MindSports

A Game of Everything

   

- Acknowledgement

- History

- Territories & walls

- Moves, mergers & captures

- Only doubles can breed

- The four sub-grids

- Loose ends

- An endgame example

- Actual play

- Epilogue

Acknowledgement I love to play Storisende, but that is personal. It's a very long game with a varied scenario that includes many sub-goals and violent clashes as well as puzzle-like manoeuvring, so it's not every player's cup of tea. I especially value the game as an inventor because it is homogeneous and organic. As a consequence of its architecture its formal territorial goal gives rise to an emergent existential sub-goal: to be able to acquire territory you have to stay alive in the first place. What makes it different from other games with that feature is that Storisende shows semi-separate armies taking care of different things. The terrain is made up of territories that are separated by 'the Wall'. The fights that take place up on the Wall are crucial for securing territory down below. Losing them is sure to put you in a predicament. Inventing the game was a process of imagination that took place without so much as touching a checker, and the different roles that emerged were unforeseen and unintended. It was very hard to think the whole game through, especially the endgame. I never could quite make that out during the process itself, but I trusted the outcome.

A short history The ancestry of Storisende includes Sid Sackson's game Focus and my own game Mu. Mu behaves incredibly complex, yet completely logical, "like a hybrid between a boardgame and a pinball machine". But it scares people. The thought of replacing its volatile growth and movement protocol by something more digestible, had been lingering as a loose end for a while. And when I finally and unsuccessfully decided to call it a day in 2018, it was for all I knew the only loose end. Addressing it meant treading on familiar ground and it was remarkably easy. The key thought turned out to be "only doubles can breed" - we'll come to that. But being less complex than Mu didn't mean it was simple. At the BGG Abstract Games Forum I at times came under the impression that trying to misinterpret the rules had become a sport in itself. Not a bad thing of course because it all was done with the best of intentions. But for the sake of this article I hope that emphasis on the overall flow of the game will make digesting the rules a bit easier. Territories and the Wall Storisende is a layered game, not only metaphorically but quite literally. Its board is a grid of hexagons each of which is initially covered with a tile. Such tiles can be removed to reveil the territory underneath, or be replaced to become part of the Wall.
	Here's an empty tiled board, in this case a hexagonal one but in principle any sufficiently large board of any shape will do, including modular boards.
	Here some tiles have been removed to reveil seven small territories. Territory, counted by the number of cells, is the game's formal goal. At the game's end you own a territory if you're the only colour present there. Now imagine that each of these territories would expand by having its border tiles removed, but always leaving in place those tiles of which the removal would result in merging territories.
	The result would be something like the diagram on the left. The tiled board, with territories that emerge but never merge, divided by a naturally growing wall, is Mu's legacy. You can get killed anywhere, but on the Wall it's more serious. Without men on the Wall its hard to claim territory and impossible to move between territories. Armies are semi-separate: you can never get up on the Wall from down below, but you may always jump down from the Wall into the territories.

Let's now turn to Storisende's other ancestor, Sid Sackson's game Focus, of which it employs the way pieces move, but not their composition or the way they capture. Moves and captures The pieces in Storisende are stacks of arbitrary height made up of like coloured single checkers. A single is a stack height one. All moves are straight moves in one of the board's six main directions. Stacks may split and must always move a distance that is precisely equal to the number of men contained in them. You can move a stack of five checkers as a whole over a distance of five cells, but you can also move say the top three checkers over a distance of three cells, or a single top checker by one step. Between departure and landing Stacks that depart from a cell of the Wall may move over any kind of cell, whether vacant or occupied. Stacks that depart from tiled cells or territory cells may not move over a cell of the Wall unless that cell is occupied by a friendly piece. On landing Stacks that depart from a cell of the Wall may land on any kind of cell, whether vacant or occupied. Stacks that depart from tiled cells or territory cells may land on tiled cells or territory cells, but not on a cell of the Wall. Merging and capturing Stacks that land on a like coloured stack merge with it. Stacks that land on an opponent's stack remove the latter from the board regardless of size, and take its place. Only doubles can breed One issue has been left undisclosed: how to get all those pieces in the first place. It was the problem I faced from the start: how to simplify the growth process that makes Mu so scary. And what turned out to be the key thought hit me early on: only doubles can breed. The thought that you need a couple to get offspring is universal so it fitted nicely. But it was more than a nice metaphor, it had beautiful consequences that soon became apparent. If a stack vacates a tile, then this tile turns into a cell of the Wall or a territory cell, depending on whether or not existing territories would merge by its disappearance. If and only if a tile is vacated by a double, it will grow one new single on the vacated cell, regardless of whether the cell turns into a territory cell or a cell of the Wall. The four sub-grids Wall cells are characterised by the fact that their removal would merge territories, but for visual clarity they also have a different colour. To vacate a tile means moving the whole stack that occupies it. You can only grow by vacating tiles with doubles and in the beginning you'll naturally want to grow as fast as possible, so you'll want to hop around with doubles. At that point in the invention process an interesting and initially unforeseen topological consequence arose, that now pervades the opening and the middle game. The significance of four sub-grids In the image on the left, the hexgrid has been divided into four sub-grids. A double is a two-cell hopper so it can never leave the subgrid it starts on. Thus a 'single double' can never by itself expand a territory or create a wall. You need to get doubles on the other sub-grids to do that. If you want to maximise growth it is therefore necessary to switch to other sub-grids, which of course requires time and manoeuvring. Loose ends The pie rule The game starts with a pie rule: one player places 2 to 5 checkers on the board, stacked or single in any configuration of his choice. Next the other player chooses whether to play with those pieces, in which case it is the opponent's turn to place the same number of checkers, or against them, in which case it is his turn. Note: with a pie of 4 or 5 checkers it is possible to start with doubles on two different sub-grids. Counting The game ends after both players pass on successive turns. The winner then is the player who controls the most territory. Draws are possible. Territory is counted by the number of cells inside it. Tiled cells, if any are left, turn to established territory before counting. They never turn into wall cells even if they would have if the game were still on. Pieces that may still occupy them remain in place. Pieces on the Wall don't control territory, only pieces inside a territory do. For a player to 'control' a territory it is necessary and sufficient to be the only colour inside that territory. One man is enough for that. Cells in territories that are not occupied, or are occupied by both players in a stalemate situation, do not count for either player. The scenario A game of Storisende has very different phases that are not strictly separated, yet follow a basic pattern. I'll try to describe it in the usual terms. The opening This is the phase in which you want to grow as fast as possible. But after a few games you've learned the importance of men on the Wall, so you want more than anything to grow on the Wall. That means creating as many small territories as you can as fast as you can and occupying and vacating the tiled cells between them with doubles. That sounds logical but it implies being present on preferably all four sub-grids. It takes time and manoeuvring and accurate observation to do that because newborn singles are often left unprotected. So they may have to move and since merging singles to create doubles on new sub-grids is required anyway, there are opportunities to combine the necessary with the intended. Another danger are unprotected newborns on the Wall. If a single on an isolated wall cell is unprotected, it may be attacked by a double on a wall cell two steps away, or a triple three steps away. That means at best that it is forced to vacate the cell, reducing the wall army by one. You don't want to be forced to do that, you want to force the opponent to do that. The usual pattern is that the board initially gets divided in two areas with most of the interaction along the periphery, with local threats back and forth and with both sides bent on growth. That division may alter considerably in the course of a game: the territory you create may not in the end be the territory that you control. The middle game In this phase territories grow both in number and in size, but the focus on growth on the Wall remains throughout, leading to a fair number of one or two-cell territories that are not only small but hard to defend. They have a high price and little yield. In a balanced endgame they may eventually be of interest, but in this phase the focus should be on the control of larger or potentially larger territories. On the Wall there is ample room for threats, captures and exchanges, and isolated stragglers make a great target. Such fights are necessary because you can't control territory with an untouchable enemy above you. It means that numerical prevalence on the Wall is crucially important. Games are won and lost up there. The endgame Till now the game has been hot, with players eager to move, but that may change in the endgame. That is, if there is one. The outcome of any game is usually clear before the actual end, so there may not be an endgame. But if there is, it will involve clear plans by at least one side, plans that may take time by the nature of the move protocol and the lay of the land. In such cases the game is hot for one player and neither hot nor cold for the other, who may or may not move while awaiting the execution of the opponent's plan. I have a clear example of such a position right here. An endgame example It's the position after Red's 101th move in this beginners game, played on a compact board made up of ten 7-cell modules. Red has won the fight on the Wall but Purple has provisionally acquired more territory, 20 points to 10. Purple can still grow a man in the northwest but he cannot move from one territory to another anymore because his last single man on the Wall is blocked by Red. The top tiled cells should not be vacated both, because that would split and reduce the 4-cell territory. Moving the single on the tiled cell in the center will split a 6-cell territory and allow control over only one part, so that single is stuck completely. If Red can conquer the 7-cell terrtory on the right, he will win 17 to 13, so Purple will have to jump in there if he's forced off the Wall, to strengthen its defense. And he will be forced off. Six moves later Red has managed to aim the bottom 3-stack at the cell that Purple occupied, forcing him off the Wall. Purple has grown one useless man in the northwest. Successfully conquering the 7-cell territory is now less certain than conquering the four cells bottom right and the two singles at the bottom, because the four singles Red has left on the Wall are precisely enough to do it. Purple cannot prevent an exchange in the 4-cell area and Red has taken control. The other two red singles have made it to the defenseless purple pair at the bottom, eliminating them and taking control. Thus the initial 20-10 score has been turned around to 14-16. There has not been much interaction in this endgame, but a lot of drama. Storisende terminates after two successive passes. Notice that the losing side can keep moving to avoid termination. In that case '3-fold' - three times the same position with the same player to move - will guarantee termination. If a losing side keeps moving and the winning side keeps passing, then it may take ten or ten thousand moves to reach 3-fold, but it is inevitably reached at some point. For the losing side that may be an issue worth considering. After termination the score is counted as usual. Of course the game may end in a draw. Actual play If you want to make a physical set, it's better to use round bi-coloured tiles with a dark side for the wall. Hexagonal tiles never quite line up with the required orientation. Here's my own set, made up of fourteen 7-cell modules. It doubles as a way to play Mu. It's all for a hypothetical world, because in this one I hardly ever play anything over the board anymore. You can find and play Storisende at MindSports. The app looks more or less the same as my physical set. Then there's Stephen Tavener's wonderful generic java based game engine AiAi with the nice display on the left. Real brick walls, figurines instead of stacks and doubles as a couple holding hands. But more importantly: it plays a strong game and it is more than likely to beat you several times before you can put up a real fight. You can also play Storisende at Abstract Play. A great implementation!   Epilogue When I imagined Storisende I merely wanted to simplify the implementation of the idea that led to Mu so its behaviour took me by surprise. Many prominent goals that are to be found in the realm of abstract games are featured as sub-goals in the game, most notably elimination but also connection, race, blockade, invasion and even configuration. That is the reason for the sub-title "A Game of Everything". Its scenario of very distinct phases, the chess like interaction of its pieces and the sheer length of a game show a great potential for narrative. It would have made a great last invention. But my mind has a mind of its own so that didn't quite work out. Enschede, the Netherlands, October/November 2021, christian freeling