Saturday, October 22, 2005

Book Review: 21st Century Game Design

I recently picked up Chris Bateman and Richard Boon’s new book 21st Century Game Design. Chris is the managing director at the game design consulting group International Hobo (aka ihobo) and has worked on Discworld Noir and Ghost Master. Chris has been kind enough to stop by this website and I’ve always enjoyed his comments.

The major contribution of his book to the dialog on game design is the formulation of a new audience model for game developers and publishers. This is fascinating stuff that certainly got my gears churning.

A more market driven approach
Many of the game design books on the market come are the ruminations of a successful game designer. They are the equivalent of listening to Miles Davis describing in his gravely voice “Sometimes I like to blow the horn like this. And it seems to sound pretty good.” Genius certainly, but such advice is difficult to replicate in any practical fashion.

21st Century Game Design is at its best when it attempts to approach the problem of game design from a perspective that is more familiar to businessman than a creative artist. The fundamental question that the book asks is “how do I make a game design that will sell?”

This is a very different question than “How do I make a good game?” The modern game industry is a Machiavellian place, where naively well-intentioned hard work is not nearly enough to engender success. 21st Century Game Design describes a calculated strategy for getting as many people as possible to play your title. The aim is game designs that are engineered for business success, not ones that succeed through luck alone.

The book provides a thought provoking look at the subject that it tackles. However, it ends up being the start of a much larger discussion. That alone is a worth contribution to the ongoing evolution of the theory and practice of game design.

Whispering sweet cluster analysis nothings in my ear
The backbone of the book is a study intended to provide a better method of categorizing customer wants and needs. A professional statistician would likely take the resulting categories with a grain of salt, but I’m willing to give it all the benefit of the doubt.

The result is a straightforward audience model consisting of four categories that goes beyond the pop concepts of “hardcore” and “casual” that many designers and gamers toss about.

  • Conqueror: The classic goal oriented power gamer, who believes “I win when someone loses”
  • Manager: The more meticulous challenge solver.
  • Wanderer: Someone who treats games like a playground.
  • Participant: Goodness knows, but it involves other people.
Each of these categories is split into a Hardcore and a Casual group. The authors then spend the rest of the book examining the describing how the various groups of game player react to different types of game mechanics and presentations. In essence, the book describes a series of market segments and then discusses how various existing design options serve those segments.
I’ve done a cluster analysis or two in my day and it is worth noting that they are inexact beasts in the best of situations. The Myer-Briggs inventories that underlie much of the books assumptions are based off hundreds of studies using very large populations tracked for many years. The likelihood of the book’s first generation audience model being correct in all its details is approximately nil.
However, that does not limit the value of the attempt. Some of the highlights include:

  • First, it calls out the dark and inbred history of modern game designs. Most of what we consider great games were created by a freakish group of Conqueror miscreants and are poor foundation for serving the needs of the broader population. Publishers, you need to get down on your hands and knees and pay ihobo gobs of money to beat this particular message into your thick, risk averse skulls.
  • Second, by presenting the current audience model, designers are encouraged to think about their target customers and the customer’s needs in a more rigorous fashion that is uncommon in the game industry.
  • Third, a fascinating topic for additional research has been broached. I hope that ihobo and other more academic researchers pursue the topic of audience models vigorously in the coming years.
Is an audience model the right way to go?
As much as I like how an audience model encourages us to think of our target customers, I worry that it only a piece of a much larger puzzle. I’m going to step away from reviewing the book for a moment and look at some of the broader implications.

An audience model is, at its core, an extension of the marketing concepts that drove much of the mass commercialization of music and movies from the 1940’s onward. There are some critical assumptions involved that could be quite dangerous if you are attempting to tap into new opportunities. Some of the implicit assumptions are as follows:
  • There are big broad market segments that are homogeneous and exist (in varying proportions) across territories.
  • These market segments are based on basic human psychology and are therefore quite stable.
  • Game distribution is a one-way push model. If publishers execute in a technically competent fashion, passive gamers will consume it. If a game is sent out and properly promoted, and it meets the generic psychological needs of the target market, then it will do well in the marketplace.
This is a highly defensible perspective on the game industry that fits the classic packaged goods models of entertainment. Within a mature market that requires its participants to play a game of ‘king of the genre’ with highly predictable consumers, the use of such a model is bound to gain a few extra percentage points on the revenue charts.

The book briefly touches upon the economic implications of this model. Each market segment has both an overall revenue and profitability associated with producing product for it. Hardcore gamers might only sell 500,000 copies of a game. This puts limits on the amount of money you should expect from and therefore spend in developing a hardcore title. This is quite reasonable.

Classic problems with an audience model
However, troubles come into play when smart publishers use this model as a technique for maximizing their revenue. They begin to create titles that appeal across multiple market segments. Marginal titles are culled and the portfolio is optimized for maximum profit. Historically, what happens here is as follows:

  • Marketing dictates ‘required’ elements for success. In order to properly cull your portfolio, you need criteria derived from your audience model. A pop record might have a check list that includes: “Pretty young girl + hip-hop inspired beats + epic vocals + sexual lyrics.” The book takes a stab at identifying common genre mechanics that appeal to different audience segments. This is only a short step away from creating a game specific check list. EA is already working towards such a check list with their latest “1 to 2 elements of original game play + 1 major brand + best in class artwork” formulation.

  • Originality is sacrificed because it does not fit into the ‘winning formula’: Games that are outside of the winning formula are instantly dismissed. Often, there will be a list of acceptable game mechanics that are acceptable. When an original game concept does not have an obvious match either the game mechanics or the buckets available in the audience model, the risk adverse action is to toss it and go with something safer.

  • Small market segments are underserved: If a market is not a major ‘acceptable market within the established audience model, it is unlikely to get much attention. There is no room for the long tail in simplistic audience models.
This model is very new to the game industry, but it has been around in a variety of forms for many decades in other media markets. The results are interesting and predictable. Rigorous application ends up with the majority of the publisher dollars funneled into high profit segments of the market. Consolidation trends are accelerated while low profit segments are starved and eventually die off.

In the short term, this firing of undesirable customers by the entire industry results in dramatic industry growth. In the long term, it leads to stagnation. It turns out that all those little low profit markets are the source of the periodic creative renaissances that the larger market requires to grow its revenue base.

Game specific issues with using an audience model
Complicating the picture is the simple fact that games are not traditional media like movies or music. Ernest Adams makes the telling point in his introduction to the book that there are a dozens of unique classes of games. The part that fascinates me is that these games differ radically terms of functionality, not merely content.

Most music is functionally identical. There are differences in taste, but the core psychological benefits that are derived by Jazz listeners are not so different than those derived by listeners of Metal. A game of Animal Crossing, on the other hand serves a radically different purpose than a game of Risk, not merely a different audience.

Who uses the game, how they use the game, where they use the game and the benefit they derive from the game are unique to the each genre. Of course you can’t play a game of Animal Crossing when you have a group of five friends over. It isn’t a multiplayer game.

I come from a background that deals with the concept of ‘product design’, not media marketing. Product design looks at the specific ecosystem of a class of users and identifies unique gaps or opportunities for creating value in that ecosystem. These opportunities are generally composed of an intricate webs of psychological, economic, relationship based needs.

We need to stop thinking of games as disposable entertainment that, like a faceless porn movie, merely services our generic psychological needs. The reason games are so hard to classify is because they entertainment tools, not merely entertainment experiences. Every tool has a different use within a very specific ecosystem.

Some examples:
  • Pokemon acts as pre-teen social networking devices and lives within the rarified ecosystem of GBA’s portable network.
  • Nintendogs appeals to Japanese consumers and other city dwellers who are unable to own a real dog.
  • Galactic Civilizations serves a niche of passionate players burnt by MOO3, but desperate for the glory of MOO 1 and MOO2. They are older gamers that need an entertainment tool that can be paused both mechanically and psychologically when the wife yells that dinner is ready.
The direct application of audience profiling as a concept formulation technique will never directly result in any of the games above. Such models are too vague, too generic. Where in the spectrum of audience markets would you find ‘dog lover?’

Part of a bigger picture
Games, as entertainment tools, are different products than disposable experiences like movies or music. An audience model is still a useful technique, but it must be applied properly. I see as it a secondary technique the can help refine a game concept that stems from an ethnographic or anthropological study.

In short
  • Identify a unique market opportunity or under served niche within an ecosystem.
  • Use an audience model and other profiling techniques (interviews, observation, etc) to identify critical goals for the final product design.
  • Build your game design around those critical goals.
Designers should avoid using audience models as the only determination of economic feasibility and instead rely on market-sizing techniques specific to their game concept.

Next Steps
Chris and crew have kicked off a wonderful discussion and I’m very excited to see where the book goes in subsequent editions. Some suggestions from the peanut gallery include:
  • Additional studies done with more statistical rigor. I want to trust the model that is put forward as reproducible. Ultimately, I would love to see the research side of this book grow to as compelling as business books like “Good to Great” or “Built to Last” by Jim Collins.
  • Exploration of the applicability of audience models to the game design process. How can it be used to enhance both the creativity and success of a product design process?
  • Exploration of the business implications of an audience model: This model is useful for game designs, but it has serious ramifications for the industry as a whole. What are the positive aspects of its application and what are the pitfalls that should be avoided?
Conclusion
Buy this book. You are doing yourself a serious disfavor as a game designer if you don’t understand the central concepts involved in the proposed audience model. The first few chapters alone are worth the price of admission.

Don’t expect the book to answer all your questions. Instead treat it as one of the first vigorous discussions about designing for the modern business-centric game industry. The basic attitude of measuring and asking real customers about their preferences needs to infect the entire industry.

Equally important is that you question the basic assumptions behind the proposed theory. Is it the right philosophy to inform the industry’s future investment strategy? EA is already following, if not the specifics of this book, the general spirit of an audience model driven strategy. They are quite successful. The simple market-based approach has worked for movies and music and seems to also work for games. Is there a better path for the game industry? Or is this good enough?

Our young industry is at the beginning of a very lucrative discussion of how to make game and why games should be made. When books like 21st Century Game Design promote a seductive message of profitability that sparks the interest of both the money men and the creative visionaries, they can shape the future of the entire industry.

I’ll leave you with this delightful quote from a recent article on the radio industry:
“[Lee] Abrams pioneered systematic audience research and "psychographics," connecting people's lifestyles to their listening habits. He invented a music format called album-oriented rock, or AOR, which in the 1970s shifted the music industry's focus from singles to albums and showed radio execs how to hold listeners and attract advertisers - to make money in the new, boundary-free
world of FM.

But his success had a cost. The rise of AOR was the beginning of the end for the brief, storied era of free-form radio and iconoclastic DJs - "some guy in a basement in Brooklyn, burning incense and playing whatever he pleased," as Abrams describes the late-'60s scene. The format ushered in such airwave dreck as classic rock, teen pop, and … there's no easy way to say this …
smooth jazz.”[1]

The good folks a ihobo are not the first to implement the concept of an audience model in the game industry. That honor belongs to the larger game publishers of the world. However, by writing a book on the subject, they are encouraging all of us to discuss the concept and its ramification in a public manner. Perhaps we can improve on the theories that drive the decisions that occur behind closed doors.

Take care
Danc.

References