Productive Design

1
Productive Design

• Christopher Natsuume
• Creative Director
• Boomzap Entertainment

2
Topic

• Stated Goal
• Define better tools and processes to allow
  projects to embrace innovation while minimizing
  waste.

• Translation
• Spend less money
• Sell more games

3
Games are not Art. Games are Products.
Products are designed for Customers. Anything
else is a Hobby.
Success of Design and Production is Measured in
Product Success
Product Success is Measured in Profits
4
Money
price
market share
revenue
volume
market size growth
PROFIT
materials
variable costs
labor
expenses
fixed costs
5
The Bad NewsYoure not Truly Innovative

• True Innovation
• Mechanics
• Distribution
• Control
• Genre

• Breakout Products
• Pac Man
• Civilization
• Quake
• Bejeweled
• Ultima Online
• Myst

6
The Good NewsTrue Innovation isnt Required

• Iterative Innovation
• Story
• Graphics
• Sound/Music
• Characters
• Setting
• Weapons/Units
• Quality

• Profitable Products
• World of Warcraft
• Rise of Nations
• Age of Empires
• Spyro
• Halo
• Far Cry
• Luxor
• Jewel Quest

7
Customers

• Customers are real people
• Give them names
• Introduce them to your team
• Discuss features as What Jenny wants
• Know what they do and who they are
• Where do they live?
• Where do they shop?
• What movies/music/magazines do they like?
• How much time do they have? What else do they do?
• Test for THEM NOT YOU
• What you think is unimportant - what they think
  is.
• Test to their difficulty level.
• Play the other games they are playing.

8
Jenny

• DINK Double Income, No Kids
• Transitioning from single life of going out and
  partying to more mature and sedentary lifestyle.
• Busy social life, but still has time for
  television, movies, and other entertainment
  media.
• Sometimes needs brainless entertainment to
  relieve personal and professional stresses.
• Plays games, but would never call herself a
  Gamer
• Played video games on console machines at a
  younger age, may still do so casually.
• Likes to zone out when playing games
• Does not like violence or blood prefers simple
  bright colors and pretty designs.
• Broadband Capable and Internet Competent
• Has a fast broadband connection
• Completely capable of doing things online, but
  does not think of herself as a power user
  online.
• Quotes
• Im not a gamer, I just like to play games.
• I dont really shop for games, sometimes I just
  play what came with our computer
• My friends tell me about a lot of games, they
  are more into them than me.
• Games help me relieve stress from work and
  school.

• 34 years old
• Administrative Professional
• College Educated
• Apartment Renter
• Long Term Relationship
• No Children

9
Competing vs. Winning

• Competing Features
• Good enough is good enough
• Rip stuff off
• Refuse to innovate unless necessary
• Document by reference

• Winning Features
• Nothing is good enough
• Innovate from every angle
• Refuse to copy unless necessary
• Document by prototype and diagram

Your Top-Selling Product Will Have Three Winning
Features
10
Competing Features

• In this order!
• What does the audience expect?
• What can you remove?
• What is the standard of quality
• What can you build to reuse?
• What can you license?
• What can you outsource?

11
Winning Features

• In this order!
• What does your customer want?
• What is the competition not providing?
• What can you compete in?

12
Your Team

• Use work teams
• Let them choose leaders
• Allow them to move teams
• Make them work together
• Hire adaptive people
• Know what they cost
• Compute task costs
• Know your margin per unit.
• Ask will this feature pay for its development?
• RememberFull cost includes integration and
  testing

13
Productive Research

• Amateurs Play Games
• Read the Previews
• Play for fun
• Follow the path
• Play what you like

• Professionals Research
• Read Reviews
• Read Forums
• Watch others Play
• Break The Game
• Find the Rules
• Take Notes
• Play Everything

14
Productive Brainstorming

• The Steps
• Takeaway/Silent Brainstorming
• Round Robin Input
• Discuss and Advocate
• Prioritize
• Choose Champions
• Dont Look Back

15
Productive Prototyping

• Hard Stuff First.
• Winning Features First
• Ugly and Fast trumps Pretty and Slow
• Iteration, Iteration, Iteration
• If its not on screen, its just bullsht
• Tools need Prototyping, Too!
• Youre not done prototyping until its fun.

16
Productive Scheduling

• The Stages of Production
• Plan
• Prototype
• Pipeline
• Produce
• Polish
• I will bet
• You are Planning and Producing Too Much
• You are not Prototyping and Polishing Enough
• You dont even schedule a Pipeline phase
• Planning ? Documenting
• Prototyping Changing

• Things You Always Forget to Schedule
• Researching other games
• Playing your game
• Waiting for feedback
• Revising work
• Vacations

17
?? (kaizen)

• kaizen n. constant improvement
• Eliminate waste from
• Defects
• Overproduction
• Transportation
• Waiting
• Inventory
• Motion
• Over-processing
• The antithesis of the Post Mortem
• Post Mortem You are Already Dead

18
???? (pokayoke)

• poka-yoke n. idiot proofing
• Remove all places for data errors (i.e.
  validation in Excel)
• Remove the monkey touching steps
• Design things that cant be broken by art
• Build in systems and DISALLOW exceptions
• Object name enforcement in all disciplines
• Numbering standardization
• Automation Task Generated when
• 3 Mistakes
• Anything done 5 times
• Whenever hand-touching adds nothing

19
Prototyping vs. Testing
Cost of going back
3000 2000 5000 1000 7000 2000
3000 2000 5000 1000 7000
3000 2000 5000 1000
3000 2000 5000
3000 2000
2000
Step 1 2000
Step 2 3000
Step 3 5000
Step 4 1000
Step 5 7000
Step 6 2000

• Testing costs inherently INCLUDE rework cost!
• Prototyping is a bargain!

20
The Weight of Assets

• Assuming
• each step is 1000
• There are 10 units

• When you add in the Asset Multiplier you are
  CRUSHED BY THE WEIGHT OF ASSETS
• Design and gameplay should drive the game, not
  the need to maximize sunk costs in assets!

21
??? (heijunka)

• heijunka n. production smoothing
• Goes against logic batching reduces setup
  costs.
• However
• Feeds the pipe better
• Gets to the end better
• Allows for real testing and gameplay
• Allows for Vertical Slices

22
Multiple Vertical Slices
Unbalanced Development BAD
Balanced Development GOOD

• This takes more time in the short run!
• RESIST the lure of the assembly line

23
Unit Design Distribution
Bleems
Splorts
Zoinks
A
A
A
(cannon fodder)
B
B
B
(infantry)
C
C
C
(cavalry)
(artillery)
D
D
D
24
Problem Solving 5 Whys

• Toyotas 5 Whys
• Situation Welding Robot Stopped Working
• Initial Proposal Replace Fuse
• Problem Breakdown of welding robot
• Why 1 Fuse melted due to overloading
• Why 2 Bearing lubrication inadequate
• Why 3 Oil pump did not draw enough oil
• Why 4 Metal shavings sucked into pump
• Why 5 No filter on pump intake
• Countermeasures Install oil filter
• Effect Breakdown frequency decreased dramatically

25
Productive Planning

• Write Less
• Nobody reads anything longer than 10 pages. Ever.
• People read lists.
• People look at pictures.
• Learn to use Powerpoint.
• If you cant describe in a paragraph, draw it.
• If you cant draw it, prototype it.
• Read everything you write twice and remove half
  the words.
• USE REFERENCES
• Screenshots with drawings
• Google Images
• ALT-PrtScr CTL-V
• Communicating design ? Mass emailing design doc
• Communicating schedule ? Mass emailing schedule

26
There is too much wasted space here (and on the
bottom) pull the edge of the well RIGHT UP to
the edge of the screen. We dont need to see any
of the background here, and we DO NOT need 2
frames. Make the WELL frame and the MAIN FRAME
the same frame, and save more space. The frame
should lie DIRECTLY on the edge of the game
screen. THEN make the well as much bigger as you
can with the extra space!
Please use the 11 wide version of the Main Screen
for your test!
The logo should be bigger
The score should be bigger
The black outlines on all of the frames look
jagged in non-filtered mode because of the high
contrast. Maybe we can see what it looks like w/o
the black edging?
These buttons need to be twice as long. Adjust
the frame to fit if necessary
Make the yellow part semi- transparent smoky
glass, as in the stuff PH put together.
HOWEVER I want to try leaving the well 100
transparent just like the edges of the screen -
not dark. So DIFFERENT from the work PH did.
Please make these semi-transparent as in the
old version.
27
Handling Data The Rules

• Never write a lua script- generate them
• Store it all together
• Use Matrices for .jpg, .ogg, etc.
• Physics / Material / Sound collision matrix
• Weapon/Vehicle matrix
• Unit/object AI matrix
• Nothing is Permanent
• Assume you will scrap 50 of your data and
  replace the other 50
• Assume you will change every variable in the game
  20 times
• Remember
• Amateurs think about Rules
• Professionals think about Systems

28
Variables The Rules

• Fewer, more important
• Hardcode Nothing
• Normalize Everything
• If its not noticeable, make it a constant
• If it is muddy
• divide by 10
• use doubling
• make a constant

29
Game Data Formatting
Individual Records BAD
One Table Better
Normalized Table Best
30
Systemizing
Systemized Design GOOD
Hardcoded Design BAD
31
Designing UI The Rules

• Designer develops by USE , not by design
• Build it with grey boxes on day one
• Change daily
• Use only text and boxes
• Dont even think about art
• Only build the art when you stop changing it
• If you will have experimental code-driven stuff,
  add that.
• When the USE has proven the design
• The designer SHUTS UP
• The artist LEAVES THE DESIGN ALONE

• Artist builds art from the main in-game screens
  out
• In-Game Screens
• Game end/summary screens
• Popups
• Game setup screens
• Main Game select Screens
• THEN everything else
• When Building
• Build One by One - Always.
• Think Localization
• Use TEXT not textures
• Use the same stuff. Over and over.
• Less is more.

32
Game Logo
BOOMZAP
DISTRIBPARTNERLOGO
Sound
Mute
Full
Music
Mute
Full
Fullscreen
ON
OFF
Hints
ON
OFF
World High Score
ON
OFF
Cancel
OK
33
Designing Levels The Rules

• No Levels till you have a library of Gameplay
  Nuggets
• Level designs are ONLY a list of gameplay nuggets
  in time and space.
• Your Friends
• The BGB (Big Grey Box)
• Unit Sizes
• Scale Standards
• The Texture Crosshatch
• NO ART till its fun and playtested
• Every level described by
• this is the level where you
• you know, the one with the
• When you add art HARD STUFF FIRST
• Main distinguishing features
• Animated or code-requiring stuff
• Story-specific necessities

34
Sample Gameplay Nugget

• Used DOZENS of times!
• Bridge across ravine
• Shallow water between 2 islands
• Bridge across river
• Two hills with rope bridge
• Walkway in storage room
• Walkway through large laboratory
• Etc.
• Tested
• How much space was cool?
• How much cover was cool?
• What about more than one safe way across?
• What about using a high sniping-spot on one side?
• What about using vehicles?
• What about having reinforcements?

Cover
Crossing
Open
Cover
35
basic units
Faust about 32 Faust is about 3 2, but with
the bounce in his step, we expect him to be about
2 units in height.
Fausts Footprint 1x1 Unit The space that we
expect Faust to take is about 1x1 unit on the
ground. This dictates much of the information on
the sizes of the places he can go and interact
with.
Basic Unit The basic unit 2 This is the basic
building block of the entire world most things
will be sized so that it can be broken down into
hunks of roughly one unit.
Smallest Free Camera Room 16 Units Square This
is the smallest room that we can comfortably have
a free roaming camera in. Most rooms in buildings
will be at least this big.
Smallest Room 12 Units Square This is the
smallest size of room that Sully will usually go
in, period. Any room smaller than this will need
to be substantially longer in dimension, and will
become somewhat like a hall.
Smallest Walkable Ledges 1 Unit The smallest
edges that Faust would ever be able to walk on,
they are usually placed nest to walls to
facilitate their use.
Difficult Walkway 2 Units The size of a walkway
put in the air that would be a difficult path, or
a balance challenge.
Tiny Platform 1 Unit Tree-stumps, fence posts,
etc. will all be of no smaller than this size.
This is the smallest thing Faust will be expected
to stand on, and we will mnot use them in places
where falling could possibly be irritating or
painful.
Normal Hallway 4 Units The average hallway in
the game that allows a free camera, these will
still look quite narrow and are used only in
places where the hall is supposed to look a bit
narrow.
Comfortable Platform, Wide Hallway 8 Units This
is the size of a comfortable space for Faust to
walk on in the air, or a nice wide hallway. This
should be a very common size for the game.
Difficult Jumping Platform 2 Units This is the
regular size for a very small jumping platform
a quite tricky challenge for Faust.
Common Platform 2 Units This is the size of a
common platform in the game Faust will very
regularly find spaces about this big to play on.
Massive Hallways 16 Units This size space will
feel extremely large and spacious when Faust is
in them. This is a common size for places where
Faust will need to fight or perform complex
activities while on a platform or in a long hall.
36
jumping
Double Jump Button Press Available
Walk Overs Things that are 1 unit or lower can
easily be walked over by Faust. The player need
not do anything but continue pressing in the
appropriate direction. The PC makes a little hop
to clear these obstacles.
Jump Up to Stand on Object The PC can Jump up
high enough to stand on any object from 1-2 units
high. Usually they will be 2 units exactly. The
PC may Double Jump onto anything from 1-3 units
high. Usually double jump units are 3 units
exactly.
Jump Up to Hang on Object The PC can Jump up
high enough to hang from any object from 2-4
units high. Usually they will be 4 units exactly.
Anything 2 or under may not be hung on the
player will jump onto these smaller objects. The
PC can Double Jump up high enough to hang from
any object from 3-5 units high. Usually they will
be 5 units exactly. Anything under 3 or under
will not be hung on when double jumped at the
player will jump onto these smaller objects.
Basic Jumping and Double Jumping The basic Jump
takes the PC 4 units. This can be extended to 6
units by Double Jumping. The player can press the
jump button anywhere during the jump to make it a
double jump until he reaches the apex of the
jump, at which point, pressing the button has no
effect.
4 Units This is a basic jump distance the PC
can comfortably clear this distance with a single
jump. These are common for filler action, but are
never serious hazards.
6 Units This is a long jump that allows a PC top
land on the other side on their feet with a
double jump, or to grab the other side with
their hands on a single jump. These are very
common hazards.
8 Units This is a very long jump, which can only
be done with a double jump, which allows the PC
to grab on to the other side with their hands.
This is a much more difficult challenge, which
has a heavy element of fear involved with it.
10 Units This is an unjumpable distance. Usually
places that are unjumpable are at least 10 units,
if not more, so that the player can clearly see
that without aid of some kind, they will not be
able to make this jump.
37
Level Progression Development
20
1
1
5
7
13
18
20
Theme Water Nuggets A C D E J H
Theme Vehicles Nuggets D E F J K Z
Theme Jumping Nuggets Z G H K T L
Theme Ranged Weapons Nuggets K O P E W
Theme Etc. Nuggets A B F R G T
Theme Basic move Nuggets A B C D E F

• Hard stuff First!
• Hit the Highlights THEN fill in!

38
Decoupling Levels from Space
39
Level Design Rules

• Reuse, Reuse, Reuse
• Change Appearance
• textures
• lighting
• sound
• Change story/reason
• Change Navigation
• Open routes
• Fill with water/gas/lava
• Constrain time
• Add vehicles
• Add new inhabitants

• Simplify, Simplify, Simplify!
• If you dont have it prototyped dont build
• If you cant interact dont build
• 5 cool interactive things 20 non-interactive
  things
• 3 distinct units 20 art-change units
• Prune until pruning ruins the design

40
Secret Math

• 3, 5, and 7 are magic numbers
• usually good sometimes great usually great
  sometimes bad
• The value of diversity is subject to diminishing
  returns.
• As your user base grows, they approach infinite
  cleverness and infinite stupidity in equal parts.

41
Secrets About Your Customer

• Nobody cares about your Wrapper UI unless it
  sucks.
• Nobody plays the last 25 of your game.
• They turn the music off.
• They play at work.
• The Four Laws of Movement
• If you dont tell them to go there, they wont.
• If you need them to go there, they wont.
• Tell them to go there, and theyll hate you if
  they cant.
• If they break the game by going there, they will.

42
Secrets About Critics Reviewers

• Reviewers only play the first few levels.
• Reviewers read other reviews.
• They want it easier than they say.
• Previews are there to destroy you.
• You should listen to your mother.
• They actually want to write good reviews.
• They hold grudges.

43
Secrets about Testers

• You dont listen to your testers enough.
• Treated well, testers will work harder than
  anyone else in your company.
• Your testers own your customers.
• QA manager is the most critical job in the
  production team.
• Every producer should spend one project in
  testing. If they are insulted by that, fire them.
• Everyone has a test quota on every project, every
  month. Everyone. Period.
• Testers without other training make horrible
  designers and producers.

44
Secrets about Management

• Things that dont work
• Cheap sweat-shop labor
• Micromanagement
• Overpaying
• Job security threats
• Being their friend

• Things that work
• The Carrot
• Choice
• Responsibility
• Tools
• Pride
• The Stick
• Guilt
• Examples
• Competition

45
Finally.

• When all the toilets are clean

Singapore has arrived.