In Othello and its variants 'lines' are captured using the custodian method, that is: the act of completing an enclosure of a straight unboken line of opponent's stones at both ends effectuates the capture. The captured line(s) are reversed to show the captor's color.
Othellonian capture in Go variants unites both principles by enclosing groups and reversing them if captured. The most important consequence is that othelloanian capture does not lead to cycles, but at the price of not creating eyespace. It makes 'life' harder to come by.
Long ago, a century or thereabouts, I invented Medusa and Lotus. I applied othelloanian capture mainly to simplify the games' stucture. It was all rather intuitive and I had no trouble getting life into the groups because of an additional condition creating it: the 'rosette'. The rosette was an invention of Mark Berger who in the seventies got the idea to play Go on the triple contacts of a hexhex grid.
|His first idea was to simply apply the rules of Go and see how it worked out. As it turned out, regular concepts like 'ko' and 'seki' remained intact, but there was a big difference if a group was in 'atari', that is: if it had just one liberty left, like the white stone.|
In Go a point has four liberties and extending from a group in atari may increase the number of its liberties by 2. In Rosette an extension increases that number at most by 1, and this one is consequently taken to keep the group in atari.
Bummer. But Mark had a brilliantly simple solution: a group containing a 'rosette' - six like colored stones around a basic hexagon - would live unconditionally.
The rosette fitted Medusa and Lotus so well that I never even thought about the rule as 'artificial', even though it was 'wearing its intent on its sleeve' as a fellow designer once put it. So I never gave much thought to othelloanian capture 'in the abstract'. Much later I encountered it again, occasionally, in the assorted Go variants that were competing for attention on the web. Then came Sygo and with it a number of comments of inventors of other variants employing the method, so I thought it would be a good plan to give a provisional summary. Moreover, an article in Sensei's Library called 'Reversi Go' provided an incentive for a more general approach.
Sensei's Library - Reversi Go
The author illuminates the basic differences between traditional and othelloanian capture on a square grid, and considers the different consequences if suicidal moves are allowed or not, concluding in favor of the latter. One quote:
"Large and empty eye-spaces now also start to look very killable because capturing stones does not help to form eyes (but rather takes them away). Because of this, not getting surrounded would seem even more imperative than in normal Go. At the same time, actually capturing stones might be less attractive (unless as endgame play) because it does not form eyes, and thus capturing a stone solidly will not bring the strength to a weak group that it once did."
Required: an additional life insurance - or is it?
To generalize and extrapolate this: the method of capture appears to require an additional life insurance to make it work properly. A legitimation is that the mechanism becomes 'cycle free' and thus avoids any and all of the related rule problems that are vehemently denied and discussed, at the same time, in Go circles. The means to get a sufficient degree of life insurance differ. Till now we have:
- The "Rosette", as applied in Medusa and Lotus. It looks and feels natural on a hexgrid but cannot simply be extended to the square grid. In terms of balance it fits the general strategy and tactics like a glove.
- The "Eye opener", as applied in Goncrete, invented by Luis Bolaños Mures. Basically it forces (rather than entitles) the captor to remove one stone of every group that is flipped in a capture, so long as the removal does not split the group that resulted from the capture. Like the rosette, the rule is clearly a means to an end, but it is simple and effective and may soon feel 'natural' in actual play. In terms of balance it appears to fit the general strategy and tactics exellently, as far as I can see.
- "Promotion to life", as applied in Implo, invented by Luis Bolaños Mures.
- "True liberties versus false liberties", a concept applied in Loose, invented by Luis Bolaños Mures.
- The "Safe Zone", as applied in Rin, invented by Zhen Wang. The game is closely related to Kropki, aka Dots, a fairly recent pencil & paper game with an obscure origin, probably in Eastern Europe or Russia. In Rin a group is immune from capture if it touches the edge of the board. Like the previous safety mechanisms, the conditions seem to fit the general mechanism of movement and capture fairly seamlessly.
- The Symple move protocol, as applied in Sygo. This one doesn't 'wear its intent on its sleeve' for a simple reason: it wasn't intended as such. The move protocol came with Symple, as an essential implementation of a theme called 'group penalty'. The first implementation was Star, invented by Craige Schensted.
The protocol isn't game specific and has an embedded balancing system. Sygo was the result of its application to Go and its affinity with othelloanian capture. The protocol, by allowing for multiple moves per turn, implicitly and unintentionally provides the 'life insurance' that the previous games had to add explicitly. It's not to say that getting groups alive in Sygo is easy, but that the tactics required to do so are in alignment with strategy.
So an additional life insurance isn't required after all, and if 'no rule required' is better than even the smartest rule, then Sygo offers the best merger of 'Go' and 'othelloanian capture' so far.