The Hotness
The Hotness has gone cold...

BoardGameGeek News

To submit news, a designer diary, outrageous rumors, or other material, please contact BGG News editor W. Eric Martin via email – wericmartin AT
 Thumb up

Designer Diary: Canterbury, or A Game of City-Building in Dark Age Britain

Andrew Parks
United States
New Jersey
flag msg tools
Quixotic Games:
Dungeon Alliance Kickstarter - February 2017
Initial Concept

One of the first things that popped into my head after playing Caylus in 2005 was how much I loved the idea that everyone was building the same city. Of course, I also enjoyed the worker placement mechanism, the competition to build the castle, and the resource management aspects of the game, but it was the idea of watching the city's structural dynamic grow based upon player decisions that really got me thinking. Wouldn't it be fun to create a city-building game that focused specifically on this one aspect but felt different from other games of the genre?

Of course, a lot of city-building games were out there, so I immediately set about coming up with concepts that would make this game have a different feel than the others. One of the opening parameters of the design was to remove the concept of personal wealth from the game. Players would have to manage expenses, of course, but I didn't want a player to build up a hoard of gold from using his buildings. I wanted the game to be a competitive city-builder in which players would draw from a common treasury, and that treasury would grow as the city expanded. This would allow us to avoid a "rich get richer" scenario for an individual player, and all of the players' available resources would grow based upon their combined decisions as they developed the city.

But how could I do this and keep the game fair? After all, we couldn't allow one player to show up and drain the city treasury all by himself! And so I decided that on a turn, a player would either collect revenue from the treasury or spend his collected revenue on new buildings. The amount of revenue collected would be based on the city's overall progress so far. To accomplish this, we would need an independent City Marker on the victory track in addition to the Player Markers. Each time a player scored points, the city would also score points, so when a player collected revenue, he would check the position of the City Marker to determine how much gold to collect.

Okay, so I had the game's basic economics down, but I needed more than this to make the design feel different than other city-building games. In particular, I didn't want the game to be about commodities or trading. In other words, I didn't want buildings to produce goods that could then be traded, converted, or sold. Thinking about how else to distinguish each building led me to think about popular city-building video games that I enjoyed, including SimCity and Caesar III. In particular, Caesar III started players off with a little tent village in which residents had nothing but an old well and some vague dreams of a teeming metropolis. Over time, as the city would grow, players would need to provide food, soldiers, and other services in order to attract new citizens.

Inspired by this, I decided that these services would become the center of my city-building board game. Players would compete with one another to provide services to different parts of the city. Small buildings would provide their services to their immediate area, while medium and large buildings would provide services to larger sections of the city. For example, a well would provide water to its immediate area, but a fountain would provide water to several areas, and this in turn would allow those areas to focus on higher level services like food, religion, and so on. In fact, I realized that this hierarchy of services would become a fundamental part of the design. You would need water before you could worry about providing food, for example, and the high end stuff, like commerce and culture, could come only once you had all the basics.

Finding Canterbury

Canterbury was the first game I had ever worked on without having the precise theme selected before developing the mechanisms. I normally like to build the mechanisms around an established setting, so finding the right city for a game that already had an initial set of mechanisms was something new to me. I first thought of appropriate time periods. I didn't want a modern setting like SimCity because I liked the idea of the city starting off with almost nothing, like an ancient community slowly growing into a village and eventually into a city. I had already worked on a game set in classical antiquity (Parthenon: Rise of the Aegean) and wanted to try something new. I also didn't want to create a game set in the Late Middle Ages or Renaissance because those time periods were sufficiently represented by other games.

So I started thinking about the early Dark Ages, a setting which hasn't seen an abundance of games so far. As an English professor, I thought about Dark Age Britain and considered which major cities were developed during that time. As I started my research, I was astonished to discover how little I knew about Britain during the onset of the Dark Ages. I hadn't realized that for a hundred years after the fall of the Roman Empire, nearly every major city in Britain had been devastated by a continuous series of raids. Nothing could be rebuilt during this time due to all of the turmoil. It would actually be the Saxon kings who would later restore order by conquering Britain. But wait a minute, I asked myself, are these the same Saxon warlords who are depicted so villainously in the Arthurian tales I love so much? Yep, it's those same guys. As far as I could tell, these guys had actually saved Britain from eternal chaos.

As I started to pore over the topic more deeply, a city leaped out at me as being a possible candidate for the game: Canterbury. Here was an extremely famous city known for its religious significance in Britain, and also for Chaucer's famous poem – but what I hadn't realized was how significant the city had been in the late sixth century. King Ethelbert of Kent had actually converted to Christianity and had established Canterbury (formerly the Roman city of Durovernum Cantiacorum, which was itself built over a conquered Celtic settlement) as his new Saxon capital. In fact, Ethelbert ordered the creation of the famous cathedral during the city's reconstruction in 597, and this moment represented the birth of Christianity in Britain. As I researched further, I discovered that although the Roman buildings had been long gone, the orderly district layout was still set into the groundwork. The Saxon builders actually used the original district layout when rebuilding the city. Now I had a famous British city with a significant but not extremely well-known origin. I knew at once that this would be the perfect setting for the game.

The organized Roman districts would form our board. To keep things simple, I created a 5 x 5 district city with a total of 25 districts. Each district would hold up to six "building slots" (small structures would take up one building slot, medium structures two slots, and large structures four slots). Each district would also keep track of its provided services (water, food, religion, defense, commerce, and culture) with wooden cubes provided by the players.

First prototype game board

Game Structure

It was important to me that the players did not actually own the buildings. The players represented Saxon Lords working for the king as they ordered the construction of the city. Each player would seek to further his individual prosperity by providing the most services to each district, and by providing the most of each particular service to the city in general.

Therefore, after constructing a building in a district, a player would place one of his colored cubes to indicate that the service was now available in that district, and that in turn would allow new buildings to be constructed there. Medium buildings would provide their cubes to adjacent districts as well. At this point, I established a parameter that would remain unchanged from that moment onward: Once a cube was on the table, it could never be removed. Thus, even though the well I built might get demolished and replaced with a fancy fountain or even the town's water works, the people of that district would always remember that I was the one who first brought them water. This meant that the placement of medium buildings became much more strategic, as placing them early in the game would garner you more cubes since there would be more available spaces on the board.

And so, using the prosperity level of the buildings as the primary vehicle for scoring, and adding in district favor bonuses for area control, I had the basic design and was ready for the first playtest.


Figuring out the costs and rewards for everything has been one of the biggest design challenges for Canterbury, but during our first playtest, things basically worked as intended. We moved along from turn to turn with little downtime and actually played the game to completion.

As we looked up from the first test, happy that the thing had actually worked, we also realized that there wasn't enough meat to the game. For one thing, we needed to add more ways to score points. In particular, we realized that it took far more planning to put in a building that provided a higher end service like culture than a building that just offered food or water. Because of this, we added bonus points (more bang for your buck) if you were able to build these "high end" structures. But we added other types of bonuses, too, things like an Enrichment Bonus for demolishing an old building and replacing it with a better version (for example, upgrading a Watchtower to a Garrison). We also added a Breaking Ground Bonus that rewarded you for building in an empty district, and this bonus was dynamic based on the number of services there at the end of the turn. For example, if an empty district had lots of services being pumped into it from adjacent districts, then building the first structure there would garner a really nice bonus. This also increased the level of interaction as players vied with one another to receive those bonuses.

But the level of interaction was still an issue. Once a player seemed to lay claim to a district, most other players just left him alone and moved on. This was a big problem. We wanted people viciously contending with one another for these districts, not politely ceding control and going elsewhere.

As we played the game more and more, we realized that two types of games seemed to emerge. Either we had cutthroat players who risked everything on screwing over their neighbor, or we had risk-averse players who spread out to the edges and left each other alone. We realized that the risk involved with taking over another player's district was too high and a failed attempt to do so could hurt your chances of winning the game. And even if a player succeeded at taking over an opponent's distict, it had too strong an impact on the other player. We simply could not figure out a solution for some time, and focused on other aspects of the design.

The Citadel (art & graphic design by Chechu Nieto)


We continued to develop Canterbury for many months and came up with all sorts of nifty new mechanisms, including a King's Bonus Chart to reward players who focused on a particular service throughout the entire city. During this time, we also submitted the game as part of a Game Design Competition held by the Games Club of Maryland, and we were strongly encouraged by the feedback we received there. We demoed Canterbury non-stop that weekend, and we were extremely grateful to discover that we had been voted the winner of the competition by those who had participated.

But we still had this interaction problem. For most new players, it wasn't much of an issue as the game had enough going on by this point to open all sorts of interesting strategic possibilities. But to those of us who played it regularly, we knew that the game's replayability was in jeopardy.

I like to give credit where it's due. One day we were playing the game with my good friend and fellow game designer, Geoff Engelstein, who had played the game once before and was coming back to help test all of the new mechanisms. The game played well, but afterwards I mentioned to him the interaction issue and he just looked up, without skipping a beat, and said, "Well, why don't you just add secondary scoring for district favor?"

The solution was so obvious that I still can't believe that it had never occurred to me. Of course, secondary scoring would allow players to interact with one another and have a good chance of getting something out of competing in more districts, even if it was just second place. Furthermore, if you're able to steal a district from someone, chances are they'll still have second place and it won't be as devastating to them. Never mind all the potential for jumping into established, high-scoring districts for just a quick piece of the action.

The addition of secondary scoring solved more than this problem as it added at least a half dozen new strategies that needed to be tested, but now the game had the level of interaction we had hoped for. Even extremely polite players were willing to jump into other districts they would have easily passed on before, and some of them even found themselves gaining control after accomplishing an unexpected move. The interaction level was finally where we wanted it!

The Central District (final version)

Endgame Woes

As time passed and other figures from the game industry took Canterbury out for a spin, we kept receiving the same feedback: The endgame came too suddenly and was too unpredictable. This was another of those problems that persisted for some time, and this time the solution was much more difficult to find. There was something inherently wrong with the game and the endgame issue was just a symptom of that problem.

The issue had to do with the fact that on their turn players could choose to either levy funds or build. If a player chose to levy funds twice in a row, then he collected only half the amount the second time. We noticed that experienced players stopped doing the second levy in a row, and when asked why, they said it seemed like they were getting only half a turn – which of course was correct!

This was compounded by the fact that when the endgame triggered, each player had only one more turn. Some players were caught unawares and had no money, so that player's last turn had to be spent levying funds that wouldn't get spent on anything! We could solve this by giving each player two more turns during the endgame, but this was just a "design band-aid" and involved forcing some players to take a turn collecting 1/2 income just so they could do something on the last turn. It was a big mess.

So we basically had two problems that were interrelated: 1) Some players had an inconsequential final turn, and this would even lead to them having a dramatic defeat simply because their rhythm (between collecting and building) was off by a single turn; and 2) The decision to collect or build was becoming obvious to experienced players, and what had been an interesting decision was now common and predictable.

Collaboration at Gen Con

We actually identified the extent of this problem when we were at Gen Con a few years ago. Many of the team members from Quixotic Games (our game design studio) were there, and we had an impromptu collaborative session to try to solve the problem. We sat in a circle and each team member presented ideas one after the other and each was summarily evaluated by the entire team. Now let me stop to emphasize how much the Quixotic development team loves Canterbury. Some regard it as the best game we have ever created, and they cared deeply about solving this issue, so all egos were erased during this discussion. We all just presented ideas, evaluated them, and (more often than not) dismissed them for one reason or another. The entire jam session took one solid hour, and it is easily one of the most incredible collaborative experiences in which I have ever participated.

Finally, one of the developers, Catherine Weresow, suggested that on the last turn of the game, you could take a "split turn". In other words, you could levy for half the funds, then build only one building (instead of the normal two). For some reason, this idea was shot down, too, but it came back around again during the discussion – and as we discussed it, we realized that this solved the problem of someone having a terrible last turn. If you had no money during the final round, you could simply do a "split turn", raise a good bit of money since the city was so prosperous at that point, then build one building of your choice.

As we were all breathing a sigh of relief that someone had come up with this solution, I had a sudden design epiphany. This "split turn" would solve much more than the endgame problem. What if you could do a "split turn" on any turn? In other words, you would now have three options on each of your turns:

-----Levy Funds (Collect the full amount based on how well the City was doing)

-----Full Build (Build one or two Structures, with this action being mandatory if you levied funds on the previous turn)

-----Tax & Build (Collect half funds and build one structure)

Now the game's rhythm would be varied once again! Some players would be collecting, some would be building, and some would be doing a little of both. Immediately, the number of strategies available in the game had increased. So now I could save up for a big move, or place two buildings at the same time to steal a district, or if I desperately needed to place a building before someone else, I could do an emergency "knock down the farmer's door and grab some loot for the king" and build something right away. There were so many new things you could do with this new system that it literally took us months to playtest the game once again to make sure everything remained balanced.

Final player board


As we continued to playtest Canterbury for several more years with people all over the place, we found that we had gotten to the point where we were running out of things to tweak. People kept trying new things and the system became resilient enough to keep the balance in place.

At this point, we knew the time had come for our team's favorite game, Canterbury, to be published. We had shown it to other publishers over the years and received extremely helpful feedback from each one of them. And after years of implementing that feedback and continuing to hone the game, we felt at last that Canterbury had become more than just another design; it had become an opportunity to take our design company, Quixotic Games, to the next level by publishing Canterbury ourselves.

Thus, we present the gaming community with a design that has been a part of our lives for many years, and we hope it will become a part of yours for a long time as well!

Andrew Parks

The city comes to life in a three-player game on the prototype
Twitter Facebook
Subscribe sub options Wed May 1, 2013 8:00 pm
Post Rolls
  • [+] Dice rolls
Loading... | Locked Hide Show Unlock Lock Comment     View Previous {{limitCount(numprevitems_calculated,commentParams.showcount)}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
    View More Comments {{limitCount(numnextitems_calculated,commentParams.showcount)}} / {{numnextitems_calculated}} 1 « Pg. {{commentParams.pageid}} » {{data.config.endpage}}
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.