Page: DesignArticle.Article201007110816AM - Last Modified : Thu, 16 Aug 12 - 2751 Visits

Author : Eric Pietrocupo

The following articles try to give guide lines to avoid the following situations. These are just some examples, but there are many more possible outcomes.

• For one of my game which had a production system, I once designed a series of heroes that the player could hire. So I took the time to dress up a list of the heroes I wanted and made sure that their stats and their special abilities were balanced. After the first playtest, the production system did not worked well and had to be changed. Which forced me to scrap all the heroes I designed.
• I had a game design idea similar to "Race for the galaxy" or "San juan". After designing a draft of the rules, I was going to design the cards to be able to playtest the game. The problem is how am I going to be able to design 100 unique cards when I vaguely know at this point how the game will work.

Starting a new game design

Let's go with the basic. How do you design a new game? First you try to define your idea my summarizing it in one of two sentences. Then, I find that it is a good thing to define the elements of the game: Space, object, actions, rules, skills and luck. This theory is explained in chapter 10 of the "art of game design". Here is a link to the summary

Art of game Design: chap. 10 -Some Elements are Game Mechanics

Then when you have an idea how you game is going to look like, you start writing your first rule draft. Of course, you will not write all your rules because you don't know yet how the game is going to work and there might be some details to complicated to define at this point since you have no working game.

When your rules are defined, you want to playtest your game to get some feedback which is going to give you some basic information helpful for continuing the design of the game. It is a good thing to rush in prototype which mean make it as fast as possible. This way you are going to receive feedback pretty quickly and you are not going to get stuck in the infinite loop of design. The infinite loop of design is where you are constantly designing and redesigning stuff without ever progressing because you want everything designed before testing it.

Now, According to the type of game you are designing, you might get stuck with the design loop hole. First I'll try to explain the problem. To playtest your game, you need a prototype. In this prototype, there will be many "data" or information to design. For example, you might want to design event card's effect, unit stats, etc.

The problem is that it is hard to define this information when you partially know how your game works and especially when you know that near the beginning of the design your game is going to change a lot. So why would I design 50 unique event cards that changes the rules of the game when I know the rules are going to be redesigned. That would make the cards obsolete.

First, it depends on the kind of design you are making. If your early design does not imply a not of unique special ability or information, you might get a working prototype without having to define much data. This was my case with Fallen Kingdoms where I only started with a map, units and city tokens.

But if you are making a game like san juan, you are screwed because the only things there is in the games are the cards. So you need to design 100 cards before being able to do a single playtest. Designing these cards will be hard because you don't know the complete area of effect the cards could have and again, the rules can change at any time.

Possible solutions

So here I try to give some solution with some example on how could the design be continued. I think the basic idea is to go from general to specific. Define a series of steps or phases and jump on the next one when you know most of the stuff from the previous step works. Here is an example:

Let say you want to build a house and you have all the technology and the knowledge to build. Let say that you also never lived in a house or an apartment. You lived in the wilderness all your life up to now. So the big question is, how are you going to design your house? How many floors, how many rooms, what should be their size, etc.

The wrong way to do it is to build the whole house, live in it a few weeks, destroy want you don't like and build again. Repeat the process until you get what you want. This is very time consuming and you have actually not idea how much cycles you need to complete before getting the design you want.

A better solution would be to first build the floors. Only the floors. Even if you only have floors, there is still a lot of things you can test. For example, you can walk on it, you can place some furniture, etc. So you will be able to analyze space and movements in the house.

If at this point, you could realize that your rooms are too small, you can easily extend the rooms by enlarging your floor platform. You could also easily add rooms. But if your walls where already built, enlarging a room would be much more complicated because you would need to destroy the walls, enlarge the rooms and build the walls again.

After testing the floors and making appropriate changes you will reach a point where everything seems to work well. Now it the time to add wall. Walls gives you much more options. You can now have door, you can have decoration on walls, you can have windows, etc. There is now plenty new features to test and you don't need to bother about the floors anymore because they have been previously tested.

So this is the idea, you start with the general and basic stuff that you know it is going to work or that is going to be easy to test. Then you can add layers gradually and test each of them.

My Dragon Realm Example

My first example is my dragon realm design which I am currently working on. The game consist in moving dragons on an hex map in order to control territories and attack other players. The battles are resolved by rolling dice and playing special ability cards. There will be 15 types of dragons with many types of special ability cards, this is most of the data that needs to be designed.

Now the first thing I could easily test is, does the hex map works. At this point, the game should not be fun. My primary objective would be to test if the movement, the placement of new dragons and the attacks works on the map. Even I said I wanted to test attacks, I do not necessarily need to resolve combat. What is an attack? It simply the idea of moving and removing pieces of the board. The way I determine how the dragon will not is not important. I could roll a die, or choose which dragon dies.

In order to complete step, I don't need that much information, I just need the movement speed and the sex of the dragon to know hoe they move on how they reproduce. If I want a more fair game, I could give a strength to each dragon and resolve the combat as roll 1 die + strength, highest wins. This way, I could counter balance movement speed with combat strength. This would be a temporary combat resolution system.

The second step would be to test the combat system without the ability cards. Since I know the maps are working, I do not need to test the combat with the map. I could simply make duels between randomly picked dragons without having to move them on on the board. That would increase the amount of combat tested within the time you are testing. At this point, you would need all the stats of the dragons to resolve the battles. So this is the new data to design Once the battle works, you might want to test the whole game together.

The third step would consist is testing the special ability cards. Now since I have a working game, it is easier to design a set of card that will change the movement and the combat rules of the game now that you have the core of the game working. So dressing up a list of the special abilities and ranking them in power level would be the data to define here.

San Juan like Example

A San Juan like game is much more complicated to design because you need everything right from the start to play. I think the solution is still to go from general to specific. This can be done by using simplified cards at the beginning of the design and then complexifying the cards as the design evolves.

San Juan has a very odd way to make the cards circulate and there is of course the role system. So first I would test the roles, placement of buildings and the production of good. This way, I am going to test all core mechanics of the game which will be the place where most of the rules are.

But the main problem is still what are you going to place on the cards? Well what do you need on the cards to test the elements listed above? In order to place cards into play, they need to have a cost. So the first thing to do is to give every card a price. So just make cards with random values. Distribute them evenly since we don't know yet how the distribution should be.

You know that there is going to be production buildings and that this is very important for the core of the game since this is how you make money. So you should start by only designing the production buildings. Everything else should be blank cards.

Depending on your kind of game, you could design at most a dozen of cards and make many duplicate cards to get the total amount of cards you need to make the game playable. Then as the game evolves, you can replace card with new cards that has different special abilities. Designing a dozen of cards would not be too complicated and will not take too much time in case you want to change them.

Of course, the game will not be fun because it is useless to build a building that has no special abilities. But at this point you want to have a game that is running. Then once your roles and production system works, you will get new special ability ideas while playing the game. Take note of these ideas and they will help you for the design of the new cards. For example, you might realize that it could be cool to have a cards that reduce the cost of other buildings, or that gives you victory points, etc.

Then when you have more and more cards, the real challenge would be to balance the price of the cards to make sure you really pay for what you get, and determine the distribution of the cards in the deck. For example, since production buildings are part of the core mechanics, you might want to have a lot of production buildings in the deck.

Variants are easier to design

This is where you realize that variants are much easier to design than creating a game for scratch. It is one argument that could be given to people that object to variant design.

The reason why design variant is easier is that you already have a set of data to work with. That helps, because you are not floating in emptiness, you have something you can walk on.

That could be compared to buying a house. You have a working house, but it does not exactly fit to your needs. So you could decide to change the floors and the change the wall colors. Or you could make larger modification like adding or removing walls. Even if it could be a lot of works, it is still easier to do that building a house from scratch.

Also you already have something to test with. You already have a house and you can use it right now to test what does not work. While if you design from scratch, you have nothing.

So I hope this article was illuminating and will help your design progress.

---

Back to the Game Design Article List