background Ludii Portal
Home of the Ludii General Game System


Home Games Forum Downloads References Concepts Contribute Tutorials Tournaments World Map Ludemes About

Branching Coral



Category Experimental


Branching Coral is a placement game based on the principle of placing pieces, such that, after placement, every site on the boards still has a connection via empty sites to an empty site at the perimeter. Branching Coral differs from its parent game Brain Coral by having a capture mechanism: After placement, all the neighbouring stones with less than 2 adjacent empty sites are simultaneously captured. The placed piece itself must have 2 adjacent empty spaces before the captures to be a legal placement. These capture conditions assure that the game is finite, and that the ending patterns that form are varied and gracefully branched, without thick regions either of pieces or empty space. Scoring is based on the size of the largest group plus a bonus factor. The standard bonus factor is number of opponent's pieces captured. This works to reward capture and thus encourages breaking up opponent's groups, while remaining straight forward in the need to enlarge your largest group, it allows creating come-back groups, or staging combacks by capturing material. Other bonus options are: Bonus for number of your own groups, serving to encourage you to make one large group and many singletons, to break up your opponent's largest group, and to try to force the opponent to connect small groups into medium-sized groups. This is similar to Brain Coral, with the added tactical element of capture. Bonus for number of opponent's groups; which is less interesting because it encourages players to form large clumps and to break up and limit opponent's group growth, and potentially partial mimicry in play. Playing without any bonus, is similar to opponent group count bonus, but even more straight forward. It leads to clumping, though there still is a need for defence from captures severing connections. In this version the game will normally be decided early because neither player can extend nor break the critical group, and there is not enough space to create a larger one. Thus, this option needs the largest boards. The standard tiebreaker is for the underdog to win based on the largest cascading group size without bonus. If all groups are paired in size with the opponent, last to play wins. For informal play, the tiebreaker can be simplified as last-to-place.


Goal: Score the most points:
-- One point for each stone in one's largest group.
-- Bonus points are also awarded according to the selected option.

The board starts empty.
Players take alternating turns placing one of their stones onto an empty site.
Directly after placement, every site of the board must remain connected to an empty perimeter site via a continuous path of empty sites, and the stone that was placed must be next to any 2 empty sites.

After placement, simultaneously remove all the stones adjacent to the placed stone that are no longer next to 2 empty sites.

There is no passing. The game ends when no more placements can be made.

Players score one point for every stone in their largest group, then, if applicable, score bonus points according to the selected option.

Bonus Options are:
-- Captured opponent's stones
-- Count of one's own groups
-- Count of the opponent's groups
-- No Bonus

The player with the higher score wins. If the scores are tied, the game is decided by the tie-breaker option that was selected before the game:
A: Cascading largest group LOSES -- i.e. negative scores -- This excludes tied pairs; If all groups are paired, last to place.
B: Last player to have placed a stone.


Dale W. Walton

Creation date


Ludeme Description

Branching Coral.lud


Browse all concepts for Branching Coral here.

Similar Games


Brain Coral

The SmallHolder


Veloop (Hex)

Claim Jumpers


Chains of Thought

Veloop (Square)




     Contact Us

lkjh Maastricht University Department of Advanced Computing Sciences (DACS), Paul-Henri Spaaklaan 1, 6229 EN Maastricht, Netherlands Funded by a €2m ERC Consolidator Grant (#771292) from the European Research Council