Simulation and Data Exploration

Let’s take a look at what 16-bit interfaces could do.

A great simulation game begins with just a handful degrees of freedom and explodes from there.

Behold the grandeur that is SimCity for the Super Nintendo.

If you’re familiar with SimCity (1991), skip ahead to Data Exploration, below.

     

On a flat plane of pixels, you have the choice to:

• Bulldoze a feature.
• Build a road.
• Build a mass transit unit.
• Build a power line.
• Build a park.
• Build a residential zone.
• Build a commercial zone.
• Build an industrial zone.
• Build a police department.
• Build a fire department.
• Build a stadium.
• Build a port.
• Build a coal plant.
• Build a nuclear plant.
• Build an airport.
• Build a special reward building.

That’s 16 discrete choices to be made on a board of some 10,000+ spaces in two axes (Map 61 has 10,514 units of land).

That’s a lot of freedom.

The optimization objective of SimCity (1991) is total population.

Population is a function of density and number and variety of zones.

Factors preventing density are traffic, lack of power, pollution, crime, employment, tax rate, and disasters.

Factors preventing number of zones are scarcity of capital, budget, and land.

Factors causing density include building airports, ports, stadiums, lack of crime, lack of pollution, lack of traffic, employment, and proximity to special buildings (like the mayor’s house).

Factors causing crime include density and the lack of police station coverage.

Factors causing pollution include traffic, industrial zones, airports, and coal power plants.

The density of residential and commercial zones is caused by the balance between number of residential units to industrial units to commercial units, and, the absence of negative factors like traffic, pollution, crime, and high taxes (that last one is a distinctly Anglosphere concept).

Greater densities are associated with higher land values.

Higher land values cause greater budget and more available capital.

(Let’s set aside the overlapping zone hack).

Data Exploration

Data exploration is centralized in a the Information Panel.

The memory that the developers had was fairly limited in the nineties. The techniques of recoloring or adding a visual filter on top of the map wasn’t quite there just yet. (This was before Starfox, and look how far we came after that.) So the developers overlayed the map directly on top of the map, and this enabled the user to see around the edges. It’s not the best way to explore the data generated by the simulation, but it’s adequate.

It presents information about the factors on the ground: the composition of the zones, power, density, growth, traffic, pollution, crime, land value, and police/fire services, in the context of a single plane.

Not one number is presented on screen. Data is represented spatially as a heatmap, Red for max, Green for min.

There are graphs. They’re extremely basic, and the only units that appear on the screen is time. There are no rates of change. And again, the number of zones, along with pollution, land value, and crime is shown. The user can chose between the short run (10 years) against the long run trends (120 years), and no other choices.

There is an overview screen, which is really a table of descriptive statistics about the city, and not presented with gridlines, sparklines, or any time series. It delivers a snapshot for a single point in time.

It returns, as primacy, the number and ratio of residential, industrial, and commercial zones. There is a count of developed versus undeveloped zones. This is not returned as a crosstab. There is a count of things in the city, and, a count of the number of units in the city and how they’re used: park area, forest area, openland area, and water area.

The primary dashboard is on the primary interface screen. The population, money, and RCI demand bars (quite possibly the most important heads up data display of the nineties) are prominently located in the upper right hand quadrant. The RCI bars are easy to read, the higher the bar, the greater the demand for that kind of zone. The lower the bar, the lower the demand. The colors, Red for Residential, Blue for Commercial, and Yellow for Industrial, correspond to the color of zones in the simulation. In other words, the design of those few hundred precious pixels was part of an overall system of design that was central to the simulation itself.

 The secondary dashboard, the City Evaluation, contains the most direct suggestions in the data architecture.

That dashboard is divided into Public Opinion and the deceptively called ‘statistics’. The dashboard itself tells you what’s wrong with your city in the form of a stacked list. And then, that optimization objective – population, is displayed with a its delta, assessed value, and category. Score gets less prominence below it. This, here, is a fairly more advanced balanced scorecard.

The Data You Generate, The Natural Laws in the Simulation, The Data You See, The Choices You Make

When you engage with a simulation like SimCity (1991), you’re generating data. When you build 12 residential zones, 6 commercial zones, and 6 industrial zones – and don’t build a power plant – and just sit there, staring at flashing lightening bolts, you’re producing data. When you do hook them up to power, and have a mass transit line joining them, you’re producing data.

And the simulation responds to those actions by pushing them through its axioms.

A part of the fun of a simulation is to deduce the natural laws in the machine. And a part of that fun is to look at what the game designers are trying to tell you (or hide from you) in the interface. SimCity has it’s own truth which is written right into its program. This post has described an entire system of reasoning, where I have controversially asserted causality (the horror!).

The data you see has to be carefully considered. The makers of this game, so long ago, made many decisions.

• They chose to allow you to explore the relationship between space, zones, positive factors and negative factors.
• They chose to allow you to see graphs trended over time, but just a few, without numbers.
• They chose to give you descriptive statistics, but not to drown in them.
• They chose to give you an evaluation.
• They chose to give you a headsup, primary performance dashboard of population, money, and demand for various zones in the R-C-I bar chart.

The data they wanted you to see what designed to cause better choices.

The designers didn’t want you to build 100 residential zones and screaming with rage “why isn’t anybody moving in?” The designers wanted to guide you through a sequence of learning, which would cause people to discover the natural laws contained within the program.

Once you learned those natural laws, the challenge became how to maximize and minimize the various factors in the game to design better outcomes, and ultimately, optimum outcomes. What is the maximum population possible? Once a design system is through through and considered in its entirety, the search for the maximum amount of land ensues.

The Data You Generate, The Natural Laws of Nature, The Data You See, The Choices You Make

As a decision maker, when you make choices, you generate data. If you chose to write three bullet points, on a given topic, instead of ten, that’s a choice. If you chose to spend $10MM USD on a campaign featuring desirable people using your desirable product in a desirable environment, desirably, you’re making a choice. If you chose to spend $5MM in digital to support that campaign, you’re generating data.

Instead of a simulation, with very clean axioms and functions, you’re dealing with the laws of nature.

Markets respond to different messages, in different mediums, with different types of behavior that can aggregate in different ways. Ultimately, the manifestation of those choices causes the aggregate demand to curve to twitch. If it twitches in your favor, the optimization objective of profit is realized.

The data that you see isn’t always structured against those natural laws.

Because we have to turn that data into an experience that aligns both with the data you throw out into the world, how to world chews that up and spits it out, and, the final leg of how that data will reinforce a true conclusion or a false one.

The Architecture of the Data You See

There are a number of choices that the designers of SimCity (1991) made that are worthy.

• They restricted the reporting for real-time figures to just five pieces of data summarized in 3 units, population, money, and zone demand.
• They enabled the user to see the data in the context that they were making decisions in.
• They did not present tables.

Even though the experience was in 16-bit and severely limited, it was usable and beautiful. They had the advantage of designing the underlining forces of nature within with simulation. They nailed the alignment between how it worked, and what they wanted you to see.

If we can only data scientists could be half as effective as they were.