I thought I'd follow up last month's Brandubh variant with another one you might be interested in. This one is based on the Welsh game tawlbwrdd. What we know about the game comes from a Welsh manuscript in 1587, written by Robert ap Ifan. He gives an imprecise description of the layout of the pieces on the board, and neglects to tell us how they move, but he does tell us about capture. Victory for the king isn't clear. Details have been filled in mainly by R. C. Bell, who I follow. But I don't bother with the dice that he introduces, as these aren't mentioned in the manuscript and they really ruin the game in my opinion.

1. The game is played with a king, twelve defenders and twenty-four attackers, on a board eleven squares by eleven. The king sits in the centre, with his defenders around him in a diamond shape, while the attackers form four hollow pentagonal shapes around the edge of the board. I won't even attempt to describe this layout in any more detail, it's best viewed here:

http://ca.wikipedia.org/wiki/Imatge:Tauler_de_tawlbwrdd.jpg

2. The object of the game for the king is to reach any square on the edges of the board. The attackers, as usual, have to capture the king.

3. The attacking side begins. Movement (it is now supposed) is in the same way as tablut: as far as you like either horizontally or vertically. There are no recorded restrictions on the central square.

4. Capture for ALL pieces is by custodianship, that is, surrounding an enemy piece on two sides by pieces of your own. The king may capture, but may himself be captured in the same way as other men.

In Bell's reconstruction, only the king can come to rest between two other pieces, but from my memory of the original rules (which I read in translation), this seems to apply to all pieces. The variant I present here seems, after a few games, to be relatively balanced.

Fwiffo: OK, consider it done. I think a bigger board and more pieces do tend to make the strategy more long term, but my only real experience is half a dozen games of this brandubh and about thirty games of tablut.

snigfarp: I just need to make myself a set!

And stop using the word "quite" so much.

harley: Thanks Harley!

Fwiffo: As from our other conversation I'll bring one of my suggestions here to see what other people think...

This one is based on Brandubh, as played in Ireland in the tenth century. What we know from literary sources and archaeology is that the game was played with a king (branan) and his four men against eight enemies. The most common board size found in Ireland is 7x7, with markings on the centre and the four corners. Poetry tells us that the central square on the board is "Tara", and is one of the "five noble squares fitting for the branan". This, together with the rules of movement and capture we're familiar with from tablut and other related games, is just about enough to put together a working variant:

1. The pieces are laid out on the board in the shape of a cross, with the king in the centre, his four defenders around him, and the attackers forming the outer arms of the cross.

2. The attackers move first, and all pieces move like a chess rook, as in tablut. Only the king can occupy the centre square or the four marked corners.

3. The pieces are captured by surrounding them on two opposite sides, as in tablut. The king, though, is also captured this way. A piece sitting next to a corner square can be captured by placing a piece on the other side of him, so that he's caught between an enemy and the corner square. The central square is different and can't be used in this way.

4. To win, the king must reach a corner square. He loses if he is captured as described in rule 3 above.

This is almost the same as the game Brandubh implemented in Zillions of Games. But that game favours the king too much, so I've adopted the layout from E. MacWhite's 1946 article on Early Irish Games, and allowed the king to be captured on only two sides as in tawlbwrdd. The result seems to be quite a balanced game. It's quite a bit quicker than our tablut, and is more about short term tactics than long term strategy, so it makes quite a good coffee table game. I just need to make myself a set!

Anyone who's interested in the general history of tafl games, of which tablut is one, might like to have a look at my site about the game, which is completely rewritten and uploaded this morning:

http://tafl.snigfarp.karoo.net/tafl/

The highlight of the site is an applet for Java-equipped web browsers, which allows you to play the game against a friend using whatever board layout and rules you select (i.e. a virtual board). I've also put a drawing on there, something in the likeness of an authentic tablut game, as Linnaeus would have witnessed it (based on his own drawings).

cryptoguru: That's an excellent facility there. I especially like the way the interface mimics Brainking.

People with Java-enabled browsers might also be interested in a site I did a while ago:

http://tafl.snigfarp.karoo.net/

It's not as good as the practice pad, as you can't take back moves, and you can't quite do Brainking's rules exactly (the king can always re-enter the central square in my implementation).

The settings required for an approximation of our familiar tablut are:

Layout: Jarlshof I; King strength: weaponless; King's objective: Board edge; First move: King and Defenders; everything else is as set by default.

While it has shortcomings as a practice pad, it does allow you to experiment with other sizes of board and rules. There is also inadequate documentation.

Stormerne: Adding a new restriction to the game, when existing games were played without it, will mean that some of those games could not be replayed under the current rules. This would be an issue especially when new players are learning to play by example, by studying old games. And since my last message I've wondered: what would be done about games in progress under the existing rules? Unless tablut were to be "closed" during the course of those games, the two rules would have to run side by side anyway. I like txaggie's suggestion of having a new variant, as this site does for so many other games.

My preference for raichi is this: if we want to recreate an authentic "raichi", to prevent unsatisfying victories in which the opponent has not noticed the position, I think a more authentic way would be to have the system put the word on the screen somewhere, in the turn of the player who is threatened. The obligation for the threatened player to do something about it is not mentioned in Linnaeus, but having "raichi" in big letters above the playing board when you start your turn (and "check" in chess for that matter) would feel more authentic to me.

Stormerne: one problem I see with this innovation is that previous games will be rendered invalid by the new rules. I also like the fact that the rules on this site, unlike many others, are completely consistent with the contemporary source. Introducing a check-like construct would change the reconstruction we have here into a variant.

Fwiffo wrote: "Do you mean you find the translation of "intercludere...ut" with "prevent...that" incorrect?"

Nothing so intelligent I'm afraid. I simply misunderstood your earlier translation.

Fwiffo: my literal understanding of your initial translation (23 Jan 19:00) is that a piece can be captured by a single enemy against a central square, as some reconstructions have it. This I find undesirable. Your latter correction (23 Jan, 19:06) makes it sound like a clarification rather than a rule, and brings Brain King's rules more into line with Linnaeus's text.

As for draws by repetition, I would be happy enough if a draw was mutually agreed by the players, rather than enforced by the rules. If an unsportsmanlike Swede attempts to win by attrition, making repetitive attacks while refusing the offer of a draw, it's simple enough on this site not to play against him (or her) in future.

Thank you Fwiffo! I'm only a beginner in Latin, but I can make out some of the meaning. Rules 1 to 12 seem to correspond to the ones in English. Rules 13 seems to simply describe the layout of the muscovites, and I can't understand 14 at all yet, but I have a feeling that, like rule 13, it may have been excluded from the English edition simply because it duplicated information already given.

I think rule 11 is there simply to prevent the king from forcing a draw by camping next to the central square. The text of fig.1 is pretty clear; I can't imagine the phrase "no [other] person can be admitted" being used to mean "no [other] person can pass", with or without the "other". As I don't have access to the original Swedish, nor the knowledge to understand it, I don't feel in a position to doubt Smith's translation.

No, as in these variants, the corners act as if there was a piece of the currently moving player sat upon them, so you can capture a piece by surrounding it between one of your own pieces and a corner square. There is evidence that, on some boards (the alea evangelii 19x19 board), there were fixed men attached to the corner squares. In others (like the Ballinderry 7x7 board) the corners had more subtle markings on them.

Given that some variants reverse the colours of the pieces, or even use colours like silver and gold, I tend to use the term "attackers" for those attempting to capture the king, and "defenders" for those who try to stop aid his escape.

The reason I say twelve pieces is that the corner squares are often used as "hostile squares" for capturing, so eight pieces standing adjacent to them would be vulnerable for attacking. With twelve pieces, though, you could have a formation of pieces in squares A3, B2, C1 and the corresponding squares in the other corners. These would be invulnerable.

Allowing the king to win only by reaching a corner square actually seems to make the game unbalanced in favour of the attackers. Only twelve pieces are needed to completely seal off the corners. With another three pieces left the attacker can create a line to sweep across the board, gradually constricting the king's forces. There appears to be no way out of this for the king.

I think that's nothing more than a quaint way of expressing things in Georgian English. If it were to be read as if it were current English, then the ONLY way for a man to be taken is to deliberately move between two others. I'd have trouble believing that that was the case. Unfortunately I've never been able to get access to a copy of the original Swedish, and I wouldn't understand it if I saw it. It might well be that some of these ambiguities appeared with the translation.

I think Linnaeus might well have thought of it in chess terms. The king is never taken off the board; the game ends when he's captured. There's also the technicality that four men are needed to capture him.

As promised on 9th August:

From "Lachesis Lapponica", Linnaeus ed. Smith, 1811, vol. ii. pp. 55-58. [My notes are in square brackets. Linnaeus's unusual notation has been converted to the standard we all know.]

The game called TABLUT is played with a checkered board, and twenty-five pieces, or men, in the following manner. [There follows a diagram of the board, and diagrams of the three types of piece.]

Fig. 1. is the king, whose station is in the central square or royal castle, called KONOKIS by the Laplanders, to which no other person can be admitted.

Fig. 2. represents one of the eight Swedes his subjects, who, at the commencement of the game, are stationed in the eight squares, adjoining to the royal castle. [The squares c5, d5, e3, e4, e6, e7, f5, and g5 are marked on the diagram.]

Fig. 3. is one of the sixteen Muscovites, their adversaries, who occupy the sixteen embroidered squares, situated four together in the middle of each side of the field. [The embroidered squares in the diagram correspond to the familiar attacker's positions.]

The vacant squares may be occupied by any of the pieces in the course of the game.

LAWS

1. Any piece may move from one square to another in a right line, as from d4 to d2, but not corner-wise, or from d4 to c3.

2. It is not allowed to pass over the heads of any other pieces that may be in the way, or to move, for instance, from c4 to c1, in case any were stationed at c3 or c2.

3. If the king should stand in c4, and no other pieces in c3, c2 or c1, he may escape by that road, unless one of the Muscovites immediately gets possesion of one of the squares in question, so as to interrupt him.

4. If the king be able to accomplish this, the contest is at an end.

5. If the king happens to be in c3, and none of his own people or his enemies either in b3, a3, c2 or c1, his exit cannot be prevented.

6. Whenever the person who moves the king perceives that a passage is free, he must call out RAICHI, and if there be two ways open, TUICHU.

7. It is allowable to move ever so far at once, in a right line, if the squares in the way be vacant, as from b4 to b1.

8. The Swedes and the Muscovites take it by turns to move.

9. If any one man gets between two squares occupied by his enemies, he is killed and taken off, except the king, who is not liable to this misfortune.

10. If the king, being in his own square or castle [e5], is encompassed on three sides by his enemies, one of them standing in each of the adjacent squares [e.g. d5, e4 and f5], he may move away by the fourth [e.g. e6]. If one of his own people happens to be in this fourth square, and one of his enemies is in e7 next to it, the soldier thus enclosed between his king and the enemy is killed. If four of the enemy gain position of the four squares adjacent to the castle, thus enclosing the king, he becomes their prisoner.

11. If the king be in e4, with an enemy in each of the adjoining squares d4, f4 and e3, he is likewise taken.

12. Whenever the king is thus taken or imprisoned, the war is over, and the conqueror seizes all the Swedes, the conquered party resigning all the Muscovites that he had taken.

At some point soon I'll be happy to post the laws here (they're well out of copyright). But Linnaeus used a strange notation which he illustrated in his diary: I'll have to convert the notation to the familiar grid references before I post, otherwise some of the laws won't make sense.

I can confirm that Linnaeus's rules allow the king beside his throne to be captured on the remaining 3 sides, without regard to what stands on the opposite side of the throne [Law 11]. Law 10 is an oddity: with the king on e5 and attackers on d5/e6/f5, a defender on e4 can be captured by moving an attacker to e3.