Here in the western part of the world, the game called "Mahjong" we're mostly familiar with is the solitaire variant. This label is since it doesn't have anything to do with the real traditional Chinese game by the same name, a set-collection game of skill, strategy and calculations, with a bit of chance involved, for four players. The only thing they have in common is the tiles used for the game.
With that said, the tile-collection mechanism of "solitaire mahjong" has been popular for decades, especially in video game form, and it is quite good on its own merits. There is something fundamentally satisfying in searching for a couple of matching tiles to remove and slowly demolishing a huge structure of tiles! You can even start with the tiles laid down in many different shapes to greatly alter the feeling of each game. There must be a reason these video games have been so popular ever since their invention, after all!
This is how the story of Dragon Castle started. During an IDEAG event — IDEAG being a series of Italian events where game designers show their prototypes to players and publishers — Luca Ricci approached me with the idea of making a board game based on "mahjong" (meaning the solitaire). I found the idea very appealing for all the reasons stated above and more, so after a couple of email exchanges we started to work separately on our ideas (me in Milan at the Horrible Games headquarters, Luca in Rome), comparing our prototypes every couple of weeks to decide where to bring the development next.
The core of the project has always been to use that tile-collection mechanism. Traditional mahjong tiles are divided into "simples" (numeric tiles going from 1 to 9), "honors" (dragons and winds), and "bonus tiles" (seasons and flowers).
One of the earliest decisions we made was to turn the "simples" into factions ("farmers" for the "bamboos", "merchants" for the "circles" that I've always seen as coins, and "soldiers" for the "characters"). The soldiers, in particular, would have to be swords in increasing numbers, like the other suits, instead of the Chinese characters. This apparently insignificant change would give us two advantages: We could start creating a setting for our (otherwise very abstract) game, and it would also make the tiles easier to distinguish for western eyes without losing that "mahjong" feeling.
For the prototype, we used a standard mahjong set with our artwork stickered on top
From here came the idea of the "crumbling castle", that is, a place that people (i.e., the tiles/factions) are abandoning in search of a new home — a premise that has remained in the game ever since. The early prototype of the game, though, was very different from what you see today. You had a more structured player board divided into different areas: the city, the fields, the barracks, and so on. Depending on the kind of tile you collected, and the area of your realm you placed them in, you could trigger different instant effects or work towards different end-game scoring criteria (waging war against your neighbors, removing farmer tiles to harvest, etc.). We wanted the special tiles to have thematic effects, too, but none of the ideas we tested was really convincing.
A playtest of this early version of the game
This was a phase when the development of the game proceeded slowly. It took me a bit of time to realize why, but the reason none of the iterations of the game using this system really convinced me was that we were building a quite structured and complicated Eurogame that was based on a simple and instantly recognizable mechanism with a distinct family appeal. We did want strategy in the game, but this was not the way to go.
This realization finally became crystal clear to me one day when I casually showed the game to my mother. She instantly said, "Oooh, mahjong", and for once she was really interested to know more about a game (even though she's not really a gamer at all) — yet I knew she would never, ever want to play a game like that. That's when it hit me that a core mechanism with such a broad appeal had to be used in a more accessible game because otherwise it would feel like a waste (to me, at least). Both Luca and I were struggling to find the right direction, and with many other projects on our hands, the game was slowly falling behind in our respective development schedules, but I always kept it alive in a part of my brain.
The project finally found new life when Hjalmar Hach joined the Horrible Games team. I showed him the latest prototype, and I explained the general direction I wanted the project to go, that is, a more accessible, puzzle-like game that would better fit with the core mechanism and its history. I was hoping that a pair of fresh eyes would see something new, something that me and Luca, with all our development history and layers of old ideas and game versions, found difficult to see. And luckily, that was the case! After a few hectic brainstorming sessions, we basically made a whole new game based on spatial objective cards that players needed to complete by building certain patterns with tiles of certain colors (a bit like Ticket to Ride). This is when the game really started to take off.
Me, Hjalmar Hach and Luca Ricci playtesting a semi-final prototype of the game
We quickly went through many revisions. The game became more and more accessible, and we started to finally really enjoy playing it. We had no more areas on the player boards — just a single area where you could place your tiles, score the objectives, then place more tiles on top of them to score more objectives. It now felt like you were building your own little castle. Players would compete both for the tiles they needed to take from the central castle (now finally called the "Dragon Castle") and the available objective cards. It had strategy, and it still was a very abstract game, but the improved accessibility made it way more enjoyable.
When I say many revisions, I mean it!
It sounds like the pieces of the puzzle were finally falling into the right places, right? We thought the same...yet many parts of the game were still not convincing. The card system felt overcomplicated, for example, and the level of pattern recognition capabilities required to effectively play the game was still higher than what I would have liked. When you completed an objective card, you were forced to turn some of the tiles face down to avoid snowballing, but this mechanism was very prone to analysis paralysis. (If I complete this objective, this tile turns face down and I can't complete this objective anymore, but if I complete this one first instead... You get the idea.) There were many other small details and nuances in the system that were needed to fix some of the problems we had, and they all contributed to make the game more complicated than it needed to be.
We couldn't decide whether we were in a situation where a few tweaks here and there would do, or whether we still needed major changes. Many of our playtesters liked this version of the game, despite its problems, but we ultimately felt that we needed to be bold. The game needed to go through another major metamorphosis. Striving to finally achieve that ultimate level of accessibility for which I was the main advocate, we got rid of the objective cards altogether. You would simply score points when creating sets of tiles of the same color. We made a first playtest with a very basic version of what would become the game as it is now (with no special powers and no special objectives), and something clicked. This was the way to go.
This is when the player boards started taking their final form
The game didn't really work, and it was a bit shallow compared to the previous version, but it finally felt right. We spent a few weeks working on this version of the game, making slow but steady progress. We re-introduced cards in the form of the Dragon cards, which are common "building" objectives for the end of the game that are much broader in scope than the old ones, yet much simpler to understand, and the Spirit cards (common special powers to break the rules and trigger combos). These cards kept the essence of what was good in the old versions of the game and integrated it into this new form, while also offering a great deal of variability and replayability to the game, with that variability being supplemented further by the many different castle layouts in the final rulebook (and the fact that you can also create your own layouts). A major improvement came with the addition of another key component: the Shrines.
The evolution of the Dragon and Spirit cards
Without the objective cards to constrain and guide your building possibilities, players were feeling a lack of long-term purpose in their decisions. When we added Shrines, which increase your score but also prevent you from building anything else on top of them — basically locking one of the spaces on your board forever — we made the puzzle aspect of the game as interesting as it was with the previous version, while keeping the game simple enough that you could understand what it was all about in five minutes and you could play it with your friends (gamers or not), partner, parents, children, dogs, cats — finally, a game for everyone.
A very close-to-final prototype
The game took its last step towards the published form when we made one final addition: a dynamic scoring for the sets of tiles. In Dragon Castle, you gain more points if you complete larger sets, but the score increase is not 100% linear. This creates a tension between riskier strategies (trying to go for larger sets for lots of points, but at the risk of other players counter-drafting the tiles you need), and quality strategies (completing many smaller sets that give you fewer points, but allow you to build a higher castle more quickly and build a lot of Shrines, especially on the upper levels of your castle where they are the most valuable). You can also mix and match these two extreme opposites, of course, and you also need to make sure you have enough Shrines to put the strategy you choose to complete fruition!
A few of the hand-written sketches used during development
I'm very happy with how the final game turned out, and I'm sure the same can be said for my partners in crime: Hjalmar Hach and Luca Ricci. If you'll be in Essen for the SPIEL fair, we hope you'll pay us a visit in Hall 3, booth Q106! Dragon Castle will be there waiting for you!
The final tiles of Dragon Castle; the mahjong feeling is definitely there!