, the advantage of the king's side lessens as the rating of both players increases. Furthermore, Black wins tend to take longer than White wins; novice Black players are thus more likely to make a mistake during the long, slow road to victory. It remains to be seen whether balanced play results from rules allowing for a two-man capture of the king.



References
Bell, Robert C. (1979). Board and Table Games from Many Civilizations (Revised ed.). New York: Dover Publications, Inc. ISBN 0-486-23855-5. 
Murray, H. J. R. (1952). A History of Board-Games Other than Chess. Oxford: Oxford University Press Reprints. ISBN 0-19-827401-7.