Copenhagen Hnefatafl

[crust]

There is an issue with the white-is-blocked-and-cannot-move situation.

Thanks for your reply Hagbard. There does seem to be a little confusion here - probably it's me that's confused.

It's true that in most cases, as long as black is not too careless, this type of stalemate draw can be avoided anyway, by carefully leaving a space for white to move into. And it seems a bit unreasonable for white to demand a draw this way, when clearly white is in a losing position.

If we say that white loses if all white pieces are immobilised, then that does lead to more possible winning positions for black. For instance, in Fetlar rules, if white has lost all pieces except the king, we would now say that the king COULD be captured on an edge square, by being immobilised there. This also applies if the king is trapped at the board edge with one or more of his warriors.

I think this is quite a big change in the rules, at least as far as Fetlar is concerned. I'm not saying I'm against it, actually it seems quite reasonable, especially since all it is doing is shortening the game by two or three moves without affecting the result (assuming that black would win anyway, by leaving a moving space and forcing the king away from the edge) ...but we might need a bit of rule-book-updating. It's odd that this is coming up in several games just now!

There's a time stamp to each game, which can identify the game

Of course, I had not thought of using the time stamp! Thanks

[conanlibrarian]

There is also a third option that can be considered - i.e. that the player that can not move have to pass, and the turn go back to the other player again. But in the end, I think a loss for the immobilized player, as in Skalk, is the best option, since it shortens an already decided game. In the same spirit I think that once a player is completely enclosed he should resign (assuming draw forts are not possible or allowed).

[nath]

Hi,

I'm quite new to this game. I know this since one and a half month.
But on the other hand, I'm a little bit experienced with game balancing and not the weakest player.

I only want to leave some comments about 11x11 tafl variants - I'm less experienced at 9x9 boards.

"old" Hnefatafl:
This game seems balanced, but is very boring. I wonder the statistic has not 95 % draws and 5 % black wins. I don't get how to loose with black.
When I play white I usually try to take 2 pieces to be able to draw. I won 6 games with white in the tournament, but most games I felt very unarmed and my opponents give me the win like as present.

Fetlar Hnefatafl:
Much more difficult to play for black. The problem is quite simple - if black reached a good position, most cases white is able to draw. I think it's favor for white.

Copenhagen Hnefatafl:
To be honest: This variant is the main reason I'm still playing this game.It looks very good - I'm currently not able to say which side has an advantage at this rule set.
I think the main problem is still the draw rule:
"8. Perpetual repetitions are forbidden. If the overall board position is repeated three times, the player who maintains the situation ("the threatening player") must find another move to break the repetitions." This rule is very unsharp. We have to judge every single game. Of course there some positions we can easy tell who is the attacking player. For example if the king moves form one side of the board to the other. to attack the corners. But it's more difficult. There was a position I tried to retain my pieces, but in fact I was the one who directly attacked the opponent pieces. I was just preventing me from getting surrounded, but I attacked very unimportant pieces (neither of us cared) while doing that.
That time I played a good alternative, but if you reach a position without such an option?
At other games like Xiangqi this rules make sense. The positions in this games are most times defined by the count of the pieces. Some times one of the weaker pieces is compensated though tactic or strategic advantages (controlling the center of the board, higher development of pieces). Only very few games are lost due to massive tactic attacks. Most times a single piece is won through a tactic attack and the other pieces are traded until the single piece make the huge difference.
But Hnefatafl (specially with Copenhagen rules) is none of these games. We have only one piece what is more important than the other: the king. And you can't thinking of sacrificing him, 'cause this means a loss. But the main problem is that captured more pieces than your opponent means not that you're able to impose pressure. There only very few positions you can trade a lot of pieces. Most games black is able to remain about 18 pieces or so. The threads in this game normally aren't to take pieces. This game is about covering and uncovering corners and edges, not about getting material advantage.
A rule like "If the overall board position is repeated three times, the king's side must find another move to break the repetitions." make more sense to me. White is the one who want to escape.
If we still want to allow draw forts we can add a rule like this: "White draw fort: the king's side may enforce a draw if a white piece has contact with the board edge, a white piece is able to move (they may be the same), and it is impossible for black to capture one of these." But I don't know weather such a rule would be necessary.
Btw: the minimum kill rule doesn't make sense while playing with encirclement.

Further thoughts: We know Unst edge escape is unbalanced in favor for white. What about king hammer but not anvil? The main advantage of an armed king compare to a normal piece is be able to step the black piece without be attacked instantly in revenge. Looking at usual well known pattern like the guillotine confirm the theory that anvil is the stronger ability. It might be favor for black, but I'd like to test this rule set.

[Adam]

Hi,I'm quite new to this game. I know this since one and a half month. But on the other hand, I'm a little bit experienced with game balancing and not the weakest player.

Welcome on board!

Copenhagen Hnefatafl: ... "8. Perpetual repetitions are forbidden. If the overall board position is repeated three times, the player who maintains the situation ("the threatening player") must find another move to break the repetitions." This rule is very unsharp.

People do seem to have an issue with this rule, or its wording, being ambiguous or hard to call. The board repetition part does clear it up though, in that there is always one player who was the first to enter into this repeated board position, and that is what makes him the aggressor, nothing more. It could even have been meant as a 'defensive' move, but if it leads to a board repetition, the rule states that another move should be made.

A rule like "If the overall board position is repeated three times, the king's side must find another move to break the repetitions." make more sense to me. White is the one who want to escape.

That's quite right in practice, but while it is extremely unlikely for black to be the one to start a board repetition, is is possible, so we felt the rule should cover that eventuality.

If we still want to allow draw forts we can add a rule like this: "White draw fort: the king's side may enforce a draw if a white piece has contact with the board edge, a white piece is able to move (they may be the same), and it is impossible for black to capture one of these."

What do you think of the Copenhagen drawfort rule as it stands, allowing only the 'mobile king winning edge fort', (a sort of build your own exit)? This was developed to give white an extra incentive for edge fort building. Copenhagen effectively does away with draw positions, other than those where both players agree the game is going nowhere and neither feels they have an advantage or hope of creating one. The winning edge fort adds some extra leverage options to the endgame that don't result in disappointing draws.

Btw: the minimum kill rule doesn't make sense while playing with encirclement.

Good point!

[conanlibrarian]

I think nath's sugestion for a repetition rule is good and easy to judge. If we want to allow for the rare case where black is the active part in the repetition, I suggest the following rules:

• If the overall board position is repeated three times, the king's side must find an alternative move to break the repetitions.
• If this is not possible without the king's side immediately losing the game on the next move, the attacker's side must break the repetitions.
• If neither the king's side, nor the attacker's side can make an alternative move without immediately losing on the next move, the game is a draw. The king reaching a double lane (or a double lane to a double lane) to the exit (assuming the king can not be captured there) counts as victory with respect to this rule.

But actually I think I like the simple rule that the king's side has to break repetitions better.

[Hagbard]

A game in the championship tournament 2013 reached this deadlock position:



The king reached the edge, but white is not able to complete the solid walls of an edge win fort. At the same time, despite that the fort is not solid, black is not able to break it without losing.

This deadlock situation is not covered well by the present Copenhagen rules. The closest one is rule 8:
"Perpetual repetitions are forbidden. If the overall board position is repeated three times, the player who maintains the situation ("the threatening player") must find another move to break the repetitions."

which handles only overall board repetitions, especially eternal checks, where both parties move back and forth eternally.
And who would "the threatening player" be in the above diagram?

The deadlock situation above is different, in that only one party (white) is confined to a very few moves, whereas the other party (black) has a lot of free movement with many pieces.

I suggest this additional Copenhagen rule to cope with this problem:
Draw forts are not allowed: If white repeats the white board position three times while no piece was captured, black wins.

This takes care of all kinds of draw fort situations (including the early center fort).

[crust]

Draw forts are not allowed: If white repeats the white board position three times while no piece was captured, black wins.

that seems quite clear and simple, also I guess that could probably be added to the computer program. It's a good addition to the Copenhagen rules.

[Hagbard]

Hnefatafl Championship 2013 evaluation.

The Hnefatafl Championship tournament 2013 is completed, this year using the Copenhagen rules for the first time in a championship tournament.
156 games were done by 13 participants from 7 countries on 2 continents.
Congratulations to Nath for the win!

Based on the 156 games, the game balance was found to be +1.30.
(calculated with certain adjustments: timeouted games excluced except for those with an already certain outcome; and games only included as full pairs).

When a game times out or is resigned, it's not always possible to identify exactly how it would have ended, but in the following games the ending was clear:

White won 57 games by simply reaching a corner (in 8 of which using the guillotine technique).
White won 5 games through the white win edge fort.

Black won 21 games by simply catching the king.
Black won 17 games by generally surrounding the white party.
Black won 1 game because white was unable to move.
Black won 1 game because white was trapped in a "draw fort".

Zero games drawed.

In at least one game a row of enemies was captured using the shieldwall rule.


Evaluation of the Copenhagen rules set.

The Copenhagen rules worked very well for the championship games, in my opinion. All rules were used and were demonstrated indispensable, and the outcome was a lot of interesting games. Also the rules set appeared to be complete with no more additions really needed (after the tournament led to a tuning of the perpetual repetitions rule in the case of "draw forts").

The general surrounding of white proved important, as almost half of the black-win games ended this way early instead of the prolonged, slow, inevitable "strangling" of white.

Also the white win edge fort proved to be an important element, leading to 5 direct white wins, but in all games always present as a latent white threat.

In one game Altti demonstrated how the two rules "shieldwall capture" and "white win edge fort" act together. A row of blacks were shieldwall captured, subsequently directly leading to the win edge fort.

And the tournament demonstrated that draws are extremely rare, here zero out of 156 games.

[Adam]

I agree that the Copenhagen rules performed very well once again. I was originally somewhat resistant to the shield wall rule, but it has turned out to be a very important facet of the game, especially in combination with the exit fort, and corner exits. The threat of the shield wall is often used as leverage too.

Is it time to adopt Copenhagen as the official world tafl federation modern hnefatafl tournament rules ?

[Hagbard]

Nath wrote:

We need a set of rules to increase population. And Copenhagen rules are the only choice we have so far. Let them use for official tournaments. By the way: I still claim they are in favor for black...