Recommend
7 
 Thumb up
 Hide
66 Posts
1 , 2 , 3  Next »   | 

Abstract Games» Forums » General

Subject: How to beat AiAi at Storisende rss

Your Tags: Add tags
Popular Tags: [View All]
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

Storisende Rules

I feel that of my games Storisende is the least understood. It is new, I said, but so is every game invented here. To explain its novelty in general terms isn't easy so I feel it might be more effective if it were supported by example. Here's a game between me and AiAi, set at strength 3 and 5 seconds per move. I crush it because I understand the game better. I'll show you how it's done.

But before that I'd like to remind you that stacks go exactly as far as they're high and that only doubles can breed. If a double vacates a previously unvacated cell, it sprouts a new man on that cell. That cell may be part of the territories or part of 'the Wall'.

Since both start with a 'single double' that only has 2-cell jumps, both are initially bound to one of the four sub-grids in this image:


With it you can make on-cell territories. You can't expand a territory with it nor make a wall cell. So you have to arrange the new-borns at some early stage to make doubles on other sub-grids. This takes time at a stage where growth is crucial. Two things should be kept in mind:

- Don't make a double on a cell that is under attack and unprotected. It will cost you the piece or it will cost you precious time.
- Put the new double on a sub-grid where it has convenient routes, or at least as convenient as possible.

The game took less than a hour. Last moves are indicated and move numbers appear at the top of each diagram, yellow and odd for White, black and even for Black. I won because AiAi is 'territory oriented' and if it could realise something, it would realise that territory is a literally small consolation if you're wiped out. Because Storisende starts out existential and territory is a far away promise not to be taken into account all that much just yet.



AiAi (white) - Christian Freeling (black)

09

AiAi opened and I swapped, so AiAi plays white. The move number is 9 because placing the modules also counts as moves. We're on different sub-grids and we both start breeding, avoiding being attacked just yet.

20

Here's my first wall piece. I had to move west because sw is blocked and se gets me stuck. In doing so I attack three white men and offer an exchange with two of the three. White accepts.

22

So we're a bit down on material, get up stand up!

30

White has one man on the Wall and five in the field, Black two on the Wall and five in the field, but Black's double is stuck, so it's time to rearrange.

42

Here you can see the contours of Black's 'wall oriented' strategy and White's premature territorial ambitions. But White will catch up.

55

The material balance is still even, and black's man on the middle wall section is in dire spot.

58

The black double will jump east, over the Wall, and provide protection for the previously isolated black man. Black starts to build dominance on the Wall, a strong strategy.

64

Here it's six against five on the Wall, and Black already attacks an isolated white man.

69

Here AiAi makes a mistake. In this stage you don't leave the Wall unless you're forced, not even to capture. The territorial claim is entirely premature.

80

Now watch this. There is only one established territory, the small empty one in the center. All the rest is in principle still one territory, holding five white men and five black man. At the bottom there are at least seven cells that will not turn into wall cells if vacated, part of a large chunk of territory in which White dominates and can grow.
On the Wall however, White is in a dire position. Black can stage attacks that keep White busy and in the meantime, with six men on the Wall, he can jump between (potential) territories. White is increasingly locked up between the walls.

90

Black stages an attack on the isolated white man on the wall segment top-left. The last move started from a cell that was already a wall cell (in case you're wondering why no new man appeared).

100

Here, after a one-man exchange on the Wall, the attack succeeds.

106

And it's positive feedback on the Wall, White's position is in fact already lost. His last man on the Wall is under attack.

112

And here it is trapped.

122

And here another new-born bottom-left is trapped.

141

White is off the Wall now, and the one new-born he can generate bottom-right will immediately be eaten by a black man next to it (two of them, but actually one is enough).

150

Black has captured a white man by jumping off the Wall. This is the endgame and Black now can afford to do so. At the top he can create two extra wall-men and capture the lone white stragglers at his convenience.

155

If AiAi were capable of it, then this would indicate desperation.

171

Black now can turn the score from 15-11 to 12-14 by capturing the white man on the left. Of course he could capture the white man on the right as well, and claim the empty single-cell territory, details that may matter a lot in games that are more evenly balanced but not in this case. Here Black has clearly won by winning the existential fight on the Wall!
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
This is fabulous. Thanks. I know how much work posts like this are.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
milomilo122 wrote:
This is fabulous. Thanks. I know how much work posts like this are.
Well, I'm retired and I like what I'm doing. I like what you're doing too, by the way.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
TPoG
Denmark
flag msg tools
ISBN: 1-85723-146-5
badge
Hmm...
Avatar
mbmbmbmbmb
Great report. Would be interesting to see a lightly themed(?) physical implementation of the game.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
The Player of Games wrote:
Great report. Would be interesting to see a lightly themed(?) physical implementation of the game.
I've had the same thought. Christian may have inadvertently created a game with some commercial potential. Not sure, but it's a possibility.

[edit] No insult intended Christian!
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
milomilo122 wrote:
The Player of Games wrote:
Great report. Would be interesting to see a lightly themed(?) physical implementation of the game.
I've had the same thought. Christian may have inadvertently created a game with some commercial potential. Not sure, but it's a possibility.
Here's where I stand. It's my last game and I'm more than pleased with it because of it's novel behaviour and its potentionally wide range of strategies and tactics. That's it really, I'm a sideliner now.

Besides, I don't think so. It's the type of game that takes time to play, and even more time to play well. It's not unlike games like Chess or Go, and they wouldn't be marketed if they weren't Chess and Go.

On the other hand it's pretty versatile and multi-player versions (someone will come up with it) would be collusion infested, unbalanced and weird. But I understand that may sometimes be the very point. Anyway, I'm not against philosofying about themes. If it were a car, then I just want the engine to be good. If themed, the bodywork should be good too, but that's hardly my forte.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
christianF wrote:

Here's where I stand. It's my last game and I'm more than pleased with it because of it's novel behaviour and its potentionally wide range of strategies and tactics. That's it really, I'm a sideliner now.

Besides, I don't think so. It's the type of game that takes time to play, and even more time to play well. It's not unlike games like Chess or Go, and they wouldn't be marketed if they weren't Chess and Go.
More and more, I'm convinced there is a significant market for stuff like this, it's just that the path to market is fraught, you have to be cleverer than normal in how you try to reach that market. We've seen publishers successfully navigate it a few times now though.

Quote:
On the other hand it's pretty versatile and multi-player versions (someone will come up with it) would be collusion infested, unbalanced and weird. But I understand that may sometimes be the very point. Anyway, I'm not against philosofying about themes. If it were a car, then I just want the engine to be good. If themed, the bodywork should be good too, but that's hardly my forte.
If a publisher were interested, it wouldn't need to be your forte. The publisher would be responsible for all the stuff that bores you.

I know you're not interested, but just trying to be clear about why I think it *could* work.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
milomilo122 wrote:
christianF wrote:

Here's where I stand. It's my last game and I'm more than pleased with it because of it's novel behaviour and its potentionally wide range of strategies and tactics. That's it really, I'm a sideliner now.

Besides, I don't think so. It's the type of game that takes time to play, and even more time to play well. It's not unlike games like Chess or Go, and they wouldn't be marketed if they weren't Chess and Go.
More and more, I'm convinced there is a significant market for stuff like this, it's just that the path to market is fraught, you have to be cleverer than normal in how you try to reach that market. We've seen publishers successfully navigate it a few times now though.

Quote:
On the other hand it's pretty versatile and multi-player versions (someone will come up with it) would be collusion infested, unbalanced and weird. But I understand that may sometimes be the very point. Anyway, I'm not against philosofying about themes. If it were a car, then I just want the engine to be good. If themed, the bodywork should be good too, but that's hardly my forte.
If a publisher were interested, it wouldn't need to be your forte. The publisher would be responsible for all the stuff that bores you.

I know you're not interested, but just trying to be clear about why I think it *could* work.
If you want I can donate the game to you. I keep a say in any rule changes (because I feel the rules are necessary and sufficient) but the theme work is yours. I won't claim any revenues, only my name as the inventor of the 'engine'. And I care about the name 'Storisende' because it marks the end of my own journey in this fascinating field.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Craig Duncan
United States
Ithaca
New York
flag msg tools
designer
Avatar
mbmbmbmbmb
Quote:
If you want I can donate the game to you. I keep a say in any rule changes (because I feel the rules are necessary and sufficient) but the theme work is yours. I won't claim any revenues, only my name as the inventor of the 'engine'. And I care about the name 'Storisende' because it marks the end of my own journey in this fascinating field.

I'm curious about the name. Could you explain its significance, Christian?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
christianF wrote:
milomilo122 wrote:
christianF wrote:

Here's where I stand. It's my last game and I'm more than pleased with it because of it's novel behaviour and its potentionally wide range of strategies and tactics. That's it really, I'm a sideliner now.

Besides, I don't think so. It's the type of game that takes time to play, and even more time to play well. It's not unlike games like Chess or Go, and they wouldn't be marketed if they weren't Chess and Go.
More and more, I'm convinced there is a significant market for stuff like this, it's just that the path to market is fraught, you have to be cleverer than normal in how you try to reach that market. We've seen publishers successfully navigate it a few times now though.

Quote:
On the other hand it's pretty versatile and multi-player versions (someone will come up with it) would be collusion infested, unbalanced and weird. But I understand that may sometimes be the very point. Anyway, I'm not against philosofying about themes. If it were a car, then I just want the engine to be good. If themed, the bodywork should be good too, but that's hardly my forte.
If a publisher were interested, it wouldn't need to be your forte. The publisher would be responsible for all the stuff that bores you.

I know you're not interested, but just trying to be clear about why I think it *could* work.
If you want I can donate the game to you. I keep a say in any rule changes (because I feel the rules are necessary and sufficient) but the theme work is yours. I won't claim any revenues, only my name as the inventor of the 'engine'. And I care about the name 'Storisende' because it marks the end of my own journey in this fascinating field.
I'll tell you what: it looks like Blooms is going to be published by a real publisher, and if that's a success, then of course I will pitch more games to him. I'll put Storisende in the mix of games I pitch. That's probably at least a year off. But I'll insist you take a royalty, because game designers deserve royalties goddamit.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
cdunc123 wrote:
Quote:
If you want I can donate the game to you. I keep a say in any rule changes (because I feel the rules are necessary and sufficient) but the theme work is yours. I won't claim any revenues, only my name as the inventor of the 'engine'. And I care about the name 'Storisende' because it marks the end of my own journey in this fascinating field.

I'm curious about the name. Could you explain its significance, Christian?
Storisende is a town in the realm of Poictesme in my favourite trilogy 'Figures of Earth', 'The Silver Stallion' and 'Jurgen' by the American writer James Branch Cabell. It's where all things come together at the story's end.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
milomilo122 wrote:
I'll tell you what: it looks like Blooms is going to be published by a real publisher, and if that's a success, then of course I will pitch more games to him. I'll put Storisende in the mix of games I pitch. That's probably at least a year off. But I'll insist you take a royalty, because game designers deserve royalties goddamit.
Sounds great to me and I may deserve royalties but I don't need them. But if you insist and anything comes of it, I'll donate them to an animal charity.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Craig Duncan
United States
Ithaca
New York
flag msg tools
designer
Avatar
mbmbmbmbmb
christianF wrote:
cdunc123 wrote:
Quote:
If you want I can donate the game to you. I keep a say in any rule changes (because I feel the rules are necessary and sufficient) but the theme work is yours. I won't claim any revenues, only my name as the inventor of the 'engine'. And I care about the name 'Storisende' because it marks the end of my own journey in this fascinating field.

I'm curious about the name. Could you explain its significance, Christian?
Storisende is a town in the realm of Poictesme in my favourite trilogy 'Figures of Earth', 'The Silver Stallion' and 'Jurgen' by the American writer James Branch Cabell. It's where all things come together at the story's end.
Ah, makes sense. So I suppose it's pronounced "Story's end" and not (as I have been mentally pronouncing it so far) "Story send." And not "Story SEN day" either.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

Storisende Rules

Here's another one, now with a large lake in the middle. I chose this lay-out to ensure that the players initially would be more or less 'on their own'. It shows that AiAi, at strength 3 with 5 seconds per move, undervalues men on the Wall, both in terms of breeding them and keeping them on the Wall. Eventually this results in a clear defeat, although by the looks of it, in terms of territory, the program seems to do fine. But as said before, territory is a far away promise not to be taken into account all that much till far in the middle game.
The game again took less than a hour. Last moves are indicated and move numbers appear at the top of each diagram, yellow and even for White, black and odd for Black.



AiAi (white) - Christian Freeling (black)

09

AiAi opened and I didn't swap. The move number is 9 because placing the modules also counts as moves. We're on different sub-grids and we both start breeding.

22

I have my first wall piece and two doubles. White switched sub-grids by making two single moves with its double.

40

Here AiAi sacrifices two men for no reason that I can see.

50

White has one man on the Wall and I got six. It's at the cost of territory and White indeed seems to do well in those terms.

67

Black has ten on the Wall, White five. Black's last move guards the target cell of the white double below. The four at the top await rearrangement.

79

Here Black sacrifices a double after an exchange that will be accepted because after Black recaptures, White can also recapture, but with a man on the Wall. I judged that a two-man sacrifice was worth it for that very reason.

99

Black's four attack the white man to the south-west. White will protect it by a double on the adjacent wall section, but Black will back up and force an exchange.

113

Here's the offer to exchange and White accepts although turning the double into two singles would probably be better.

127

Black prepares to jump over the lake and attack from the rear, grabbing patches of white territory in the process.

135

The white man on the bottom left wall section is trapped and White's position is crumbling apart.

155

Time to round up the stragglers.

171

Which is a process to be carried out at ease. Against a human opponent the game would have been considered a clear loss far before this point. Against AiAi I'll indulge in making it utterly clear.

211

Of course I can capture the last ones too, but this seems utterly clear enough.

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb

I don't think I appreciated how colossal the game tree is for this one, at first. And yet the positions have clear meaning. I'm loving this.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
Quote:
There's room for improvement!
That's what I wrote yesterday in a message I accidentally deleted while writing this one, but there sure is. Stephen sent me a message correcting his previous one. '5' is the strongest level, not '0'. I had been wondering about that, that's why I played previous games at '3', middle of the road so to say.

So the game in which I crushed AiAi yesterday at level '0' is very misleading and there's still hope for AiAi. Fortunalely I hadn't posted it yet. I'll start a new one now, at level 5.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

Storisende Rules

There's room for improvement after all. In this game AiAi was at maximum strength, with 15 seconds per move. I used the same lay-out as in the previous example game for the same reason: it ensures both players are initially more or less on their own. AiAi uses that freedom to prematurely grab territory, I'm more focused on wall domination. The lay-out may have affected AiAi's performance. Overall it played better than before, but it still makes the occasional weird sacrifice.

Occasionally there are endgames in which 3-fold may arise around a patch of territory, with a piece on the Wall attacking and a piece inside evading. It's never a forced cycle so I'm inclined to make it a draw regardless of the score. In that case the player leading the score will have an incentive to avoid it.



AiAi (white, max strength, 15 sec per move) - Christian Freeling (black)

09

AiAi opened and I didn't swap. The move number is 9 because placing the modules also counts as moves. We're on the same sub-grid and I placed my double so that White cannot jump to the south-east.

21

White cannot jump south-east with its double and Black cannot jump north-west with the double facing it. But Black can still jump north-east and then with his second double proceed with building at the Wall.

26

White's double isn't very useful anymore and would have to invest two steps in the switch to an adjacent sub-grid. Instead AiAi offers it for an exchange that Black accepts.

49

Twentythree moves on the situation on the right is provisionally settled and Black prepares a wall-building series of jumps to the right and up.

61

Another twelve move onwards White 'controls' two thirds of the board, but its men are thinly spread and there's only one white man on an isolated wall cell.

68

White has captured an isolated man top-right and is now on the way back. Black in the meantime prepares troops to reinforce the bottom section.

85

The black double at the bottom created the wall cell that now is occupied by four black men, ready to jump the lake.

99

And there they are, untouchable by White who already lost a piece in the process. Black has now set his eye on the lone white man on the double wall cell on the left. That's a 6-cell jump, 7 if the piece retreats one step (as it will).

111

There they are, ready to jump. White has no way to defend the piece or escape with it, other than to leave the Wall.

137

Twentysix moves onwards Black wants to capture the bottom left patch of six cells, but at the same time get the bottom double out of its one-cell enclosure. But he needs 3 men to make a jump, so one will have to come off the Wall.

151

Here one man has come off the Wall to capture a white man inside, and another has come off the Wall to join the two bottom men, ensuring their release.

157

Black has taken control of the six-cell territory and White prepares an attack on the black man that is provisionally stuck on the two Wall cells on the left. Black in the meantime goes around to the top-right territory which is hardly defended, or defendable for that matter.

169

Here the job is done and Black can create another wall cell with another man on it. A good thing because men are getting scarce.

180

The score is now White-4 and Black-15. The top 3-cell territory cannot at the moment be taken by Black because after capture/recapture/recapture the remaining black man falls victim to the white man on the Wall.
On the other hand, if White uses that man to claim the four cells below it, then Black can win the 3-cell section.
Which are all superfluous details, Black's win is clear. Maybe AiAi is stronger on a board without a big lake, because the interaction is more immediate and dreams of large pieces of territory that seem hardly contested wouldn't emerge so quickly.

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

Storisende Rules

This is a final one, for your eyes only.
I've selected 5 images up to the middle game, and then the final position. I've added a couple of tentative conclusions at the bottom.

AiAi operates again at its maximum strength and with 15 seconds per move.



AiAi (white) - Christian Freeling (black)

13

AiAi opened and I didn't swap. The move number is 13 because placing the modules also counts as moves. We're on different sub-grid and I placed my double so that White is forced down or sideways.

31

Eighteen moves onwards we both have our first wall cell and White already tries to cover more terrain.

64

A big jump and White seems to be doing fine, but its last move is one of those unnecessary sacrifices it sometimes makes. The next image shows another one.

84

Here White jumps north-west and gets captured, where it could have jumped west and keep the double. The Black double can, after it makes the capture, be captured itself by the white man on the Wall. But for me that's worth it: one white piece less on the Wall.

147

This is a bit of a sorry move too. I would think twice before capturing a double in the territories with a man on the Wall, but now that it jumps and leaves a newborn on the Wall, I of course capture it.
Now consider this position and see the final position below. It happens most likely because heuristics are primarily aimed at 'territory'. And territory it has: six cells with its inhabitants neatly locked in. Black has just captured the last white man in the last other patch that White occupied.

265

Of course the game was decided long before this point. I just want to illustrate that AiAi currently doesn't seem to have the right heuristics to deal with a wall-oriented long term strategy. Giving more value to wall-cell breeding would seem an appropriate measure.

I also want to illustrate that Storisende has a long term strategy and that it's a wide strategy and a deep strategy, permeated with details and far from being fully understood.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

Storisende Rules

Don't interpret this wrong..
I've always emphasised Storisende's novelty as being both territorial and existential. Other than with Go or other territorial games that feature capture, the embedded existential game takes place on a different plane and with a different timing than the territorial one. This is one of the reasons is why AiAi plays very poorly on large boards. But it gives me the opportunity to illustrate the point: AiAi operates again at its maximum strength, with 60 seconds per move. It took me some 120 moves to eliminate it. Here it is in 5 frames with 50 move intervals.



AiAi (white) - Christian Freeling (black)

52


102


152


202


252

My last move was the capture of the last white man. Note that I had priorities throughout the game other than say claiming the 7-cell territory top-left. Territory is the formal goal and eliminating the opponent if you get the chance, is a convincing way to do it. Not that it would happen in actual play between human players because humans see at some far earlier stage that a game is won or lost.

I'd very much like to hear any comments on this interaction of two goals under a common denominator. If you've ever seen an abstract game with similar behaviour please let me know.

Don't interpret this game wrong, AiAi isn't all that bad on smaller boards. However, it is unclear, even to Stephen, what the actual differences between AiAi's playing levels are. So I will play six 4-module games on levels 0-5 and a minute thinking time and see how it pans out.

3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

On very small boards I encountered a new difficulty in sorting out the levels. I played Black in two games, swapping AiAi's placement both times, so in both cases I made the actual first move. AiAi got 30 seconds per move.
Here's level '0', the first diagram after we both made a triangular round, the second after I gave up.




And here is level '5':



That's not much of a difference is it? Actually I haven't really experienced any difference between the levels. On larger boards its easy to win because of the increased branching factor and the fact that humans can make long term plans. On small boards the number of iterations increases dramatically, there's no 'long term' and tactics become heavy. I'll need more time on small boards to sort out the levels.

 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

Storisende Rules

Getting colder.
I am less and less hesitant to call Storisende my favourite game. This is off topic regarding this thread but has become the general context of my posts about it.
I'd like to illustrate and share a remarkable property that emerged on 4-module boards. Generally speaking Storisende is a hot game: you want to move and you can move. But on 4-module boards it's different and you're forced into very crammed positions in which every move may seem bad. The fight for territory and the existential fight on the Wall are also closer intertwined. The former enters the equation at a much earlier stage.
AiAi runs three million iterations in 30 seconds, over an increasingly lower branching factor, so for me it feels as if it has the game in the bag from the onset. I switch sides to see what it moves, then make that move myself and lose again. Very enlightning in terms of refinement of tactics in a very small area.


For illustration I use a symmetric 4-module lay-out. The initial deliberations differ on any other 4-modules shape, but their general aim is the same: where do I start?


On the red dots the initially placed double has five directions in which it can jump. On the yellow dots it's four. If you place on a red dot (or a yellow one for that matter), the only sensible reply is the dot diagonally opposite of it. The other ones allow an immediate capture and other options appear to allow the first player too much freedom. So whether or not the swap is accepted, the game would probably start off like this:


AiAi (black) - Christian Freeling (white)

6

This limits jump directions to three and that's where the game initially goes without much of a credible alternative.

18

After six symmetric jumps, four of which were breeding, White still has to follow Black in perfect symmetry. Black now steps sideways with one of the three men, putting two men in the green cell. White captures those two and Black of course recaptures.

21

And we're in symmetric territory again, but Black now follows White!

29

Three moves onwards symmetry is still mainained and the first wall cells have appeared.

31

One move onwards Black breaks symmetry.

71

Twenty moves onwards Black has won, a lesson in tactical finesse. The end position is amazing. Black leaves the Wall for two extra points of territory.

- If White passes, both having four points as it is, then Black captures the white man on the green cell (!), White must recapture turning his departure cell into a wall cell and leaving him with two points.
Note that if Black captures the white man on the beige cell, the game ends in a draw!

- If White makes a double of his two singles, it is captured.

- If White moves the top single away from the other Black may capture either.



Given that it's a pie, first placements on the ring around the central one is possible. I haven't tried it yet but there are four cells on the second ring that allow jumps in four directions. There are two possible replies to such a placement on the central ring, one with initially four directions to jump, one with three. There's one on the same ring diagonally opposite, inviting symmetric play as in th example. And there's the small green dot that only has one direction to jump, that's why I made it small. But on closer investigation the second player can get away with symmetric play nonetheless. Only problem is he may still lose.

Conclusion
On small boards the game is quite suddenly getting much colder. I've not had this experience on a compact convex 7-module board at any time.

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

My idea to make Storisende modular was because I wanted to shift somewhat from the 'sport weapon' idea to a more recreational game. But the more I play it the more I realise that it would make a formidable sport weapon, in particular on hexhex base 5 or 6. So I asked Stephen if it could be implemented and got a reply with the implemented version attached! That was great! (although I overlooked the option, fortunately I don't play that stupid).


So here it is and as expected it's colder and AiAi is stronger on smaller boards. But base-4 very enlightning if you want to find out about emerging refinement of tactics.


Edit:

This was on strength '5' and I won (the last white man on the Wall will fall without compensation). But I've seen instances where level '1' appeared stronger than '5' so this is very provisional.
This is also an instance where the formal and the territorial goal are clear. Black could (after capturing the white man on the Wall) hop the remaining diamond of beige cells. It would create wall cells while breeding new men who then could claim more territory.
On the other hand the black man at the bottom controls 7 cells of territory (if it stays put). And if Black hops into the central area he has three more cells without 'wasting' territory for wall cells.
It doesn't matter here, but in high level games it may matter a lot.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Craig Duncan
United States
Ithaca
New York
flag msg tools
designer
Avatar
mbmbmbmbmb
Thank you for these sample game illustrations; very interesting.

I'm writing with a design question.

Consider the following picture of the board at the end of a game. (I know that a real game would not end here since Black can reduce White's territory. But just suppose Black is stupid and passes too, after White passes.)



The rules state that after both players pass, any beige hexes convert to green. I'm curious from a design standpoint what led to that rule. I can imagine two other alternative rules:

Alternative 1: Beige hexes don't count as territory.

In the board above, White would only have 2 territory, not 4. White could get 1 more territory (temporarily) by moving from the beige to the green cell, thereby turning the beige to a wall cell. ("Temporary," because then Black could capture the just-moved white piece.)

Alternative 2: Beige hexes count as their own sort of territory.

In this case the current score is White 3, Black 4. White doesn't get a point for the green cell immediately southeast of the occupied beige cell.

I'm not saying that your choice of the beige-turns-green-at-end-of-game rule is flawed. I'm just curious from a design perspective why you preferred that rule.

P.S. I posted some days ago a different question on the Storisende BGG page. Apologies if you already saw this and just chose not to post a reply; I don't mean to hassle you into replying. But just in case you didn't notice the post, I'm mentioning it here.


1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
cdunc123 wrote:
Thank you for these sample game illustrations; very interesting.

I'm writing with a design question.

Consider the following picture of the board at the end of a game. (I know that a real game would not end here since Black can reduce White's territory. But just suppose Black is stupid and passes too, after White passes.)



The rules state that after both players pass, any beige hexes convert to green. I'm curious from a design standpoint what led to that rule. I can imagine two other alternative rules:

Alternative 1: Beige hexes don't count as territory.

In the board above, White would only have 2 territory, not 4. White could get 1 more territory (temporarily) by moving from the beige to the green cell, thereby turning the beige to a wall cell. ("Temporary," because then Black could capture the just-moved white piece.)

Alternative 2: Beige hexes count as their own sort of territory.

In this case the current score is White 3, Black 4. White doesn't get a point for the green cell immediately southeast of the occupied beige cell.

I'm not saying that your choice of the beige-turns-green-at-end-of-game rule is flawed. I'm just curious from a design perspective why you preferred that rule.

P.S. I posted some days ago a different question on the Storisende BGG page. Apologies if you already saw this and just chose not to post a reply; I don't mean to hassle you into replying. But just in case you didn't notice the post, I'm mentioning it here.


Hi Craig,

Thanks for your interest and I hadn't looked at the Storisende page yet. You ask some in deep questions and it's rather late to turn away from sorting out levels by playing hexhex base-4 at all levels with 15 seconds per move. Very cativating, so I'll try to answer your questions tomorrow morning (CET).

Regarding your 'beige' question, it was a simple and intuitive decision and I don't regret it. It has inherent and in this case fairly immediate consequences and so of course would any alternative have. But the main consideration was to see the board as territory which then is divided by the emerging walls. And it's nice for AiAi to not have to distinguish between two types of territory (it's busy enough already!)

The 'no territory' scenario you describe would be the same under the current rules but Black couldn't make the one mistake he can make in this position under the rules as they are: capturing the white man on the beige cell. That would be a draw, while if beige were no territory, it would still be a black win.

All in all my intuition led me to believe that the way in which the rules now cover territory is the simplest. And I feel its consequences (as you may have noticed against AiAi) should be at least as interesting as the alternatives you suggest.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

Playing Storisende to me feels like playing Chess and Go at the same time. This is not suggested at all by its mechanics, so it's maybe better to say that it merges an existential and a territorial goal unlike any other game I know.
Initially I envisioned it as a recreational game meant for casual players who love variety in terms of strategy and tactics, and not so much as a 'sport weapon' tailored to tournament play. Now I've come to realise that the game is so deep and wide that it may actually drive away casual players and that it would be a great game for players inclined to dig deeper if (and only if) the effort is worth it. And in Storisende's case the effort is more than worth it!

Modular or hexhex
Here's the difference: the modular board prevents almost any attempt to record a game and it totally prevents any opening preparation. The hexhex board doesn't. I have already mentioned the fact that on a 4-module board the game gets very cold. That's due to the very limited number of options for the initial double jumps and the subsequent single ones that must be made to switch to other sub-grids.


A bit of that is still present on the base-4 hexhex which although it has only nine cells more that 4-module board, allows for more initial jumps. I played six games on it, level '0' to '5', and in all cases AiAi swapped my first move (a cell adjacent to the central one). Next came five jumps by each, back to the original cell, and we're in the diagram above. This happened in all games except the level '0' game. It shows that 'theory' immediately takes center stage. Hereafter the lines diverged. All games were set on 15 seconds per move for AiAi, good for a million or so iterations and 6-ply depth.


Here are the six final positions

Level-0: White's 56th move: win
Here AiAi deviated from a line leading the above position and took me a bit by surprise. But it handled the endgame badly. It's still 10-6 in Black's favour but the 2-cell top territory can be captured with two white men to equal the score and White then still has two men left to claim one of the single cell territories without bothering about the large black one.


Level-1: White's 35th move: win
Level 1 was totally bad.


Level-2: White's 59th move: draw
This was much better: White can't capture the man on the beige cell and capturing the other (if Black would allow it in the first place) then Black can recapture off the Wall. Black has 6 points, White has 5 and ample 1-cell territories to put his last man in, but that's not more than a draw.


Level-3: White's 56th move: win
This one I won quite convincingly and the endposition is more characterised by my attempt to trap the last man on the Wall (which I did in this position).


Level-4: White's 56th move: draw
The second draw! White has 3 points, Black has 6. But White cannot afford to capture a man on the beige cell because it is recaptured and then the score is definite. And he cannot capture the other one because it keeps moving to and fro over the beige cell. And even if he could, then it would have to be on the top left green territory cell to ensure a win. So White can only claim the far right 3-cell territory.


Level-5: Black's 28th move: loss
I don't want to talk about it! yuk


 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
1 , 2 , 3  Next »   |