General rules

Tafl rules
User avatar
Hagbard
Posts: 710
Joined: Sat Mar 14, 2015 6:07 pm
Location: Copenhagen
Contact:

A comment about testing of rules

Post by Hagbard » Sat Aug 20, 2011 2:27 pm

There is in Lejre in Denmark a "Centre for Historical-Archaeological Research and Communication". Here archaeologists do live experiments like fx. building iron age houses as authentically as possible, families live iron age lives in the houses, and one day a house is burned down and left in ruins for many years, to be excavated after even many more years and compared to real iron age house excavations. The same thing is done at the Viking Ship Museum in Roskilde, building as authentic as possible Viking ships and crossing the North Sea with a hundred men.

I think we are doing a bit of the same here with the historical Hnefatafl game. A group of experienced players try out in live games various variations and combinations of game rules known from historical sources, sometimes clearly, sometimes not completely clearly described. If a rule set turns out to work really well in such skilled games, we are on the right track. Whereas rule sets which does not work really well, are probably not valid (but can have been used in "unskilled" games between players of unequal strength).

Imagine that when a Viking had some spare time in the evening, he would certainly not turn on the TV like modern man does - he would go visit some neighbours, keep up the good relations so that they'd not kill him but he'd have their friendly support in serious situations, he'd have a good talk, tell and listen to good stories, and they'd have a good game or two of Hnefatafl or other board games for pastime.

Now, these people, using this board game for hundreds upon hundreds of years, would not come out with a rules set that doesn't work!

Neil N. Peterson in his student's research paper on Hnefatafl from 2001 gives many good considerations on the Hnefatafl rules. He and a team (fellow students?) did a series of test games using various rules variants, but an important source of error was that the test players were not experienced. Another source of error is that the team was only a few players, and the testing of variants amounted to just 48 games altogether.

aagenielsen.dk also gives some considerations on the Hnefatafl rules.
Last edited by Hagbard on Sat Dec 14, 2024 3:19 pm, edited 1 time in total.

User avatar
Hagbard
Posts: 710
Joined: Sat Mar 14, 2015 6:07 pm
Location: Copenhagen
Contact:

making a tafl set

Post by Hagbard » Sun Sep 30, 2012 3:32 pm

skallatorc wrote:

https://docs.google.com/open?id=0B38L8n ... URvQlVVR0U

Over the the week I have had my heathen brother out from Nebraska and having the opportunity to get some in person Tafl in was something I was NOT going to pass on.

Earlier I hit the dollar store and picked up three dollar chess sets to make a Talf board while at work. I printed out a pretty decent board from boardgeek.com for just such a situation. The day before my brother came out I realized that I had thrown out the board by mistake and would not be back to work for the week. So I improvised a board on the back of a press board clipboard, which turned out rather well.

I then dug out a box of slingshot steel ball ammo and the Elmer's glue and weighted down the VERY light pieces and was ready to play.

I also found on the board geek site a pocketmod set of rules for Tafl and Tablut and wanted to improve on it. So I redid it and made it for just Tafl and added the rule variants for Copenhagen and Berserk. This is 1.0. I am thinking of redoing it so that it contains just the variant rules since the basic rules can be taught by the host.

crust
Posts: 127
Joined: Sat Mar 21, 2015 9:29 am

Re: Tafl pocket mod with rules variant

Post by crust » Sun Oct 07, 2012 8:23 pm

Great work Skallatorc, lovely set of rules.

I found it necessary (when I was writing out the rules for a tournament) to specify that, although you can capture two or even three warriors in a single move, this did not mean you could capture two or three warriors standing together in a row, just by putting one of your warriors on each end of the row. I "caught" a couple of players doing that in the tournament, and they rightly stated that my rules were not clear on that point, so I had to change them! :oops: :oops: :oops: Maybe they were anticipating the Copenhagen "shieldwall" rule which did not exist at that time. It's bloody difficult writing rules which are unambiguous, but I think you've done a great job.

User avatar
Hagbard
Posts: 710
Joined: Sat Mar 14, 2015 6:07 pm
Location: Copenhagen
Contact:

Re: Balanced 9x9 and 11x11 tafl variants

Post by Hagbard » Sat Dec 08, 2012 6:34 pm

Scandinavian Museums Edge 11x11 rating or not?
arne64 wrote:I don't think Scandinavian Museums Edge 11x11 is balanced! I had the strong perception that in a high level game white has no chance. I think the results only suggest that it's balanced because the most players are very inexperienced in this variant. An example is that many black players have some fear attacking the king directly because in the other tafl variants they usually play it's a bad idea but in this variant the king is often seriously threatened by this move.
crust wrote:I totally agree. If anyone thinks this form is balanced, I'll happily play them for money, as long as I can always play black! I wondered if maybe there were strategies to be discovered, but I do not think that is the case. Even armed, and even with the whole edge of the board as a target, the two-side capture king is just too vulnerable.
The result of the Scandinavian Museums Edge 11x11 tournament is listed above. Counting the whole game archive gives for Scandinavian Museums Edge 11x11: 12 players did 66 games with 30 white wins, 35 black wins, 1 draws. Counting the game archive only for players rated above 1500 gives for Scandinavian Museums Edge 11x11: 9 players did 38 games with 16 white wins, 21 black wins, 1 draws. There's in all counts an overweight in black wins though not so much by the look of it. But there's clearly been a subjective experience from the tournament that with more practical gaming experience the black overweight could increase considerably; so maybe the time has not come yet to rate the Scandinavian Museums Edge 11x11 variant.

Draw by perpetual repetition.
conanlibrarian wrote:I think that "Scand. museums Hnefatafl edge 11x11", and indeed any edge Tafl version, should forbid repetitions and draws completely. Draws are probably needed in corner Tafl, to stop the attackers from blocking all corners too easily, but in edge Tafl there is no such reason for having them, and also no possibility of building edge forts (since you already won then).
Adam wrote:Perpetual board positions are perhaps the only thing some of us would like to see the back of
conanlibrarian wrote:Question is if rules forbidding draws should be included? Since there were not many draws, perhaps this is not a problem, but I have the feeling that people simply did not go for a draw even if they had the chance, instead mostly playing as if draws were not possible. The rules as they stand are susceptible to draw forts and infinite raichi. And the rules are definitely playable without draws.
Adam wrote:Forbidding board repetition removes draw forts unless they are specified in the rules (as in Copenhagen). In order to avoid perpetual check threats one would need to create a rule that allows draw forts, but forbids other forms of perpetual move draw positions. Tricky perhaps?
Fotevikens Museum now forbids perpetual repetition for their Tablut, which is halfway an edge tafl. Dragonheelslair forbids, after a lengthly discussion, perpetual repetition for their Tablut, which is an edge tafl the same as our Sea Battle tafl plus a throne. As Adam points out, there might not be a simple way to forbid perpetual repetition and at the same time allow center fort draws. But since the 9x9 board is too small for center forts, it's not a problem to ban perpetual repetition for all 9x9 edge tafls.

Draw by center fort.
arne64 wrote:what do you think of this idea: White can get only a draw if at least 3 black pieces are killed. So in order to draw white has to do some kind of strong attack.
Hagbard wrote:It looks like a suitable threshold to abandon draws with 0 or 1 black pieces killed and to allow draws with 2 or more black pieces killed?
A minimum-kill rule as suggested by arne64 is important for tournaments to avoid Monstrous Forts and similar early draw abuse. But since such early draws would be very rare and just as spoiling for non-tournament games (friendly games), then why not require the minimum-kill for all games, tournaments as well as friendly games?

So, for a future update? :
Ban perpetual repetition for all 9x9 edge tafls. Minimum-kill for all games: White can get only a draw if at least 2 black pieces are killed.

Adam
Posts: 117
Joined: Fri Mar 20, 2015 1:28 pm

Re: Balanced 9x9 and 11x11 tafl variants

Post by Adam » Sat Dec 08, 2012 8:37 pm

Hagbard wrote: So, for a future update? :
Ban perpetual repetition for all 9x9 edge tafls.
Minimum-kill for all games: White can get only a draw if at least 2 black pieces are killed.
An interesting suggestion. I would like to propose that we keep Copenhagen as it is, simply because it is a rule set designed specifically to allow some draw forts while forbidding others and perpetual moves. And I think it is an excellent variant as it stands. Also berserk functions very well as it is I think.

If Fetlar is going to continue allowing perpetual moves it will be open to abuse, and a weaker player will always be able to force a draw as white with a perpetual threat with only a 'between player agreement' preventing it. If a rule preventing perpetual threats but allowing drawfort board repetitions could be formulated, this would be ideal.

The minimum kill would seem to do away nicely with 'unfair' forts.

So, anyone for a game of copenhagen? ; )

crust
Posts: 127
Joined: Sat Mar 21, 2015 9:29 am

Re: Balanced 9x9 and 11x11 tafl variants

Post by crust » Sun Dec 09, 2012 12:59 pm

This is a deep and complex thread, more a sort of yarn with lots of threads in it. I have re-read it to enjoy again Hagbard's forensic analysis of the statistics, Conanlibrarian's historical perspective and Adam's expertise in game dynamics. What a trio. We've got two main themes here, the scand. edge tafl form, and draw by repetition, which I'll come to in a moment.

Scand 11x11 edge tafl
Hagbard wrote:There's in all counts an overweight in black wins though not so much by the look of it. But there's clearly been a subjective experience from the tournament that with more practical gaming experience the black overweight could increase considerably
The figures for the Scand 11x11 edge tafl are closer to balance than I would have thought, but you're right, my subjective experience is that I feel unbeatable as black, and relatively powerless as white. I won every game when playing as black, and Arne64 won all of his black games except one, where Luigi005 scored a draw by centre fort. Hagbard's black games were all wins except against Arne and me, and I certainly felt I was lucky to win that one.

The best strategy for white seems to be to try to gain a foothold on the edge (near the corner is the only space available) and then try to get the king to join them. But being captured on 2 sides makes the king so vulnerable to check, that he cannot afford to stand next to any enemy pieces, and therefore it is usually easy for black to prevent him from joining up with his comrades at the board edge.

So, I enjoyed the tournament but it seems the variant is not balanced. Perhaps a rule change could improve it? For example, I wonder if having the king captured on 3 sides would be better balanced? It might be worth a try, it's exactly the sort of tinkering that this site is uniquely well-placed to conduct.

I have encountered 3-side capture once before, on the Shetland island of Unst, where they had an Alea Evangelii game set up in their reconstructed longship. According to their Alea rules the four white pieces around the king were elite guards, who could only be taken by being surrounded on three sides. The king was captured from 4 sides, all the other pieces on 2 sides. I played a game through with Peter Kelly, (Adam is right to say that Alea should be played live, preferably in quickplay mode, it's really too long for correspondence format) and the 3-side elites worked pretty well. Somewhere on the island of Unst there must be a clever hnefatafl innovator, but I can't find out who it is at the moment.

Draw by repetition Draws by repetition come in two flavours: perpetual check and draw forts. I think we would all like to get rid of perpetual check, but there is a case for keeping draw forts, apart from the early centre draw forts.
Adam wrote:keep Copenhagen as it is, simply because it is a rule set designed specifically to allow some draw forts while forbidding others and perpetual moves. And I think it is an excellent variant as it stands. Also berserk functions very well as it is I think
Adam wrote:The minimum kill would seem to do away nicely with 'unfair' forts
Well said :D

User avatar
Hagbard
Posts: 710
Joined: Sat Mar 14, 2015 6:07 pm
Location: Copenhagen
Contact:

Re: Balanced 9x9 and 11x11 tafl variants

Post by Hagbard » Mon Dec 10, 2012 10:30 am

Software updated :
Perpetual repetition banned for all 9x9 edge tafls (plus Copenhagen and Berserk).
Minimum-kill for all games: White can get only a draw if at least 2 black pieces are killed.

Update, NB:
Perpetual repetetion now banned in all variants except Fetlar.
Minimum-kill (draw only if 2 black pieces killed) now only used in Fetlar.

User avatar
Hagbard
Posts: 710
Joined: Sat Mar 14, 2015 6:07 pm
Location: Copenhagen
Contact:

Re: News about Scandinavian Museums' Hnefatafl

Post by Hagbard » Fri Dec 21, 2012 10:32 am

crust wrote:It could be worth trying these rules
If Norse America have their own rules, we'll implement them as a variant. I've come to the view that it makes sense to focus on the commercial Hnefatafl game sets. By and large, internet pages on Hnefatafl echo those - somebody met the game in some commercial version, became fascinated by the game and wrote a page about it on the internet (the present site also started that way, based on Skalk 11x11). The commercial game sets for their part are presumably based on academic books and articles and professionals (fx. the Skalk version is).

User avatar
Hagbard
Posts: 710
Joined: Sat Mar 14, 2015 6:07 pm
Location: Copenhagen
Contact:

I dont know

Post by Hagbard » Tue Mar 05, 2013 1:05 am

Chuck Ward wrote:

I was talking with Ric Snead about Fetlar rules who said: There is a wrinkle in their ruleset that allows 2 men to 'camp' on the edge, gumming up the works with no possibility of capture. I have never had a prob with this.

crust
Posts: 127
Joined: Sat Mar 21, 2015 9:29 am

Re: I dont know

Post by crust » Thu Mar 07, 2013 12:57 am

Hello Chuck Not just Fetlar rules - this is a feature of all rule sets except Copenhagen, where the shieldwall rule allows capture of a pair or row of pieces at the edge. In all other variants, two pieces together at the board edge are impregnable. But like you said, I never had a problem with this either.

Post Reply