The following is a selected excerpt of Chapter 2 from 21st Century Game Design (ISBN 1-58450-429-3) published by Charles River Media.
Why is game design often overlooked as an important factor contributing to game sales? Perhaps because when most people in development companies talk about “good game design,” they mean “game design that produced a game I really like.” This sort of subjective validation of game design is of no use in business, which thrives on repeatable methods based around capturing a target audience—the market. Unable to see the profit resulting from “good design”— especially since many allegedly well-designed games fail commercially— most businessmen ignore design entirely.
Design is not suggested to be the only (or even the primary) factor in the sales of a game. Marketing, for example, is hugely important in making a product visible in a crowded market. Similarly, the sales of a game depend greatly upon the budget for development. A game developed on a budget of $100,000 should not be expected to achieve sales figures equivalent to a game developed on a budget of $5,000,000. However, mechanisms such as word of mouth transmit individual opinions of a product, opinions that will be swayed by the design content of the game.
Therefore, we face a great need to make game design relevant to the business side of the game development process. Once the ragtag market has stabilized, we will have plenty of time to pursue the artistic side of game development, but for the time being, that is a luxury we cannot afford. We would not see inventive filmmakers like the Coen Brothers were it not for commercially motivated film makers like Spielberg and Bruckheimer, because the commercial success of a medium clears the way for artistic expression, not the other way around.
Demographic Game Design
A first step is to consider a criteria for success—what is a successful design? Notions such as elegance, a criteria famously applied to the design process by Ernest Adams, are great aspirational concepts, but less useful for business purposes. Systemic production rules, such as Noah Falstein's “400 Project,” provide neither a success criteria nor aspiration and are useful mainly as a means of provoking discussion.
The concept of demographic game design is that game design inherently targets an audience, and therefore the success criterion for a design is how effectively it satisfies the needs of that audience. This factor is not directly related to sales figures and is not intended as a means by which to consider the success of the game as a whole—only the success of the design. If the target audience is satisfied by the game (which can be determined by appropriate sampling techniques), the design can be considered a success.
However, before these criteria can be applied we must know the demographics that are available to be targeted, so the first step in demographic game design is to study the audience. If the game designer is to act as a player advocate in the development process—as if they are an elected politician reflecting the diverse needs of their constituency—they must first acquire a useful audience model.
A Warning on Statistics
It has often been said that you can make statistics prove anything you want, and this is true—provided the people you are talking to lack the critical faculty to see the flaws in the presented argument. Nonetheless, statistical principles are a vital part of modern business and science. Quantum mechanics, which all modern computing depends upon, is essentially statistical in nature, and even the concept of “species” is not a Platonic ideal, but a Gaussian distribution of diverse life forms arranged into clusters, which we choose to term “species” only by convention.
The most important thing to remember when dealing with statistics of any kind is that showing a correlation (a connection between two events, details, or tendencies) does not prove causality; it is merely a clue to something interesting. For example, in one famous incident a statistician found a statistically significant correlation between babies and storks in Switzerland. Storks were consistently nesting on houses with newborn children. This fact did not prove that babies were brought by storks, of course, and on investigation it was discovered that the houses with newborn babies were kept warmer than other houses. This extra warmth attracted the storks. The lesson is that statistical correlations tell you nothing of the underlying causal mechanisms.
The other important aspect of statistics is that statistical data about a group tells you nothing about individuals in that group. For example, it is well known that the majority of college students drink alcohol, but this statistic does not allow you to know whether any given college student drinks alcohol. Reasoning about the general tells you nothing about the specific. The advantage of statistics is that whatever does not average itself out to insignificance in a given set of data is a tendency that can be counted on. For example, statistical analysis has demonstrated to the movie industry that roughly 50% of the audience of a profitable film return to see a sequel, allowing for strategies involving producing cut-price sequels for short-term gain.
Market Clusters And Audience Models
The notion of a market cluster (or market segment) originates in marketing. In recent years, with the advent of narrowcasting channels such as specialist TV stations and personal e-mail, cluster analysis has fallen out of favor in marketing, but the technique still has value in other disciplines. The basic principle is to analyze a data set containing information on a particular group of people and look for common traits that when taken together define a coherent group or cluster.
For example, the vacation travel market identified three distinct clusters: the demanders, whose priorities are exceptional service; the escapists, who want to get away and relax; and the educationalists, who want to see new things, experience new cultures, and so forth. These categories emerged from a cluster analysis on data taken from a pool of vacation makers; this data was then sorted by a clustering procedure. Note that the categories were named after the cluster analysis—the names were created to capture the feel of an abstract cluster of people who shared some common traits.
This cluster analysis approach is one of the more formal ways of producing an audience model, but a simpler method exists that anyone can apply: observation and hypothesis. In essence, you observe many different people from the audience (or look at statistical data in general) and draw a hypothesis from the observation. In science, this would be followed with an attempt to validate the hypothesis. Alas, in the games industry, many hypotheses are treated as a priori facts. However, provided you remember that such a hypothesis is only a working assumption and needs to be tested to determine its value, we find nothing wrong with building an audience model in this way.
Hardcore and Casual Split
This is the most basic audience model at use in the games industry today. It is in essence a consensual hypothesis—that is, a hypothesis which the majority accept as factual—and almost all people working in the games industry know what is meant by the Hardcore (or Core) market and the Casual market. Some data at use in the industry might confirm this split, but since no formal definition for each group exists, it remains in essence a hypothetical model.
The essence of Hardcore players can be summarized as follows:
Capcom characterized the Hardcore approach at the start of Resident Evil on the GameCube (Capcom, 2002) as “Mountain climbing.” This ego-neutral characterization was used at the start of the game to determine which players were Hardcore in their approach and therefore required greater challenge. Selecting this option ran the game at a higher difficulty level.
On the other hand, Casual players can be summarized as follows:
Capcom characterized the Casual approach in the GameCube Resident Evil as “Hiking.” Players who selected this option played the game at an easier setting, allowing them to have more fun and enjoy the experience without the greater emphasis on challenge (which often equates to greater emphasis on repeated failure). The full wording of the sorting question at the start of this game is as follows:
Question: Which best describes your opinion about games?
I. MOUNTAIN CLIMBING—Beyond the hardships lies accomplishment.
II. HIKING—The destination can be reached rather comfortably.
The value of this somewhat unusual question over a straight choice between “Easy” and “Normal ” (or “Easy” and “Difficult”; or “Normal ” and “Difficult”) is psychological. A Hardcore player faced with “Easy” versus “Normal” will pick “Normal,” but a Casual player is equally likely to pick “Normal,” thinking that choosing “Easy” makes them deficient in some way. The choice between “Normal” and “Difficult” is likely to cause some Hardcore players to select “Normal” (on the grounds that “Difficult” settings are for replay value) and then complain that the game is too easy. Finally, a choice between “Easy” and “Difficult” is likely to mislead both Casual and Hardcore types as they try to decide which of these two options is the normal setting.
In principle, the advantages of this approach are that tailoring the gameplay to the audience—and sorting the audience correctly—improves the reception of the game, equating to stronger sales. Unfortunately for Resident Evil on the GameCube, the slow-burning sales of the platform somewhat interfered with the actual unit sales. However, informal observation shows that Hardcore players were satisfied with the degree of challenge they received in “Mountain climbing” mode, while Casual players had no difficulty completing the game in “Hiking” mode.
This example clearly shows the value that even a simple audience model can have when used to approach the design process. The sorting question was a novel approach, and although it provoked some confusion in game-literate reviewers, the basic approach seems sound and could be refined to a more subtle approach.