## Hnefatafl Internet Championship

### Re: Hnefatafl Internet Championship

Hi,

I want to talk here about rules. Actually not the rules of the game, but the rules of the championship. If the game popularity increase (and I hope so) we can't continue to have Hnefatafl Internet Champions like this. Last time we had 13 participants => 24 games per player. Some games ended via timeout early, but still 20 games are pretty much. I have just one day per move and if I have a busy day I have just half an hour for all my moves. 30 mins/20 moves=1.5 minutes. Shame on me, but there're a lot of position in which I can't make reasonable moves in 2 minutes.

My suggestion:

We divide all players in groups of 6 to 8 players.

We make rounds of with int(participants/8) groups. The groups sizes are hold as equal as possible (Ex. with 20 players we make a group with 6 and two groups with 7 players). The players are totally divided randomly over the groups. In each group each player play against each player (this are at most 14 games).

After each round the better half of each group (rounded up). If the number of players would reach 9, 10, 11 or 17 we need withdraws. 3 withdraws if we would reach 9, two if we would reach 10 and one if we would reach 11 or 17 players. The best two players of the last group fight afterward a finale (two games - one with each color).

This would lead to a longer tournament (like three rounds) instead of just one, but with moves of higher quality and a games that will be decided through skill and not through spent time. It's not just a small tournament, but a big event.

Example: 13 players (like we had last time): They would be randomly paired into two groups of 7 and 6 players. 4+3 players advance to the next (last) round.

19 players: Paired into 3 groups (randomly) with 7, 6 and 6 players. 4+3+3+(2 withdraws) advance to the next round. Randomly split12 players in two groups of six players 3+3 players advance into the last round.

27 players. Paired into 3 groups (randomly) with 7, 7, 7 and 6 players. 4+4+4+3 players advance. 15 players are split in groups of 8 and 7 players. 4+4 players advance into the last round.

We should also determinate how we sort players. We don't know what to do if we have two players with the same amount of points. Without further thinking I would suggest the following:

1. points (he archived this round)

2. direct comparison (skip this for withdraws)

3. Sonneborn-Berger (just from this round)

4. coin flip (shouldn't happen anytime)

If the finale is a draw (one win for each player) the direct comparison of the last round is used. If that don't lead to a result they play again two games. If there isn't a winner, they play again two times. Afterward I would suggest a coin flip (so we don't need half a year for the whole stuff).

It might be also interesting to choose just one player from the candidates tournament and let him (or her) play against the current world champion.

Just some ideas. I'd be happy to get comments about it. Of course it would be some afford to organize such an event, but it's much more interesting than just a "normal" tournament at this site.

I can offer help with coding/organizing if that sounds also interesting for you.

I want to talk here about rules. Actually not the rules of the game, but the rules of the championship. If the game popularity increase (and I hope so) we can't continue to have Hnefatafl Internet Champions like this. Last time we had 13 participants => 24 games per player. Some games ended via timeout early, but still 20 games are pretty much. I have just one day per move and if I have a busy day I have just half an hour for all my moves. 30 mins/20 moves=1.5 minutes. Shame on me, but there're a lot of position in which I can't make reasonable moves in 2 minutes.

My suggestion:

We divide all players in groups of 6 to 8 players.

We make rounds of with int(participants/8) groups. The groups sizes are hold as equal as possible (Ex. with 20 players we make a group with 6 and two groups with 7 players). The players are totally divided randomly over the groups. In each group each player play against each player (this are at most 14 games).

After each round the better half of each group (rounded up). If the number of players would reach 9, 10, 11 or 17 we need withdraws. 3 withdraws if we would reach 9, two if we would reach 10 and one if we would reach 11 or 17 players. The best two players of the last group fight afterward a finale (two games - one with each color).

This would lead to a longer tournament (like three rounds) instead of just one, but with moves of higher quality and a games that will be decided through skill and not through spent time. It's not just a small tournament, but a big event.

Example: 13 players (like we had last time): They would be randomly paired into two groups of 7 and 6 players. 4+3 players advance to the next (last) round.

19 players: Paired into 3 groups (randomly) with 7, 6 and 6 players. 4+3+3+(2 withdraws) advance to the next round. Randomly split12 players in two groups of six players 3+3 players advance into the last round.

27 players. Paired into 3 groups (randomly) with 7, 7, 7 and 6 players. 4+4+4+3 players advance. 15 players are split in groups of 8 and 7 players. 4+4 players advance into the last round.

We should also determinate how we sort players. We don't know what to do if we have two players with the same amount of points. Without further thinking I would suggest the following:

1. points (he archived this round)

2. direct comparison (skip this for withdraws)

3. Sonneborn-Berger (just from this round)

4. coin flip (shouldn't happen anytime)

If the finale is a draw (one win for each player) the direct comparison of the last round is used. If that don't lead to a result they play again two games. If there isn't a winner, they play again two times. Afterward I would suggest a coin flip (so we don't need half a year for the whole stuff).

It might be also interesting to choose just one player from the candidates tournament and let him (or her) play against the current world champion.

Just some ideas. I'd be happy to get comments about it. Of course it would be some afford to organize such an event, but it's much more interesting than just a "normal" tournament at this site.

I can offer help with coding/organizing if that sounds also interesting for you.

### Re: Hnefatafl Internet Championship

**The championship schedule 2014 could be:**

All August participants sign up for the championship.

September 1st start of first round.

About October 1st start of second round.

### Re: Hnefatafl Internet Championship

With that suggestion you take the advantage of the 20 day-limit for the final away again. If I loose first round (10-day limit) through a time out that might disqualify me to reach the top. My idea was to get the best players to the top. "I lost through timeout against a strong player in the first round, so I loose the third round." I hope my suggestion also promote creative play, because you have the opportunity to astonish good players with new variants at early rounds, without taking to much risk.

Anyways I like a good reason for players to play against each other. If we are lucky we can get a lot of excellent games during just one tournament.

What about sign up in July, start the first one 1st August and don't set a date for the second round now, but start it, when the first one is finished?

Anyways I like a good reason for players to play against each other. If we are lucky we can get a lot of excellent games during just one tournament.

What about sign up in July, start the first one 1st August and don't set a date for the second round now, but start it, when the first one is finished?

### Re: Hnefatafl Internet Championship

Perhaps just as well go the whole length and use a 20 day-limit for all games of a championship tournament.nath wrote:With that suggestion you take the advantage of the 20 day-limit for the final away again.

If matches which were carried out already in the first round, were to be repeated in the second round, this would be an asymmetry; you'd have two matches against some strong opponents (who accidentally shared your group in the first round) but only one match against other strong opponents.

The only worry is: summer holidays. If a strong player has holidays in August, he's lost for the championship tournament, which would be most unfortunate.nath wrote:What about sign up in July, start the first one 1st August

### Re: Hnefatafl Internet Championship

That double the needed time. If we have to many participants, that take a lot of time.Hagbard wrote:Perhaps just as well go the whole length and use a 20 day-limit for all games of a championship tournament.

Where is the problem? I don't want to compare players. A single game can be never used to say who's the better player in general. Where is the problem if we have more than one game between players? We have some preliminary rounds with certain groups of 6 to 8 ppl. Comparing the old system each group is a "tournament" itself. Each of these rounds are just used to sort weak players out. To get candidates for the final group of 8 players. And they a tournament as known before. Each round offer totally new chances to all players. It doesn't matter what kind of mistakes he made before - he get a new start. We are much faster with this system and it's very easy, because we don't have to look at old rounds, what would be very confusing, even for players that are used to organize tournaments. I don't want to fill in a matrix - I want to sort out good candidates. I don't care about any kind of matrix.Hagbard wrote:If matches which were carried out already in the first round, were to be repeated in the second round, this would be an asymmetry; you'd have two matches against some strong opponents (who accidentally shared your group in the first round) but only one match against other strong opponents.

That's the reason we can use 10 days for the early round. If the players belongs to the upper players, he will not get into the lower half of a group, even if he loose one or two games through a timeout.

OK. I'm never at vacation at this time (I played last tournament from vacation), but you're probably right. So it should be like that:Hagbard wrote: The only worry is: summer holidays. If a strong player has holidays in August, he's lost for the championship tournament, which would be most unfortunate.

"All August participants sign up for the championship. September 1st start of first round. The second round start, when the first round ended."

### Re: Hnefatafl Internet Championship

So, perhaps something like this:

First round groups of about 8 players, 10 moves per 10 days. Players are evenly distributed with strong and weak players in each group.

Best half of each group continues to next round.

...

Last round one group of about 8 players, 10 moves per 20 days.

First round groups of about 8 players, 10 moves per 10 days. Players are evenly distributed with strong and weak players in each group.

Best half of each group continues to next round.

...

Last round one group of about 8 players, 10 moves per 20 days.

### Re: Hnefatafl Internet Championship

That's nearly my fist suggestion. Just the algorithm of contributing the players to the groups differs. We should keep the algorithm clean, but you might be right that it is better if we use the rating for distributing the players over the groups.

I suggest that kind of creating the rank:

I suggest that kind of creating the rank:

1. points (he archived this round)

2. direct comparison (skip this for withdraws)

3. Sonneborn-Berger (just from this round)

4. coin flip (shouldn't happen anytime)

### Re: Hnefatafl Internet Championship

Draft for the 2014 internet Championship tournament.

An average game lenght is 60 moves. If a tournament takes 3 rounds and 2 days per move is allowed, then the whole tournament could last for more than a year! To avoid this, the tournament could start with preliminary rounds of only one day per move.

24 participants.

First preliminary round:

4 groups of each 5 players. One day per move. 2 winners from each group continue to the next round. (4 highest rated players do not participate here).

Left are now 12 participants.

Second preliminary round:

2 groups of each 5 players. One day per move. 3 winners from each group continue to the next round. (2 highest rated players do not participate here).

Left are now 8 participants.

Final round:

1 group of 8 players. Two days per move.

Same games are not repeated over again, that is, if player A already met player B in an earlier round, then the result is transferred to the following rounds.

24 participants.

Preliminary round:

4 groups of each 6 players. One day per move. 2 winners from each group continue to the next round.

Left are now 8 participants.

Final round:

1 group of 8 players. Two days per move.

**Rules:**Copenhagen Hnefatafl 11x11.**Sign up:**all August. There will be a facility for signing up for the tournament.**Start:**September.An average game lenght is 60 moves. If a tournament takes 3 rounds and 2 days per move is allowed, then the whole tournament could last for more than a year! To avoid this, the tournament could start with preliminary rounds of only one day per move.

**Example:**24 participants.

First preliminary round:

4 groups of each 5 players. One day per move. 2 winners from each group continue to the next round. (4 highest rated players do not participate here).

Left are now 12 participants.

Second preliminary round:

2 groups of each 5 players. One day per move. 3 winners from each group continue to the next round. (2 highest rated players do not participate here).

Left are now 8 participants.

Final round:

1 group of 8 players. Two days per move.

Same games are not repeated over again, that is, if player A already met player B in an earlier round, then the result is transferred to the following rounds.

**Example 2:**24 participants.

Preliminary round:

4 groups of each 6 players. One day per move. 2 winners from each group continue to the next round.

Left are now 8 participants.

Final round:

1 group of 8 players. Two days per move.

### Re: Hnefatafl Internet Championship

If players in rating order are A, B, C, etc., the grouping is

1 2 3 4

A B C D

E F G H

I J K L

. . . .

. . . .

In the preliminary games with one day per move, player A would pay special attention to games against players E and I, because most likely one of these players proceeds to the final.

In the final with two days per move, player A meets the strongest players B, C and D for the first time.

Score: 1 for win, 0 for loss, 0.5 for draw.

If the final ends up with a tie fx. between players A and B, but A won his games against B, then A wins the tournament. If the games A-against-B were also a draw, then A and B take new rounds until one wins.

1 2 3 4

A B C D

E F G H

I J K L

. . . .

. . . .

In the preliminary games with one day per move, player A would pay special attention to games against players E and I, because most likely one of these players proceeds to the final.

In the final with two days per move, player A meets the strongest players B, C and D for the first time.

Score: 1 for win, 0 for loss, 0.5 for draw.

If the final ends up with a tie fx. between players A and B, but A won his games against B, then A wins the tournament. If the games A-against-B were also a draw, then A and B take new rounds until one wins.

### Re: Hnefatafl Internet Championship

Even in the preliminary rounds? If we are unlucky player

A, E, I are in the same group, 2 players should advance, but player E and I both got for example 8 points and both won the black game against each other.

I suggested this

You don't know at all what players will be in the final. You can't judge a player from the rating. I recommend to make at least at the final round new games.

In general I wouldn't mind to participate more often against single players. Of course it's boring to face the same player again each round. But we get interesting games and aspects from playing more games against strong players. Note: only strong players make more games than necessary: the lower half drop out after the first round. I would appreciate more games between strong players. We're lacking of that currently. Such a tournament could be a good opportunity for that.

A, E, I are in the same group, 2 players should advance, but player E and I both got for example 8 points and both won the black game against each other.

I suggested this

You just suggested 1&2. In the final round, I agree with you to make further head up games. But in the preliminary round we shouldn't loose to much time - even if have to use a coin.1. points (he archived this round)

2. direct comparison (skip this for withdraws)

3. Sonneborn-Berger (just from this round)

4. coin flip (shouldn't happen anytime)

You don't know at all what players will be in the final. You can't judge a player from the rating. I recommend to make at least at the final round new games.

In general I wouldn't mind to participate more often against single players. Of course it's boring to face the same player again each round. But we get interesting games and aspects from playing more games against strong players. Note: only strong players make more games than necessary: the lower half drop out after the first round. I would appreciate more games between strong players. We're lacking of that currently. Such a tournament could be a good opportunity for that.