Dice are fun. They are probably the oldest game component in human history! As humans we like tactility, we like throwing things, we like winning and losing based on chance. It's fun to give up your control in a game and see what happens. Attaching this to a city builder is nothing profound – it's a fun cycle to collect dice from your city, roll them, and use the results to shape a city.

That is a good entry point, but there are larger ideas that follow. Tactility is about how the dice present to the player; control is simply that dice create random outcomes and what does that do to the idea of a city builder; and design resolution is one way to think about where the overall focus of the player and designer should lie.

As the design got started my thinking with tactility were the pitfalls to avoid, in order to keep the dice rolling easy to predict and process. The first is the number of dice: in my view, players can roll up to about eight dice and still process what’s happening. More than that and it's a big jumble of information that might require time or a computer to sort out, which I wanted to avoid. Typically in Amberspire a player rolls two to five dice at a time, and the early design work focused on how the player would naturally arrive at those numbers, which will be discussed at length in a future design diary.

The other thing I set out to avoid was an overload of information on each face of a die. I knew Amberspire would be using a lot of resources unique to city builders generally, so an early constraint was to say that each die face could only have one symbol on it. This is partly a complexity budget concern, but also one of readability. I don't want to demand the player learn a bunch of symbols and be able to tell them apart in a group on the side of a die.

This general simplicity also communicates a lot about the game and where its goals lie, and what it expects of the player. Looking at the game you can infer that there are quite a few resources, but each building only rolls one die, that die always has six faces and can produce at most six resources - usually fewer! You can easily look at a die and evaluate what it can do for you and whether you want to roll it or not. This hopefully tells the player how complex the game is and what they will have to learn in order to play.

Combining dice with a city builder is a statement of intent, that cities are not perfectly ordered creations that immediately respond to one person's whims and desires. The scant few cities in the real world that were under that kind of control were usually only so for a short amount of time or within other constraints. For Amberspire I am not interested in those kinds of cities. I want the player to feel that while they do have a certain amount of control and input on their city, it will not be a complete.

Dice are one of the major components of this effect. For dice that come from buildings to produce resources, the game is telling the player 'you will never be able to perfectly predict what this building will do'. Despite the extreme level of abstraction in ideas like buildings and resources, the dice represent some real-world messiness.

Other portions of the game using dice and randomness are aligned in this. My goal for the player is while they can control a lot in their city, what isn't in their control begins to create something weird and unique that otherwise would not happen. For example: why is this specific area of the city so under-populated, or beset by competing rust blooms and floodplains, or weirdly cut off from the rest? Through a series of your decisions and dice outcomes that pushed residents, buildings, and terrain into that configuration. You will have a sense of your own city's history grounded by both your decisions and environmental forces.

As you play the game and the city expands, those small decisions and outcomes pile up into a city with a history beneath it.

A city builder could have a limitless amount of detail and complexity. I’ve talked before about the resolution of a simulation, and for me dice are perfect way to put a hard limit on that resolution or complexity. When I say ‘a building rolls a die to produce a resource’, the die producing a random result encapsulates a lot of what a very detailed simulation accomplishes.

For example, a Kiln building might produce a Brick, Earth, or Salt resource when its die is rolled. All the player needs to know is one of those resources got produced, and then decide what to do with it. A complex simulation might have to know how many workers are at the Kiln, what input materials it has, the temperature of the fires, the disposition of the workers, and so on. This is not a judgement on complex games (which I love dearly), simply stating that every game has to draw a line somewhere and I’ve drawn the line for Amberspire to exclude all of those details.

My goal is for the player to focus less on the individual buildings and more on the city as a whole.

I'll have much more to say on dice in future design dairies of course, but this encapsulates where I started with the general ideas of how dice could shape a city building game.

This is the first Amberspire design diary! I'll be posting these roughly every few weeks, going into the backstory, design, and production of the game.

As production on The Banished Vault came to a close, I knew I wanted to make a city builder. I also knew I wanted more dice. Dice are fun to work with and I felt would be a good way to immediately add a lot of interest and variety into a city builder. How exactly that would work was not immediately apparent, however. My goals were not to just make a city builder with dice attached to it, but to think about a city builder in a larger context. Dice produce random outcomes, and how would that affect how you played, or how your city was created? A city also does not grow in a vacuum so I knew I wanted some kind of interaction with the environment, and dice would play a role here. Likewise any other ideas that would present themselves in the future.

In The Banished Vault dice represent one element, your character's faith and their ability to overcome hazards. In Amberspire, what a dice represents changes depending its context. This broad but vague starting point meant working through a lot of different ideas for how they would work.

My starting point was ‘every building rolls a die’. This leads to an obvious question: what are the faces of the die? They could be resources, or whole new buildings, or different kinds of ‘labor’ (also just a resource in a sense).

More questions: When players roll dice, do they roll every die their city can produce? Depending on the scale of the city this could be fine or so large to be meaningless. I wanted the player to build a large city, but conversely a small set of dice to be rolled, so that each roll and its outcome can be followed by the player.

And more: once a player rolls, can they use everything they rolled? If so, rolling dice might feel perfunctory; if not it might feel punishing. The middle ground here is extremely hard to find, less like finding the solution to one problem but a system that will cover every possible outcome, and remain engaging!

I describe this as a tension put upon the player, which is simply giving them mutually exclusive goals or desires. A simple tension for a dice game: it should be appealing to both roll greater or fewer amounts of dice. For the player there's no right answer, just a best guess given their game and personality. Another potential tension: using as many resources as you can while also being efficient and pursuing one goal more directly.

A quick tour of some early ideas that didn’t pan out: dice have a 'quality’ based in a building’s 'morale'; unspent resources affect some kind of fatigue value; dice create groups of ‘labor’ that are combined with resources, etc. At the start of preproduction I didn't know exactly which ideas would work, but these did not express the game I had in mind.

The final version, as the game works now: each building has a die, and each face of that die produces a resource. Resources are used to construct and operate buildings. Some of the die results add a die to the weather roll, which happens every three times you roll a group of dice as the player.

I’ll elaborate on the nuances of this system in a future design diary, but it would end up satisfying the needs of tension, pacing, and approachability. As a pitch, it remains simple: each building rolls a die for resources, and you don’t know what resources you’ll have to build with.