Brian Moriarty, ImaginEngine

Our group focused on case studies of its members who were interested in interactive natural language artistic experiences, generalizing their conclusions for those who may be interested game development community.

Solutions

• Case Study #1: As a professional endeavor
• Case Study #2: As an amateur (Victorian definition: not for those unskilled in the craft but rather experienced people who love it and do it their spare time)

• Prose Adventure
• Parsed Artistic Experience
• Don’t get hung up on the term ‘text’ in ‘text adventure’ – it’s ‘expression in / expression out’
• Natural language is the most expressive form of input and output
• You can afford to do much more in prose than in art
• Parsers will take more and more of a place in future products
• Stand-alone or imbedded in other game structures
• Create PBJ’s in a format entirely controlled by the author
• Not so much the author needs to DO everything
• But the creative effort needs to be entirely directed by the author
• A work of passion, a statement of art
• Adventure games are seeing a rebirth, even more pronounced in Europe

• Lots of existing amateur/community games
• Quick fix experiences, 1-2 hours of gameplay
• Not professional quality
• Some companies trying to move into this space
• Malinche is out there
• TEXTFYRE – becoming a publisher of other people’s content, using a book publishing model
• They are reaching out to the community
• Other casual game competitive product
• Adventure Mysterious Island (3D)
• Cute Night (2D RPG)
• Depth of experience
• About ¼ the size of an Infocom game, about a 5-6 hour experience
• Need to consider the ‘value per hour’ proposition for price points of $20 to $25
• The typical casual game player does not need – or want – 3D interfaces
• Some aversion against “Scumm-based” interface, even if overlayed on top of the natural language parser as an option
• Restrictive to player input

• The PBJ format will support any story style.
• Three things of adventure design
• What the player does to get to their goal
• What the player tried to do while getting to their goal
• What they do to test the limits of the universe

• Needs community forums
• Author participation

• Authors assign ACTIONS on OBJECTS
• Mostly through use of verbs, nouns, and prepositions
• Questions have traditionally been shunned
• Adverbs are difficult to support
• Interaction with NPCs has been awkward in the past
• TADS 3 and LUA were mentioned as authoring workbenches
• TADS 3 is open source
• Natural language processing has advanced significantly in the last 20 years
• Voice recognition as input mechanism
• Can be included, but it’s someone else’s technology/problem
• The tools are all easy, readily available
• PBJ is not a technical risk – the (only) difficult part is the creativity

• Infocom games took 9 months to a year to complete
• Doing a PBJ as a part-time effort would take too long to complete
• Case #1: The business case needs to recognize all of the full-time contributors
• The only full-time resource would be the author
• Case #2: Can’t produce a product in a year in a part-time basis
• Main goal is to connect people of like minds
• Distributed design/authorship, possibly from fan base
• Like the EA distributed model presented by Mike Verdu
• Collaboration through other amateurs
• AmateurGameDev.com – waiting to be kicked off
• Create an ‘invitation only’ community of amateur game developers Approach it like a food co-op
• People of different trade skills work for/with each other
• Personal referrals, rankings, etc.
• Can be free services, barter services, or paid services
• People may want to contribute for credit alone
• The mod community is very passionate, spends a lot of time crafting free stuff (Sims)
• GameFire may collaborate
• Elance.com, RentACoder, but these can be costly & may not specialize in what is needed

• Non-gamer’s expectations are going to be high in natural language processing
• Business case

Case study #1

• Try-before-you-buy
• A small segment of the story is made available for hook-em-and-hold-em play
• This can even be done initially, before kicking off the entire effort, to estimate market interest and a conversion rate

• Needs to utilize Other People’s Money
• Angel fund
• Good for small amounts of money – single product effort
• Venture capital
• These guys only make BIG plays – multi product effort

• Proposed: make this game work for $250k
• Author fee + architecture
• Suspect this is underfunded, doesn’t include technology, testing, marketing, subscription, community maintenance, etc.
• At $20 / copy, break even at 17,000 units
• At $25 / copy, break even at 12,500 units
• Sanity check: Legend games sold 20K units without much effort
• Online try-to-by: the 1% rule
• Of the people who try the game, 1% will purchase
• To make the above targets, that means 1.7M players (more, if the dev costs are suspected to be low)
• On the other hand – targeted marketing to those who are likely to want to buy

• Other sources of funding
• SBIR – retool for blind
• Societies for blind people
• Voice recognition technology companies

• Getting a prototype off the ground
• Doesn’t need to REALLY do all of this stuff
• Just needs to LOOK like it does this stuff
• Fake it with existing tools, build the real pipeline later

Case study #2:

• Boutique approach
• Because it’s not being done for livelihood, rigorous financial models fall away

• Will work with “name brand cache” of ex-Infocom author(s)
• Once architecture is in place, you can get recognized authors(s) to contribute
• Requires some story/puzzle editing (Hitchhiker’s Guide)
• Question: who would buy this who had not played an Infocom game before?

We offer two financial models for Case Study #1, wherein the costs reflect a one-year development effort (Author, technology, web design, ending with publicity), at the end of which the product is launched, at which time certain other costs occur (administration, hosting…).  (The costs are recognized jointly, while they will be spent in different fiscal years.) The first is a self-publishing model, where the author and/or author’s business enterprise undertakes all of the responsibilities of publicity, site administration, and other costs, with the second using one of the existing game portals as a distribution partner. 

Note that most of these figures are common, well-thought-through, industry standard WAGs*.  We invite the reader to put in their own assumptions on cost of authorship, profitability goals, marketing requirements, sales projections, and so forth, and to draw their own conclusions about the viability of the approaches.

* "Wild Ass Guesses" – Sheri Graner Ray

While at first glance the profitability seems higher for the self-publishing model, the determining factor will be in how easily high-end sales goals are made.  For example, a self-publishing model might sell 20,000 copies with extraordinary difficulty, but a portal partner with an enormous reach may sell more copies.  (Portal partners quote a 1% conversion rate, meaning if 4 million people tried the game, sales would reach 40,000 copies.)  In this case, economics would indicate going with a larger portal partner; but it all depends on sell-through in either case.  Also noteworthy is that most portal partners do not sell games that require back-end hosting, but if that were the determining factor, the Server-Side Parser Technology budget could be dedicated instead to a downloadable game engine.

section 6

next section

select a section:
1. Introduction  2. Speakers  3. Executive Summary  
4. FabulaRasa: Story in Games
5. Games that Encompass a Diverse and Unusual Range of Emotions
6. Interactive Natural Language Artistic Experiences
7. Professionalizing Game Design
8. Schedule & Sponsors

HOME   SITE MAP    ABOUT   REGISTRATION INFO    SPONSORSHIP INFO   REPORTS   PRESS INFO   CONTACT INFO

Copyright 2000-2008, Fat Labs, Inc., ALL RIGHTS RESERVED
Last update 2/6/2008