Tải bản đầy đủ - 0trang
Chapter 8. Getting into the Design
Go through the list yourself, reprioritize the features, and make the team defend why each feature should be in the game.
We guarantee it will be a sobering exercise, at the least. It will probably also create some true anger among some members of the design
team, which is where the tough love comes in. As a team leader, you'll probably have to cut one or more features that the team
considers crucial or to which they have formed an emotional attachment. Be ready for histrionics and fireworks, and use the tools and
information in this book to back up your decisions.
It may seem that throughout this book that we're hard or "down" on designers and design teams. In truth, we
are, but not out of any disrespect for individuals or their responsibilities. To make these things work right, they
have to be the brightest, most educated, and versatile people in the process. Designers have the hardest job in
the industry because, at the end of the day, the game succeeds or fails financially and as a game by what they do.
If you think having $10 million in development funds and the jobs of 30-50 on your shoulders is an easy
responsibility to carry, you're in the wrong business.
[ Team LiB ]
[ Team LiB ]
Acquisition and Retention Features
"I try and think of what every feature does in terms of acquisition and retention. More specifically, I try and rationalize
every feature on how it ties a player to (1) other people within the game environment and/or (2) an in-game reminder
of a valuable game accomplishment. There are dozens of other variables, but these are the big-ticket items in my
Most computer and video games have only one distribution problem: how to acquire customers—that is, how to get the customers to
walk into a store and buy the box. This includes retail hybrids, as most publishers don't provide an online distribution or multiplayer support
solution for customers; they depend on the players to do that for themselves, or for one of several online hybrid gaming portals such as
GameSpy or The Zone to do it for them. In neither case does online in-game support become a factor. Hybrids today are a product, but
they are not yet a full service in most aspects.
Online games in the persistent world (PW) category, being both a product and a service, have a unique problem: They not only have to
acquire customers, but they also have to retain them. If a subscription fee is involved, the retention factors have to work for a minimum of
some months or, preferably, years.
It makes plain sense, then, to know just who it is you want to play and pay for the game and what features those people want that will keep
them coming back month after month. If you accept the market demographic niche definitions found in Chapter 1
This document was created by an unregistered ChmMagic, please go to http://www.bisenter.com to register
. it. Thanks
The Bartle Player Types
Dr. Richard Bartle
was the co-creator of the first MUD in 1978. Over the years, as MUDs proliferated on university mainframes and
eventually as commercial products, he noticed certain types of common play behavior in MUDs and made the first attempt to categorize
those behaviors. The categories he defined were achiever, explorer, socializer, and killer, and he explained the behavior patterns in an
article titled "Hearts, Clubs, Diamonds, Spades: Players Who Suit Muds." The article was first written in 1995 and gained widespread
recognition quickly. Over the years, Dr. Bartle has periodically edited and revised the article to keep it current.
See Dr. Bartle's full article and the new introduction to it inAppendix D; used by permission of the author andThe
Journal of Virtual Worlds.
Here is how Bartle described what he saw:
The four things that people typically enjoy personally about MUDs are:
i. Achievement within the game context.
Players give themselves game-related goals and vigorously set out to achieve them. This usually means
accumulating and disposing of large quantities of high-value treasure, or cutting a swathe through hordes
of mobiles (i.e., monsters built into the virtual world).
Exploration of the game.
Players try to find out as much as they can about the virtual world. Although initially this means mapping
its topology (i.e., exploring the MUD's breadth), later it advances to experimentation with its physics (i.e.,
exploring the MUD's depth).
iii. Socialization with others.
Players use the game's communicative facilities, and apply the role-playing that these engender, as a
context in which to converse (and otherwise interact) with their fellow players.
iv. Imposition upon others.
Players use the tools provided by the game to cause distress to (or, in rare circumstances, to help) other
players. Where permitted, this usually involves acquiring some weapon and applying it enthusiastically to
the persona of another player in the game world.
So, labeling the
four player types abstracted, we get: achievers, explorers, socializers, and killers.
This was the first attempt to define general player and gameplay types in a virtual world, and these motivations still hold up well in today's
market, where games with 20,000–100,000 simultaneous players are no longer unknown or even rare. These types more than adequately
explain the gross motivations of the general player base and provide a good framework from which to start designing a massively
multiplayer (MMP) game world.
Moreover, players rarely exhibit just one form of play behavior; they tend to mix and match styles and change behaviors over time.
Several years ago, Erwin Andreasen put up on the web a survey of questions, the answers to which would allow MUD and PW players to
rate their own play styles and discover how much of their play time was as an achiever, socializer, explorer, or killer. Table 8.1 includes
some of the results from the Bartle Quotient Survey
on the web, showing how players rate themselves in various combinations of
those four general categories created by Dr. Bartle, sorted by response rates, combinations of play in the categories, and percentages of
the total respondents. The responses to the survey gives you an inkling of how players see themselves and their own playing styles in
All stats used by permission. For more information, see Erwin S. Andreasen's "Measuring the Bartle Quotient"
online test at www.andreasen.org/bartle/.
This document was created by an unregistered ChmMagic, please go to http://www.bisenter.com
to register it. Thanks
Table 8.1. Survey Results from "Measuring the Bartle Quotient"
From the "Measuring the Bartle Quotient" web site, http://www.andreasen.org/bartle/, March 4, 2002 at 2:28pm EST.
Abbreviations: E= explorer, S= socializer, A= achiever, and K= killer. The numbers in parentheses represent the total number of
respondents scored in that percentile. The Combination Play column represents the total percentage of players scored by
combination players, such as socializer/explorers (16% of the total respondents). The Combinations of Three column represents
an even finer gradation of play, such as 12% of the total respondents showed elements of socializer, explorer, and achiever in
play style. As of the date and time this sample was taken, there had been a total of 111,926 respondents included in this
Despite the fact that the survey is unscientific, it has been the experience of the authors that these percentages fairly closely match the
reality of the customer base in current commercial PWs, with the exception of the "importance" of the killer class of players. The
importance of that class has been overblown by a vocal minority and by PvP combat-oriented designers. For example, check out the
specific survey results for the top five commercial PWs in Appendix C "The Bartle Quotient Survey and Some Results," and you'll find
few people rating themselves as pure killers in most games.
Lineage: The Bloodpledge, which claims four million subscribers, was not included because fewer than 30Lineage
players had rated themselves in the survey at the time the sample was taken.
The takeaway from this exercise should be obvious: 55–60% of players in both the general population and those playing for-pay games
classify themselves primarily as socializers and explorers. These are people who form and join guilds and teams, spend quite a bit of time
in-game chatting or engaging in social events, and wander about the world, exploring its secrets. For three of the five commercial
games, 20% or fewer of the players rate themselves as killers, with 26% for UO and 25% for Dark Age of Camelot, two games that are
attractive to the killer class of player due to the faction-based conflict inherent within their designs.
What this says is that, even if you build a game heavily weighted toward the killer classes via PvP and faction or team/guild conflict,
chances are you're going to be attractive to only 20–25% of the total player base. The dichotomy of this is that most of the PWs that
launched in 2001 or are being developed today for launch in 2002–2003 are heavily weighted in design toward the killer class. This is
likely to cause increased churn rates for the socializer and explorer classes, whom the killers look at as their intended victims by virtue of
that same design.
Unless you're firmly set on appealing to one player type and you're comfortable with the thought that such a design may greatly limit your
subscriber total potential, the idea is to have a balanced design that appeals to a number of player
Features: Acquisition and Retention Over Time
For purposes of the following discussion, we're going to assume that you want to appeal in some fashion to all four Bartle types, in an
attempt to maximize the subscriber total of your game.
Now that you have a general idea of the size and scope of the basic customer groups, you can take the design treatment discussed earlier
and begin to match up the vision of the game with the features the players will want to have available.
Exactly what those features are, however, is a matter of some controversy within the industry. Some general features that apply to all four
player types, such as secure player-owned housing and no-conflict safe zones, are unanimously accepted among the design literati as
required to attract and retain customers for the long term. After that general feature set, however, few design teams actually give much
thought to applying features to player types. They have a general understanding of the features they want in the game and which features
have worked or not worked in other games. They rarely dig very deeply into player needs and motivations or list them out by the Bartle
types, or any other analytic measurement for that matter. Thus, the features list tends to be scattered and a bit incoherent for the design as
a whole, and that usually comes back to haunt the team
[ Team LiB ]
[ Team LiB ]
The Themis Group Player Satisfaction Matrix
"As we talked with various developers and publishers," says Alex Macris,
CEO of The Themis Group, an online game consulting and
full-service support company, "we began to notice a trend: The teams had no firm grasp of who the potential customers were or how their
feature set appealed to those customers. They just assumed that everyone played for the same reasons and wanted every possible
feature. Trying to make the teams understand the differences between the general player types and what they wanted in a game was
frustrating at times."
Disclaimer: One of the co-authors, Jessica Mulligan, was president of The Themis Group at the time this book
was being written.
To get a handle on this problem, Macris devised an unusual and innovative solution. What he realized is that, although many people in the
industry knew what the separate features were, no one had taken the time to plot them out in a logical, easy-to-see manner. He also knew
that development teams understand issues and concepts best when arranged as data points in a matrix; trying to just explain features to
them as concepts made things more difficult, not less. So he sat down and drew out a basic matrix chart that correlated the Bartle player
types against the game features they needed or wanted at various stages in their lifecycle for an online game.
The format Bartle arrived at was one of those simple solutions that the industry has needed but somehow missed for years. Table 8.2
shows how Macris expressed the player lifecycle and satisfaction factors in a
Table 8.2. Basic Player Satisfaction Matrix
The Player Satisfaction Matrix is copyright Themis Group 2001–2002 and used with permission.
Down the left side of the matrix are the first three periods of the average player lifecycle, as noted inChapter 1. Across the top run the
major activity types that parallel the four Bartle types: socializers, achievers, explorers, and killers. The boxes that cross- reference each
lifecycle period with a player type list the features those players want or need at that stage of their "career" in the game.
What The Themis Group does for a game is first fill in all the relevant features for the "complete" PW, with the features prioritized from
highest to lowest needed, as shown in Table 8.3.
Table 8.3. The Player Satisfaction Matrix, First Step
Examining Table 8.3, you'll see at the left are the first three phases of the player lifecycle. At the top are the four main PW player types
according to the Bartle index. In each box are the general features each player type needs or wants at each phase of the lifecycle.
Within this matrix, any experienced design team member at Themis can go through the filled-in features boxes and highlight the features
that exist in the game, as denoted by the bolded features in Table 8.4.
Table 8.4. The Player Satisfaction Matrix, Partially Filled
When the work is finished, the team has an
invaluable tool; it can tell at a glance which features are currently planned or existing in the
game and which features are missing for specific player types.
This method has proven to be of enormous help in focusing design and development on just what should be in a game at launch to help
acquire subscribers and which features are needed to retain subscribers long term but can be added at a later date, say, through an
expansion pack or server-side patch.
The trick is to know what feature set is needed for your game to function competitively at a base level in the marketplace and build from
there. Table 8.5 is a partial list of features we believe necessary for the acquisition and retention of subscribers at all levels, from hard-core
gamers to mass-market non-players. Bear in mind that while many acquisition features are also retention features, a retention feature does
not always work for
Table 8.5. Basic Acquisition and Retention Features
[ Team LiB ]
This document was created by an unregistered ChmMagic, please go to http://www.bisenter.com to register
. it. Thanks
[ Team LiB ]
The Critical "New Player Experience"
Designers love complexity; nothing gives them more of a sense of satisfaction than watching interlocking game mechanics work
together or providing an interface that can do anything and everything, including walking the dog and making coffee. There is also an
element of competition among designers to provide more features than the previous guy did, under the assumption that more is better.
What many designers forget to plan for is how long it will take a new player to learn how to operate the necessary features in a game
compared to the average new player's patience level. Designers play games for years before they get the financial go-ahead to work on
their own game. Then they spend dozens of months building their game. Naturally, they know how everything works (or how everything
is supposed to work). By the time they get funding approved for their game, they are strangers to the sense of wonder and frustration that
a new player experiences.
Space shuttles are wonderful, but the number of people qualified to pilot them is small. The average middle-school youngster could drive
you to a hospital ER these days—seat, mirror, ignition, gas, brakes (maybe), and you're there. If a new player has to be fairly adept at
using most of the capabilities in your everything-under-the-sun feature set to grow a character's stamina, wealth, skills, and so forth fairly
quickly in the game world, then the only players your game will retain over time will be from the hard-core segment. Remember: These
customers are not buying a car for $15,000—they're test-driving a virtual world that costs $25–$50 to enter and $12.95 a month to rent a
life in (and you're giving them the first month's rent as an incentive to stay). You have 30 days, and often less time than that, in which to
Having nothing but hard-core players can still be a winning formula, assuming that within the next 3–5 years, nobody develops an
interface that walks the dog, makes coffee, kills spiders, and then takes out the garbage. If you could keep 50% of the hard-core gamers
playing your PW for a year, someone would probably come along and try to seduce your shuttle pilots with their newer, better feature set
and interface. As you will see, what you drive is important (getting there may be fully half of the fun, indeed), but so are where you go,
what you do when you get there, and with whom.
Another thing that is often ignored is whether the new player experience is compelling enough and entertaining enough to make the
player stick around, or whether it is a frustrating experience that causes him/her to churn out and go looking for entertainment
"Churn" stands for "change-turn." It is most often used in the industry to describe players who change from
customers into non-customers.
The quality of the new player experience is your key retention factor. The player has already decided to try you out; if he/she can't
figure out the interface easily, or the environment is so hostile the player can't succeed at something early on, you'll probably lose the
player in the first month. Historically, the churn rate of new players from online games, after garnering the hard-core players in the first
three months of the game being available, is well over 80%, and in some cases, exceeds 90%. Overall, long-term retention (two months
or more) varies, but 40% retention of all those who try the game is pretty standard.
There are a number of reasons
for this churn; fixing these reasons during the design phase should be of paramount concern.
Following are some of the worst offenses:
Complexity of the interface— If the client interface is cumbersome to use, employs non-standard commands or methods
for everyday issues, or employs so many buttons and capabilities that even a shuttle pilot would have trouble figuring out how
to use it effectively, you can expect only a small, dedicated core of players to bother learning it.
Complexity of the game mechanics— Sure, lots of interlocking moving parts are cool and represent a triumph of design;
they also tend to limit your subscriber base to those players with the determination and sheer grit to work through them.
Experience has shown that this is a limited market. It is fine if your game world is interesting enough that hard-core players
spend time on Internet sites sharing special hard-won information, but new players should not need to go to those sites to
master the basics quickly.
A hostile new player environment— If the new player represents nothing more than a crunchy snack to experienced
players, or NPCs, or the environment, you can expect a lot of churn. For example, UO's environment on launch was
completely hostile to new players because player killers (PKs) had free rein. Those who were trying the game tended to get
"ganked" by the PKs repeatedly. This is not conducive to creating "fun" in the minds of most players and caused tremendous
churn from UO in the early days. It was only when EA reined in the PKs and provided more of a safe haven to allow the new
players to get to
know the game that subscription growth began to rise again. It is important to provide protection to the
new player, to allow him/her to get to know the game and learn how to gain protection.
An unsatisfying initial experience— If the new player can't achieve something memorable pretty quickly or be rewarded in
some fashion, the game is simple drudgery and unentertaining. Some quick reward, one that preferably teaches something
about the game or interface, can provide an accomplishment "hit" and keep the player hooked for more. SSI's old Gold Box
series of Advanced Dungeons & Dragons (AD&D) computer games solved this one by starting the level-one characters within
a few points of level two and advancing the player within minutes of starting through some simple, easily completed actions.
A lack of information or training in-game, in context— Most players never read the game manual; they just jump right in
and start playing. If they can't figure out how things work or have easy access to that information in-game, they are soon gone
to try the next guy's product. Our history has been that we tend to dump the new players into our complex games without
even an interactive tutorial to get
them started; we just leave them to sink or swim on their own. The first game to
actually provide a tutorial was AC; players were led through a series of training actions to get to know simple commands such
as walking, running, chatting with other players, arming weapons, and so on. Few online games since have followed this
example, though one or two have retrofitted tutorials post-launch.
Additional devices, such as interactive tutorials, tool tips, NPCs standing at new player entry points to provide information, or
even human players or gamemasters (GMs) to provide an immediate welcome and offer of help, are not luxuries anymore.
Some of the other factors that are listed, such as the complexity of the interface and game mechanics, can be blunted with
thoughtful and timely tutorials, training, and assistance to the new player.
Consider every feature you want to design into the game as a potential block to new subscribers, and make sure you build in
mechanisms to let the new player learn the basics quickly and easily and survive long enough to start enjoying the
[ Team LiB ]