Page 1 of 1
For a draw in Kalah, one should still award half a point
Posted: 21 January 2026, 19:57
by Trepiastok
For a draw in Kalah, one should still award half a point, as in chess and checkers; otherwise everything looks strange. It’s unclear to whom and for what the point is given, and for a Swiss-system tournament this is unnatural.
За ничью в калахе всё же надо давать пол-очка как в шахматах и шашках, иначе всё выглядит странно. Непонятно кому и за что очко, и для турнира по швейцарской системе это неестественно.
Re: For a draw in Kalah, one should still award half a point
Posted: 30 January 2026, 19:25
by Trepiastok
Oh! A draw has been introduced in Kalah! Many thanks to the site’s developers! We’ll see how the Swiss system works, but in any case, my thanks.
Re: For a draw in Kalah, one should still award half a point
Posted: 30 January 2026, 19:58
by Jellby
Re: For a draw in Kalah, one should still award half a point
Posted: 30 January 2026, 22:25
by Trepiastok
If you look at the history of games between two players, you’ll see that only the wins count.
Lesha: 52 Победы в 75 играх (69%)
Actually, it’s not 69%, because there are also draws.
Re: For a draw in Kalah, one should still award half a point
Posted: 31 January 2026, 12:41
by TakedaYamaku
I would like to give an overall respond also regarding your previous post about why the Swiss System wouldn't work in here.
One of the basic conditions as a draw has already been existing for a long time and has been implemented long before. When the player's win percentages are calculated, only the games that are won or lost are giving consideration. If we look at the draws as a whole we would get the same result.
Regarding the Swiss System I'm not sure what makes you think that in a draw the players are not getting half a point and the players randomly assigned a 1 and a 0. It is just not true. In a case of a tournament using a Swiss System the players will get half a point in case of a draw. Also at the end of the tournaments in case of a tie, the order of players are decided by who played against stronger players the most.
For Arena and Simple games, when there is a draw. The game result is also assigned as a tie. You can already see it in their player stats. But because of their elo differences the site is re-evaluating their elo's based on their current opponents. (higher elo players will drop while lower elo player will rise e.g.)
I'm hoping this will be helpful for clearing some misunderstandings.
Re: For a draw in Kalah, one should still award half a point
Posted: 31 January 2026, 16:38
by Trepiastok
TakedaYamaku wrote: ↑31 January 2026, 12:41
I would like to give an overall respond also regarding your previous post about why the Swiss System wouldn't work in here.
One of the basic conditions as a draw has already been existing for a long time and has been implemented long before. When the player's win percentages are calculated, only the games that are won or lost are giving consideration. If we look at the draws as a whole we would get the same result.
Regarding the Swiss System I'm not sure what makes you think that in a draw the players are not getting half a point and the players randomly assigned a 1 and a 0. It is just not true. In a case of a tournament using a Swiss System the players will get half a point in case of a draw. Also at the end of the tournaments in case of a tie, the order of players are decided by who played against stronger players the most.
For Arena and Simple games, when there is a draw. The game result is also assigned as a tie. You can already see it in their player stats. But because of their elo differences the site is re-evaluating their elo's based on their current opponents. (higher elo players will drop while lower elo player will rise e.g.)
I'm hoping this will be helpful for clearing some misunderstandings.
That’s acceptable.
However, in a Swiss system you can’t play the same opponent twice, and the seeding should be determined, for example, by rating.
And there must be a proper balance between White and Black (i.e. the right to move first). Just a standard Swiss system.
Re: For a draw in Kalah, one should still award half a point
Posted: 01 February 2026, 05:29
by TakedaYamaku
Trepiastok wrote: ↑31 January 2026, 16:38
That’s acceptable.
However, in a Swiss system you can’t play the same opponent twice, and the seeding should be determined, for example, by rating.
And there must be a proper balance between White and Black (i.e. the right to move first). Just a standard Swiss system.
If you're creating a tournament using the Swiss System in bga, you have an option to enable or disable two players playing against
each other more than once. If there is enough player of course you should disable and play as intended. But as I've mentioned
before in Kalah Club, if total number of players is not enough to allow such system. Playing against the same opponent twice is the better option.
About the first round matchings.
Not being fully sure, It seems the site is using a random matching and not considering the ratings. However if there are enough matches in the tournament the initial pairings are not mattered much.
I agree about the point that the players should have an equal number of games going first and second in tournaments for general.
I'm not sure if the site is already accomodating such a system but it's worth while checking.
Re: For a draw in Kalah, one should still award half a point
Posted: 01 February 2026, 11:27
by Trepiastok
I think 14 players are enough for 7 rounds. And it's better to play fewer rounds than twice with the same player.
Re: For a draw in Kalah, one should still award half a point
Posted: 02 February 2026, 10:53
by TakedaYamaku
The way to calculate the number of rounds in a swiss tournament is to "round the number of players in the tournament to the nearest power or two"
4: 2, 8:3, 16:4, 32:5, 64:6 etc.
(In the case of 14 the number or rounds should be 4.)
That's why if necessary, players should play against each other again.
Re: For a draw in Kalah, one should still award half a point
Posted: 02 February 2026, 11:37
by Trepiastok
In practice, it’s simpler. On the Lichess chess site there is a program that can run five rounds with six participants. In reality, this turns into a round-robin, but the players meet only once. The same program also calculates the Buchholz coefficient, performance, and so on.