Modular China
When playing China last Friday, I had a notion to make the map modular, to randomize the setup for each game. In general, I love the concept of modular boards for games and any game with random tile setup automatically gets a bump in my interest level. In both games last time, I ended up playing houses in very similar places, which is something that random setup helps avoid. A fan did something very similar for Power Grid, so it's not a radical idea.
To begin my toying around with the idea, I looked for a modular variant that someone might have already done. Not finding one, I went about understanding the characteristics of the existing map. There are actually two maps (not counting variants like Border Disputes): one for 3-4 players, and one for 4-5. I don't really understand why there has to be two different maps, though I appreciate that they would play differently. This is, after all, one of the goals for having a modular board. I just couldn't tell you why one is better for 3 and the other better for 5. The maps have an identical arrangement of cities and regions and the only difference is how the roads are connected. To help me conceptualize the differences, I recreated the two maps into one, using a grid-based format.
I don't know if this exercise taught me anything, but now that you can see the map, I'll note some of my observations which influenced how I made the tiles.
There are 5 colors and 9 regions in the game. Most regions have 4-7 cities; purple has 8, but there is only 1 purple region. There are no 4-way intersections in the 3-4 player game, and only 3 with 4-5 players. There are plenty of dead-ends. In some regions, all the houses are connected while in others, you can't make a path between them without entering some other territory.
Looking at the regions and their colors, it is clear that they aren't "equal" (purple being a prime example), but each color as the same number of cards. China, like the original Web of Power, is inspired by real-world geography (though I'm doubtful how closely China follows actual geography). But when thinking about an abstract, modularized version, how should the asymmetry be generated? I have "an overdeveloped sense of symmetry" and automatically want to make each region and color equivalent in some way.
I settled on the idea of having 1 tile per region, 2 of each color. From there, I decided to make 6 cities per region/tile with the option to change that later. Though I am still inclined to make the total number of cities per color to be equal, I admit that I don't have a good reason to do it that way.
I prefer hexes because they tessellate well and arranging 6 cities is easy to do. I started with a basic pattern of 7 cities, then removed 1. (The following diagram presented in lovely Courier because Blogger can't handle preformatted text tags.)
..O...O...........O...O...........O...O..
O...O...O.......O...X...O.......O...O...O
..O...O...........O...O...........X...O..
I decided that each color would have 1 tile with the center missing and 1 tile with the edge city missing. I then had to create 5 patterns of roads for each tile type. Of course, there are far more than 5 ways to connect 6 cities, so I had to make some rules for myself. And my "had" I mean, I'd never be able to decide which patterns to use unless I had a set of rules for me to follow. With that in mind, each city could connect only to it's closest neighbors. No tile would allow for a 6-city road. And there would always be paths leading out from the corners to the adjacent tile. Even with the rules, there were plenty of arbitrary decisions.
Working with hex tiles on your typical drawing program (I use OpenOffice) is a pain. Oo has a great feature which allows you to set the vertical and horizontal grid to different spacing, which you can use to create the 1:0.866 ratio of a hexagon length:width. But to hash out my ideas, I went with squares instead. When laid in an overlapping brick format, you can get them to approximate hexes.
The tile in the lower left is my template. The diagram in the lower right shows how they should be laid out. After printing and cutting out the tiles, I noted the need for the arrows so that they could be oriented properly. Hexes rotate 60°, 120°, or 180°; when using squares to approximate hexes, you only get 0° or 180°.
I played around with some tile arrangements and noted that there was a tendency to be too many stragglers along the edges. Several of the edge tiles had 1 to 3 cities that weren't connected to anything (or each other). In China, there are no islands--all cities connect eventually. Web of Power, you have England, which is a 5-city island. I don't mind islands, but because road-building is such an important part of the two-pronged strategy of placing houses and roads only count if they are 4-cities, having small islands is unacceptable. In some cases, you can simply remove those isolated cities from play (since we have 60 cities rather than China's 51), but in others, this created regions with fewer than 4 cities (another unacceptable situation).
I'm not sure it is possible to create tiles in such a way as to prevent these problems, at least without very convoluted placement rules . One thought is to allow islands, but make them their own region for scoring purposes, similar to the harbors from Border Disputes. But this still doesn't solve the road issue. Or I could add a frame around the 10 tiles that allows the edge cities to be connected, but it gets a bit wonky around the corners and breaks the "only connect to closest city" rule. And it still doesn't address any islands that appear in the middle.
Overall, it was a project with mixed success. I still think they concept has merit, but there are obviously kinks that need to be worked out.
Aside 1: When randomly arranging the tiles, what should be done when two tiles of the same color are adjacent? You can either arrange them so that this doesn't happen or let them be one single region (like purple in the original). I think this is something that would need to be shaken out during playtesting.
Aside 2: Yes, I know the red and orange are awfully close to each other. I would have changed it, except I'm paying homage to the fact that red and orange are hard to distinguish in the original game. Yeah, that's it. ;)
To begin my toying around with the idea, I looked for a modular variant that someone might have already done. Not finding one, I went about understanding the characteristics of the existing map. There are actually two maps (not counting variants like Border Disputes): one for 3-4 players, and one for 4-5. I don't really understand why there has to be two different maps, though I appreciate that they would play differently. This is, after all, one of the goals for having a modular board. I just couldn't tell you why one is better for 3 and the other better for 5. The maps have an identical arrangement of cities and regions and the only difference is how the roads are connected. To help me conceptualize the differences, I recreated the two maps into one, using a grid-based format.
I don't know if this exercise taught me anything, but now that you can see the map, I'll note some of my observations which influenced how I made the tiles.
There are 5 colors and 9 regions in the game. Most regions have 4-7 cities; purple has 8, but there is only 1 purple region. There are no 4-way intersections in the 3-4 player game, and only 3 with 4-5 players. There are plenty of dead-ends. In some regions, all the houses are connected while in others, you can't make a path between them without entering some other territory.
Looking at the regions and their colors, it is clear that they aren't "equal" (purple being a prime example), but each color as the same number of cards. China, like the original Web of Power, is inspired by real-world geography (though I'm doubtful how closely China follows actual geography). But when thinking about an abstract, modularized version, how should the asymmetry be generated? I have "an overdeveloped sense of symmetry" and automatically want to make each region and color equivalent in some way.
I settled on the idea of having 1 tile per region, 2 of each color. From there, I decided to make 6 cities per region/tile with the option to change that later. Though I am still inclined to make the total number of cities per color to be equal, I admit that I don't have a good reason to do it that way.
I prefer hexes because they tessellate well and arranging 6 cities is easy to do. I started with a basic pattern of 7 cities, then removed 1. (The following diagram presented in lovely Courier because Blogger can't handle preformatted text tags.)
..O...O...........O...O...........O...O..
O...O...O.......O...X...O.......O...O...O
..O...O...........O...O...........X...O..
I decided that each color would have 1 tile with the center missing and 1 tile with the edge city missing. I then had to create 5 patterns of roads for each tile type. Of course, there are far more than 5 ways to connect 6 cities, so I had to make some rules for myself. And my "had" I mean, I'd never be able to decide which patterns to use unless I had a set of rules for me to follow. With that in mind, each city could connect only to it's closest neighbors. No tile would allow for a 6-city road. And there would always be paths leading out from the corners to the adjacent tile. Even with the rules, there were plenty of arbitrary decisions.
Working with hex tiles on your typical drawing program (I use OpenOffice) is a pain. Oo has a great feature which allows you to set the vertical and horizontal grid to different spacing, which you can use to create the 1:0.866 ratio of a hexagon length:width. But to hash out my ideas, I went with squares instead. When laid in an overlapping brick format, you can get them to approximate hexes.
The tile in the lower left is my template. The diagram in the lower right shows how they should be laid out. After printing and cutting out the tiles, I noted the need for the arrows so that they could be oriented properly. Hexes rotate 60°, 120°, or 180°; when using squares to approximate hexes, you only get 0° or 180°.
I played around with some tile arrangements and noted that there was a tendency to be too many stragglers along the edges. Several of the edge tiles had 1 to 3 cities that weren't connected to anything (or each other). In China, there are no islands--all cities connect eventually. Web of Power, you have England, which is a 5-city island. I don't mind islands, but because road-building is such an important part of the two-pronged strategy of placing houses and roads only count if they are 4-cities, having small islands is unacceptable. In some cases, you can simply remove those isolated cities from play (since we have 60 cities rather than China's 51), but in others, this created regions with fewer than 4 cities (another unacceptable situation).
I'm not sure it is possible to create tiles in such a way as to prevent these problems, at least without very convoluted placement rules . One thought is to allow islands, but make them their own region for scoring purposes, similar to the harbors from Border Disputes. But this still doesn't solve the road issue. Or I could add a frame around the 10 tiles that allows the edge cities to be connected, but it gets a bit wonky around the corners and breaks the "only connect to closest city" rule. And it still doesn't address any islands that appear in the middle.
Overall, it was a project with mixed success. I still think they concept has merit, but there are obviously kinks that need to be worked out.
Aside 1: When randomly arranging the tiles, what should be done when two tiles of the same color are adjacent? You can either arrange them so that this doesn't happen or let them be one single region (like purple in the original). I think this is something that would need to be shaken out during playtesting.
Aside 2: Yes, I know the red and orange are awfully close to each other. I would have changed it, except I'm paying homage to the fact that red and orange are hard to distinguish in the original game. Yeah, that's it. ;)
Labels: Other Games
posted by Patrick at 4:26 PM
<< Home