GeekGold Bonus for All Supporters at year's end: 1000!
20 Days Left
Calimala is a Euro-style game in which players are members of the guild of merchants in foreign cloth, in Florence, around the 13th century.
The game has a few twists on the classic worker placement genre. The main idea is to have nine main actions in a three-by-three grid, randomly arranged at the beginning of the game. These spaces are connected by "streets", and players take turns placing one of their workers on a street and executing both actions. Since the actions are placed randomly, the possible pairs of actions available change from game to game.
These actions allow players to collect basic materials (wood, bricks and marble); use them to build workshops, ships and trade houses; produce cloth and deliver it to various cities; contribute materials for the construction and decoration of churches; etc.
Another aspect of the game is that workers (discs) are always added and never retrieved: Players have a fixed number of workers (15 with three players, 12 with four, 10 with five), and when placed on an action space, they stack on top of each other. Whenever one disc is placed on a stack, all discs in that stack perform the two actions in order from top to bottom (so extra actions can be triggered in other players' turns).
When the fourth disc is placed on a stack, only the top three discs are activated, while the bottom disc is "promoted" into the city council, triggering a scoring. The city council has 15 seats that are filled in order when workers are promoted (i.e., when stacks grow to more than three discs). Each seat has a scoring tile (assigned randomly at the beginning of the game) that determines which category to score (e.g., most contributions to a given church, most deliveries to a given city). Majority scoring is used for all categories, awarding 3, 2 and 1 victory points to the first, second, and third player respectively.
Where to place your early workers becomes an important decision because if well placed, they will be reactivated by other players two more times.
Seats in the city council also break ties, so when choosing an action space where to place a disc, players have to be careful about which scoring can trigger it and how the balance in the city council will change.
At the beginning of the game, each player receives two scoring cards and secretly picks one that will be revealed at the end of the game and that will score for 5/3/1 points. Each player thus knows of one city or building that will score again at the end (the card they picked), and one that will not score (the card that they discarded).
The game is very tight, and players have to choose what to focus on, especially since with more players it's not really possible to participate in all categories, and these scoring cards add tension, as well as the possibility of bluffing (with people trying to guess other players' scoring cards).
What follows is the story of how I designed this game.
Calimala is my first board game design, although I've been regularly playing board games for more than fifteen years.
When I moved to the UK in 2013, I joined London on Board, a board games club with a few thousand members, with daily meet-ups in various locations around the city. There I met a few game designers and somehow I got the design bug and I started thinking about making a board game of my own.
The basic concept was some variant on the worker placement mechanism in which the available action spaces would change from game to game. The players would then have to come up with a different strategy on each new game.
This is probably the only thing that survived from that inception to the published game.
The idea was to have the action spaces on eight cards in a three-by-three grid (with a hole in the middle). Players would then place a worker between two cards and take both actions. The optimal sequence of actions to achieve the various goals would therefore change from game to game.
The first prototype was just eight handwritten cards, some workers (gray cubes), and a bunch of colored discs.
On your turn, you could either place some cubes on a space between two cards (equal to how many cubes were already there) and take the actions on the cards, or collect all the cubes between two cards. This allowed a continuous flow of play (with no need to collect your workers at the end of a round). The actions on the card would provide discs or convert discs into other discs or into victory points.
It was very boring and uninteresting, but it showed some promise, so one evening I brought it to a Playtest UK meet-up where I played it with a few other designers and where it fell apart very quickly.
After more iterations, I started thinking about a theme and, maybe not too originally, I went for medieval Florence.
The game was still card-based then, with the eight basic cards providing materials like wood and clay or allowing you to hire specialists, along with a set of advanced buildings (more cards) in construction that required those materials.
Players would take actions to contribute materials to the advanced cards, e.g., by taking the "clay" action, I would put a cube of my player color to a clay slot in the building. Once a building was complete, players who contributed to it would score points and the complete building would go into play. (There was some kind of rotation mechanism in which action cards would move in and out, and each new building would enter that rotation.)
This still had several problems, but it's the origin of the buildings in Calimala (like the Cathedral and the other churches).
At this point I took a step back and started studying a bit more in detail the historical period when these buildings were built. There are several Eurogames set in medieval Florence, but none of them really tries to be historically accurate: There were no princes in Florence, and the Medici didn't really trade in spices...
I wondered who built these great churches and why, and I found out about the guild of Calimala.
In the Middle Ages, Florence was a mercantile republic, and the various trades were organized in guilds, whose elder members would take turns ruling the city. The most powerful among these guilds was the guild of Calimala. This was the guild of traders in foreign cloth; during the late middle ages, they were buying rough woolen cloth from all over Europe (England, France, the Flandres, etc.), bringing it back to Florence where they would refine and dye it, then selling it back for a much higher price.
They were producing very high quality cloth, in colors that were not otherwise available in other places. The members of this guild quickly became extremely wealthy, and moving all that gold across Europe and back to Florence was not practical, so they ended up establishing a more permanent presence in the major trading centers where they held their business, keeping the gold there and instead using letters of change to move money, giving birth to the first banks.
Incidentally they also started lending this money to various kings, financing the wars between England and France in that period. (The first bankruptcy happened when the king of England defaulted on his debts.) At home they would then use the money to build palaces and churches and sponsor art works (which would eventually lead to the Renaissance).
The Medici were among the most influential families within the Calimala guild, and within a couple of generations they managed to take full control of the city. (Lorenzo il Magnifico was never formally a prince or a ruler, but with his influence he controlled the majority of the city council.)
Back to the Drawing Board
This research provided some new ideas for elements to add to the game. I decided to focus on the cloth production and the trade network.
I started working on a proper board, with streets connecting thirteen different action spaces, each street with three spots for workers. I didn't come up with the idea of triggering previous players when stacking discs until quite late in the game development; players didn't even need discs in different colors at the time as each street had three slots and by placing in the second or third slot, players would get a better action. More specifically, placing the second or third disc you would do some actions two or three times, while some other actions would be more cost effective.
This allowed for doing more stuff with fewer discs. As the game proceeded, actions became more powerful so that four players with just twelve rounds could be able to complete buildings and fulfill cloth demands from cities.
I had one more building material (stone) and various actions that eventually went away. Each player had an artist meeple, for example, that would move around the city, with an action to move the artist and another action to make an artwork (with a certain number of slots for artwork in each neighborhood of the city).
A "recruiter" action would let you hire an employee (i.e., a card that could be used once at any time matching one of the twelve other basic actions), while a "prestige" action would let you draw a bonus card for endgame scoring.
The scoring was different at the time: Points were awarded right away when delivering a cube to a slot, and extra points were awarded on completions or at the end of the game.
Needless to say, all this was very complicated and playtests revealed many issues, especially with the random placement of action tiles. It was sometimes extremely tedious to do even simple things (collect one marble, then move the artist somewhere with a free slot, finally take the artwork action, etc.). Also, having an artist meeple on the board in addition to the actual workers confused players.
I needed to streamline and simplify; I cut the number of actions down to nine (on a three-by-three grid), and various actions went in and out until I settled on the final ones.
I also simplified the scoring, using majority scoring everywhere. (When an area was completed, points were awarded to the players who contributed the most.) Even artworks were gone, although they eventually came back at a later stage; I instead kept the "recruiter" action that provided an action card to play at any time.
Majority scoring is tricky to get right. Two important design decisions are about when to trigger the scoring and how to handle ties. Some games do scoring at the end of specific turns, but that didn't really fit with the game. I wanted the scoring to happen in a more flexible way because depending on how the action tiles are set up at the beginning of play, some areas might fill up faster than others.
Another important decision is about how to handle ties (more on this later).
I also had another issue: Having fewer action spaces meant fewer slots available to place discs, so I had to revisit the idea of having at most three workers per pair of actions. Instead of having a fixed number of slots, I introduced the idea of placing workers in a stack; in order to keep the stack from growing too much, when the fourth disc was placed on a stack, the bottom disc was removed.
Initially I placed that disc as a "statue" in one of the four quarters of the board (to commemorate the career of the worker who just retired). Each quarter of the board would then trigger the scoring for one category: port cities, trade cities, buildings, and most artwork. Each category would score at most four times per game.
Another concept introduced around this time was that of triggering other players' actions when placing discs. In the initial iterations, when players placed their second or third disc on a slot, they would carry out both actions two or three times in a row. (This helped in maintaining a high number of total actions per game so that there would be enough to make progress on all fronts.) This had a drawback, though, as lots of things could change between one player's turn and their next turn (e.g., in a four-player game, the other three players towards the end of the game could take a total of 18 actions).
That's when I had the idea to invert the flow; now when a player placed a disc on a stack, each disc is activated in order from top to bottom and the owner of each disc performs the actions. The total number of actions per space doesn't change. What's more, the first player to place a disc on a spot will now benefit from three pairs of actions, spread over time. This greatly improved the flow of the game, and players were engaged on everyone's turn.
Playtest, Playtest, Playtest!
Something that came up with more playtests was that players tried to place their discs so that other players could benefit less from their moves, e.g., playing a build action when the owner of the previous discs didn't have enough building materials to benefit from it.
My first attempt to compensate for that was to introduce a "Feld" track, that is, a track used to break ties in scoring; whenever a player couldn't perform an action, they would advance on that track. This maybe overcompensated as players then tried to advance on that track by setting themselves up to not be able to take actions.
With more tweaks and lots of playtesting, I fixed a few problems at once:
I removed the recruiter action; instead players would gain an action card whenever their worker was not able to perform their action. (So that the total number of actions per player didn't change, the action card would let them do another action at any other time.)
I replaced the recruiter action with the "artwork" action (and the "stone" resource with "marble") and added extra slots for artwork in the buildings.
Then I introduced the city council. Now when the fourth disc is added to a slot, the eldest worker (at the bottom of the stack) is promoted to the city council and triggers a scoring. (Scoring tiles are randomly placed during setup in the city council.) In case of a tie, the city council decides the winner (the player with most seats). All this tied together very nicely and made thematic sense.
Playtesting was extremely useful, and every week I would come back home with a new problem and a deadline to solve it before the next playtest session. Slowly but surely, a few more tweaks were introduced over time, such as the white discs which when placed perform each action twice, but are not triggered again later and the scoring cards (which add some more uncertainty, provide a longer term goal during the game, and allow a player to keep contributing to areas that already scored, which was sometime an issue in the last rounds).
By the end of mid-2015, I was quite happy with the game: It played smoothly and within 75 minutes, even with five players. (The total number of discs doesn't change much between player counts: between 45 and 50.)
The game had undergone several playtest sessions, and I was now focusing on writing the rules, including going through a few "blind playtests" (where players learn the game from the rules and play without me, while I watch in silence and take notes). After a few iterations, the rules were clear enough.
In October 2015, almost by chance, I heard about the Hippodice competition when another designer from my playtest group mentioned it in conversation.
I checked online, and I thought that it could be a good way to do some actual blind playtests: Hippodice is a board game club in Germany, and every year they organize a competition for new designers where they play some prototypes for a few months and at the end, in the summer, they provide feedback to the authors.
So I applied (that was just a couple of days before the deadline) and sent the rules, and after a few weeks they asked for a prototype.
The winner is decided by a jury made mostly by German publishers, and every year one or two games among the finalists get usually published. I didn't really think I had a chance, and I was mostly interested in the feedback from the players, so when in March 2016 I got a quick message from a German email address telling me that my game won the competition, I thought it was some kind of joke from one of my fellow designers, moreso because it said that six publishers were interested and they couldn't agree on who should take my prototype, so they asked if I had a preference.
In the following days, a few publishers contacted me directly, and only then was I assured that this was not an elaborate prank. I quickly made a couple of prototypes and mailed them.
Eventually I signed a publishing contract with ADC Blackfire; Uli Blennemann (their main developer, who is also owner of Spielworxx) was very excited about the game and eager to publish it in time for SPIEL '17. Harald Lieske worked on the art, Uli kept me in the loop during the development, and I was able to provide input and feedback.
The game was well received at SPIEL. ADC Blackfire had a large booth with several tables, and Calimala was played constantly on at least six tables at a time during the whole fair. I had the chance to play it a few times with various people, and it was a lot of fun!