The Settlers of the United States of Catan

Houseal Lavigne Transforms the National Land Cover geospatial data into the iconic board game Settlers of Catan

I wanted to share a poster and process my team and I completed for the Map Gallery at Esri’s User Conference this past July. While we managed to win an award for another map, one I’ll share in a few weeks, this map came up short. However I think just about anyone at the conference would agree it was one of the most interesting and popular maps at the event. If you are interested in a high-resolution JPG or PDF of the full poster (84″ x 48″) connect with us on our company’s LinkedIn page and let us know. A low-res JPG is at the bottom of the story.

In the new Golden Age of gaming, board games are no exception. Over the past decade, board games have seen a resurgence driven by independent games that feature innovative game mechanics, artistic design, and online versions to engage new players. Millennials have had a large part in the board game boom, embracing ‘table-top’ games as social events. This includes dedicated board game shops with spaces for casual and competitive events and the most popular games are developing regional and even international leagues where players test their skills and strategy.

At the foundation of this resurgence is the board game The Settlers of Catan. First published in Germany in 1995, ‘Settlers’ has long acted as a first introduction into the world of detailed, strategy-based board games. Since it’s publication, Settlers has sold over 20 million copies in 30 different languages including various expansions and spinoffs, making it one of the world’s most popular board games. This includes versions themed to Star Trek and Game of Thrones. Settlers of Catan Copyright © 2019 Catan Studio, Inc. and Catan GmbH. All rights reserved.

Embracing our inner nerd, we set out to reimagine our world within the confines of the Settlers of Catan, using the game’s distinct hexagonal tiles to create a game board of the continental United States. Using Esri software we applied the rules and game pieces of Settlers to the United States, developing a new board informed by land coverage, cities, population, and other data. While perhaps unplayable, this new board game of America provides an interesting lens through which to view geospatial data of the United States.

Our Process

Classifying the Tiles

The first step in developing our Settlers map of the United States was to identify a method to classify the individual hexagonal tiles. We knew that once we applied a hexagonal grid to the county, we would need a way to enrich the tiles with data to inform what landscape each would be. Our ultimate goal was that each hexagonal tile would be reduced to a single landscape, similar to the game itself. For this we used the United States Geological Survey’s national land cover database, which offers recently updated and credible data. Reviewing the defined categories, we determined that land cover could be sorted into the six landscape types of the game. Using the data would allow us to identify a majority land cover for each tile that would inform what landscape each hexagon received.

USA into Hexagons

Our next step was to apply a hexagonal grid to the United States and create the individual hexagons that would represent the tiles of the game. We used the Generate Tessellation tool which creates a grid in either hexagon, square, diamond, or triangle shapes. Using the extent of the U.S. as our input, we created a hexagonal grid reflective of the board game. To make sure the shape and orientation was consistent with Settlers of Catan, we created ‘transverse hexagons.’ This process required some trial and error to find the preferred size of the hexagons that would be small enough to pick up attributes of the land cover while big enough to be legible as tiles on the map. We eventually decided on hexagons 50,000 square miles in size.

Reclassifying the Raster Data

To prepare the data for our map we used the Reclassify geoprocessing tool within Spatial Analyst to convert the National Land Cover Database’s 20 categories into our 6 identified landscapes from the board game. Given the large size of the data, encompassing land cover for the entire country, the raster data was simplified with the Majority Filter, which identifies and changes individual pixels completely surrounded by a different land cover. In addition, we used the Boundary Clean tool to smooth out boundaries, making the data easier to manage. 

Summarizing Data within the Hexes

Having developed our hexagonal tiles and prepared the raster data, our next step was to join data in order to identify what Settlers landscape each tile should be. For this we used the Zonal Histogram geoprocessing tool, which summarized the quantity of individual raster pixels, each representing land cover, within each hexagon. The resulting table informed us of the land cover breakdown of each hexagon. We exported this data and used Microsoft Access to summarize and determine, based on land cover and our Settlers categories, what landscape each tile would be.  

Weighting the Results

Upon completing this step, we were unhappy with the results, which showed the majority of the U.S. as pasture, forest, and field. Reviewing our work, we realized that the land cover categories which we had attributed to hills and mountains were very few, even in areas we knew they should be present. A great example was the rocky mountains, where the land cover was often defined as shrubs or forests even if on an intense slope. Understanding this limitation, we developed a weighting system to better emphasize areas of hills and mountains and achieve a more even distribution of tiles on our map. Once we were happy with the output of the weighted system, we brought our table back into ArcGIS and joined it with our hexagonal tiles. This acted as the base for our final map, which was stylized in CityEngine to add texture.

Adding the Chits

To develop the chits for each tile, we needed to enrich our hexagons with necessary demographic data. For this we added data from the populated places from Esri’s Living Atlas and summarized data within each hexagon. We summarized total population which determined the board game’s probability scale – This scale is based on the probability of a dice roll, with the most favorable numbers being those with the highest probability; 6’s and 8’s, and the least favorable being those with the lowest, 2’s and 12’s (7’s are the single highest probability, but are reserved for the robber). Applying this scale to our population data, we gave the largest populations the highest probabilities scaled down to the lowest receiving the lowest probabilities. We also textured the probability counter on each chit based on this information. We also used the summarized data to identify the largest place by land area. The first letter of the largest place was applied to the chit. 

Finishing In CityEngine

To finish the board, we exported the geodatabase and brought it into CityEngine. In CityEngine we wrote and applied CGA (procedural) rules to the hexagons and the spaces between to help stylize the board and incorporate other components of the game. This included rules to apply textures to each landscape, using pictures of the board game tiles, as well as rules to add the settlements, cities, roads, and chits. We also used CityEngine to extrude each tile to add depth. We extruded the hexagonal tiles, chits, and roads as well as the settlements and cities while applying appropriate roof types to mimic the game pieces. The result was a fully textured, 3D Settlers game board of the United States. 

About Houseal Lavigne

We are an innovation-based urban planning and design firm. We pride ourselves on creativity, collaboration, and delivery of quality. Our team approach is built on strong relationships, the exchange of ideas, and a commitment to the integration of technology. Our priorities are to do good, have fun, work hard, and provide responsive, visionary, and viable solutions to our clients and partners. We are a proud partner of Esri, and believe that the integration of the most current technologies should be used to improve the planning process and product—increasing communication and involvement with the public, gathering and assessing vital information, and producing more effective plans, concepts, and recommendations.