Make Something That Makes Something:
A Report On The First Procedural Generation Jam
Michael Cook
Computational Creativity Group
Goldsmiths, University of London
ccg.doc.gold.ac.uk
Abstract
We report on the first procedural generation jam,
PROCJAM, an event designed to bring together artists,
researchers and game developers to experiment with
new techniques and applications for generating content
for videogames. Much of the event’s resulting work has
applications beyond videogames, however, and we believe
the event may be a strong platform for engaging
creators and programmers in Computational Creativity
in the future. We discuss the structure of the event, the
results it yielded, and the potential future impact of such
events on the Computational Creativity community.
Introduction
Procedural content generation (PCG) is a crucial and rapidly
developing area of videogames technology (Togelius et al.
2011). PCG is a rich area of games culture – it has been
used as a supplement for human effort (Interactive Data Visualization
2002), a source of wonder and unpredictability
(Toy et al. 1980), a tool for artistic expression (Betts 2014),
and a unique mechanical design tool (Yu 2009). Its increase
in popularity and its growing importance in the culture of
videogames has also been mirrored by a surge in the aesthetic
of generative software in art, web culture (such as
Twitterbots) and other creative media. These are all areas
which have considerable overlap with Computational Creativity
in terms of the techniques they use to generate artefacts,
and represent a great opportunity to share the field’s
philosophy and theory with a vibrant, active community of
people.
Game jams are increasingly common events within the
game development community where people develop games
under the constraints of both a time limit and some kind
of common theme (which might be a technical constraint
such as containing the game within 4 kilobytes of Java1 or
a creative constraint such as incorporating a theme like fishing2).
Entrants to game jams typically fall within one of two
categories: novices looking to use the event to create their
first game, and experienced developers looking to experiment
and innovate (Zook and Riedl 2013). In both cases the
1
http://www.java4k.com
2
http://www.fishingjam.com
short timescale helps encourage entrants to set themselves
projects which are small enough to be easily completed.
Popular game jam formats are repeated at regular intervals
throughout the year. Ludum Dare,3 one of the most popular,
runs every four months. Entrants make a game in 48 hours
from scratch, including game design, code, music and visual
art, following a theme voted on by entrants in the week
prior to the game jam. In December 2014, Ludum Dare 31
received over 2000 entries for the theme Entire Game On
One Screen. By running repeatedly, these regular game jams
build communities of creators who meet to create together,
share ideas, give feedback on games (there is an extensive
period of reviews and ratings after the jam) and often form
collaborations or extend their jam entries into full commercial
releases (Zucconi 2014). They form strong global communities
who share ideas, draw in new practitioners, and
push forward the state of the art (Gray et al. 2005).
In this paper we present a report on the first procedural
generation jam, or PROCJAM, an event held in November
2014. The jam ran over nine days, starting with a streamed
day of talks about procedural generation and ending with
138 entries being submitted in the form of games, tools, experimental
prototypes and artworks. Although styled as a
game jam, PROCJAM deviated from the traditional format
in several important ways, which helped expand the appeal
of the event beyond game developers and draw in people
interested in generative techniques in general. We will go
into these changes to the format in depth later in this paper,
as we believe they are crucial to the success of PROCJAM
and point to a format for generative events that could form
the basis of Computational Creativity outreach in the future.
We also will outline how PROCJAM itself is fostering work
related to Computational Creativity and how this can grow
in coming years.
We believe that the community-building and experimental
aspects of game jams are extremely valuable, and make
the game jam format ideal for engaging communities of
programmers such as Twitterbot writers, programmer-artists
and game developers with Computational Creativity, as well
as being rich sources of inspiration and code which would be
of benefit to everyone working in and around this field. Additionally,
events like PROCJAM can also be valuable ways
3
http://www.ludumdare.com/compo
Proceedings of the Sixth International Conference on Computational Creativity June 2015 197
to expose people to Computational Creativity for the first
time, in much the same way that game jams encourage people
to try out making a game, and may serve as a useful
model for student workshops and similar activities.
In this paper, we will outline the format for PROCJAM
and explain how and why we deviated from several common
elements of typical game jam organisation to create a
better community for creating and sharing ideas. We then
give several specific examples of entries to the jam and discuss
their relevance to Computational Creativity. We follow
this with a general analysis of entries, identifying issues related
to Computational Creativity that arose in them, and
also areas where our research could contribute to future entries
to the event. Finally, we discuss the jam format as a
model for outreach and engagement, and look ahead to the
future of PROCJAM.
PROCJAM Format and Organisation
PROCJAM took place from November 8th to the 17th 2014,
co-ordinated across the web using Twitter hashtags and a
central website where people could submit their entries.4
Subsequently, the jam has registered its own website to coordinate
the community and future events.5
The most common format for a game jam is as follows: at
the beginning of the jam a theme is announced, normally on
the jam’s website so that people can take part from around
the world. Participants then have 48 hours to develop a game
from scratch, including art and music assets, that somehow
incorporates the jam’s theme. Entries are then submitted at
the end of 48 hours. A review phase then takes place in
which people vote for their favourite entries, with the voting
pool consisting either of the general public, other entrants to
the jam, or a select panel of judges. Prizes may be awarded
to the winners.
This format for a game jam is very popular and is replicated
hundreds of times a year from large-scale jams with
hundreds of entrants down to small-scale local jams run between
small groups of friends. With PROCJAM we made
several changes to the standard game jam format with the express
aim of increasing participation, particularly with those
who had relevant experience writing generative software but
had not interacted with game developers before. A secondary
aim for the jam’s format was to encourage experimentation
and allow people to prototype unusual projects
that stretched the state of the art in generative techniques for
games.
Unlike most other game jams, making a game was not the
only way to enter PROCJAM. Entrants could alternatively
submit a piece of software that simply generated something
(the jam’s slogan was Make Something That Makes Something).
Developing a game is a highly specific skill that people
are unlikely to have unless they already work in games,
and developing a game in the timeframe of a game jam is
even more difficult. By relaxing this constraint, people who
have interesting ideas, knowledge or skills can contribute
generative systems to the jam that might spur on projects or
4
http://itch.io/jam/procjam
5
http://www.procjam.com
inspire developers to integrate new kinds of system in their
future games. As a result, the jam received many entries in
the form of complete games, but equally saw systems which
generated dungeons and planets, weapon ideas and fabric
designs, music loops and more. Bringing in people from different
backgrounds helped make PROCJAM feel more like
a melting pot of ideas and less like a competition.
We removed the requirement to produce original artwork
and music for the jam, too. Since the primary focus of the
jam was on new ideas in procedural generation, rather than
testing game development skills, it didn’t make sense to require
people to put effort into elements of a game that were
unrelated to their main contribution. This encourages people
to enter the jam by relieving pressures on them to take
on more work. We also removed a similar requirement that
all code should be written from scratch. Game jam games
tend to be very simplistic in nature because of their short development
cycles, which works well for the goals the jams
often have. However, in order to allow people to spend the
week focusing on procedural generation, it made sense to
allow them to use existing codebases or even entire games.
One group of developers took a game they had been working
on and added a procedural generation system to it as part
of the jam. This doesn’t just make the jam more appealing
to outsiders, it can also allow deeper work to be done that
builds on existing efforts (Hecker 2012).
By allowing entrants to make anything from a small script
to a full game, and removing the restrictions on existing code
and art assets, the process of evaluation becomes an issue.
This raises the question of how to rate and compare entries
when they are so varied in their origins. PROCJAM circumvents
this simply by removing the ratings process - people
are encouraged to comment on each others’ entries and share
them among one another, but there is no numerical rating
system and no winners are declared. This solves the issue of
comparing, say, a script which generates quilt patterns with
a full murder mystery game. However it simultaneously encourages
people to try out more experimental ideas without
the intimidation of being judged and ranked by someone
else’s idea of what a good jam entry should be.
All of these changes have the same ultimate aims: to encourage
people to take part, particularly those who are not
game developers by trade, and to encourage experimentation
and the sharing of new ideas.
We supplemented PROCJAM with a day of talks which
we livestreamed on the web on the first day of the jam6. 80
people turned up to attend the day of talks, with 200 viewers
tuning in to each talk throughout the day, and many hundreds
more have viewed the recordings of the talks online
since. The talks provided inspiration to jam entrants, with
many citing the talks during the development of their jam
entry, but they also provided an opportunity to be exposed
to new views on generative systems – the speakers included
an academic researcher, an artist and a creative director at
an indie games studio. One of the aims of the event was
to elevate procedural generation in games beyond “random
levels”, and having a variety of speakers giving talks was a
6
http://www.procjam.com/talks/2014
Proceedings of the Sixth International Conference on Computational Creativity June 2015 198
Figure 1: A screenshot from Dreamer of Electric Sheep.
good way of doing this. We hope to have a speaker at next
year’s PROCJAM event to promote Computational Creativity
as a new philosophy for procedural generation in games.
Selected Entries
In this section we will briefly describe and discuss three entries
to the jam. We look at their most interesting features
and the responses some of them received from the games
community. We selected three projects that we believed
would be of most interest to the Computational Creativity
community, either because of their philosophy, innovative
concepts, or the relationship between the techniques used
and ideas within the field of Computational Creativity. We
give details of where to find these entries, as well as all other
jam entries, in the next section.
Dreamer of Electric Sheep
Dreamer of Electric Sheep is a text adventure submitted to
PROCJAM by Tom Coxon, who also gave a talk at the jam’s
opening event about his procedurally-generated adventure
game, Lenna’s Inception (Bytten Games 2014). Like most
text adventures, players are presented with descriptions of
their surroundings and can manipulate the world by inputting
simple commands for their character to execute.
Dreamer attempts to procedurally generate the game world
using a combination of ConceptNet (Liu and Singh 2004),
a commonsense knowledge database, and Spritely7, a tool
which generates game art from web searches. ConceptNet
stores its data in the form of concepts, series of facts that
are all about a similar topic. These facts are linked to one
another through triplets (such as {magazine, AtLocation,
bookstore}) which can be explored through an API which
Dreamer uses. ConceptNet has seen use in academic Computational
Creativity research, such as (Llano et al. 2014).
Dreamer searches the ConceptNet API for concepts which
other concepts are linked to via the relationship AtLocation.
It then populates those places with objects and characters
that ConceptNet says should be found in them, and
7
http://www.github.com/gamesbyangelina/spritely
Figure 2: A screenshot from Inquisitor showing the conclusion
to a case. The player has correctly guessed the motive
but not the murderer or the weapon.
uses Spritely to generate an illustration of the location.
Spritely queries online image databases such as Google Images
and Wikimedia Commons, searching for images that
can be cleanly shrunk down in size with their backgrounds
extracted, to make relatively clear sprites for use in games.
The player can perform common text adventure commands
such as moving in the cardinal compass directions to
travel between places, as well as picking up objects. However,
because the game lacks deeper knowledge about the
objects it places in each location, this can result in surreal
interactions like picking up shop assistants and taking them
with you. The game gets around this somewhat by being
set inside a dream world, thereby allowing unusual things to
take place without the game’s sense of reality breaking.
The Inquisitor
The Inquisitor is a murder mystery game by Malcolm
Brown. The game tasks the player with solving a murder
by investigating the crime scene, discovering evidence,
questioning witnesses and identifying the murderer, murder
weapon and motive. The crime is procedurally generated,
generating a cast of characters and relationships between
them, simulating the movement of the characters before and
after the murder (so that evidence such as blood trails and
witnesses are realistic and consistent) and then leaving the
player to put together the details within a time limit.
Although the individual generative techniques within The
Inquisitor are not new per se, the way it uses them to generate
murder mysteries is novel and quite effective. In particular,
interviewing witnesses yields partial and sometimes
conflicting information, forcing the player to take notes and
draw up potential scenarios in which certain characters are
lying, and information is procedurally redacted from certain
kinds of evidence, leaving out the contents of a letter but
revealing its author, for example, or smudging the name of
its author but revealing unrequited love. The Inquisitor also
Proceedings of the Sixth International Conference on Computational Creativity June 2015 199
Figure 3: A screenshot from Secret Habitat showing a generated
gallery in the generative game landscape. The player
can walk inside and view the pieces, as well as exploring
outside to find other galleries.
adds little additional touches on top of the game, such as a
procedural system for applying accents to the pre-written dialogue.
This takes dialogue written in plain English and then
adds affectations to it to simulate a character who is drunk or
has a particular speech impediment. To our knowledge this
is a completely novel idea for content generation in games.
Secret Habitat
Secret Habitat is an ‘art gallery simulator’ and ambient exploration
game by Strangethink. The game has no obvious
end state. Instead, the player is encouraged to enjoy walking
around the game’s generated world, entering the various
buildings, viewing artworks and listening to audio recordings,
all of which are also procedurally generated. The
game seems to appeal to a particular aesthetic of wonder and
mystery associated with generative software. One journalist
wrote about the game:
[The paintings] seem to use a similar algorithm, or
similar parts, or similar something, as colours, patterns,
and other motifs repeat across them; you can
recognise they’re part of a series. Seeing different spins
on common themes can be delightful, and it’s awfully
exciting when you discover one painting very different
to the rest of its set. (O’Connor 2014)
Strangethink’s biography simply reads I make strange
computer worlds8, and Secret Habitat leans towards digital,
interactive art as much as it does the traditional ideas
of videogames as systems of rules and objectives. Procedural
generation has much overlap with both game development
and interactive art, and generative software has a
unique value in being able to present extremely large or in-
finite scales to a user (Betts 2014).
Analysis of Entries
PROCJAM received 138 entries in total, although more entries
may exist that were not officially submitted, since we
are aware that the jam was set as a class assignment in at
8
https://twitter.com/strangethink23
least two universities and not all students submitted their
entries to the site. All of the jam entries are available online9.
The entries include full games, prototypes, tech demos,
tools and libraries designed for developer use, as well
as standalone generators and art pieces. We encourage the
reader to visit the site and browse the entries themselves.
By way of a brief survey, we categorised the entries to
the jam into two categories: game, or tool. The categories
are defined loosely as follows: a game is any software designed
to be interactive, but not for the purposes of producing
something; a tool is any software designed to facilitate
the generation of content as part of a larger creative activity.
These definitions are not strict, but we offer them here as a
rough partition of the entrants to the jam. Overall there were
79 game submissions and 59 tool submissions.
The games typically involved some kind of generative element
in how they set up their game world, such as generating
the 3D galleries in Secret Habitat or using simulation to create
murder mystery scenarios in The Inquisitor. Most tools
fell into one of two categories: some generated common
kinds of content in accessible ways, such as world map generators
of which multiple were submitted to the jam. This is
partly because procedural generation lacks a cohesive community
and established baseline software that solves common
problems such as world generation - instead, developers
tend to reinvent solutions to common problems repeatedly.
We believe this is a key problem PROCJAM can target
to benefit the generative software community in coming
years. Other tools generated unusual kinds of content which
are not commonly seen in games, like GlyphGenerator’s alphabets
or Bootleg’s 3D shoe models. These are exciting
because they break new ground in generative techniques and
offer new applications for games, similar to those submitted
to the jam.
Computational Creativity Issues
PROCJAM’s aim was primarily to produce generative tools
and games, and to bring together both novices and experts
to try out new ideas and learn more about the field. The
theory and practice of Computational Creativity is gaining
awareness in generative communities, but we believe that
many developers are not confident about how to concretely
use these ideas in the software they are building. That said,
many entries to PROCJAM touch upon issues in the field,
and others show clear areas where they could be extended
to take advantage of results from Computational Creativity
research.
Many of the tools explore co-creation, in which the software
either creates alongside the user or tries to assist the
user in achieving a particular goal, such as (Liapis, Yannakakis,
and Togelius 2012). Synthetic Poetry allows the
user to write poetry on alternating lines, along with one
of three poet models based on Keats, Shakespeare and
YouTube Comments. Other entries were more straightforward
tools, such as Nodemancer, which allows users to specify
the components of an item, such as a sword, and then lets
the system design the specific details autonomously. Most
9
http://itch.io/jam/procjam#entries
Proceedings of the Sixth International Conference on Computational Creativity June 2015 200
tools focused on the user retaining control, however: SPARTAN
proudly announces that ‘the user has complete control
over every step of the generation’ – encouraging people to
explore ideas that break ideas like complete user control is
something that will need to be emphasised in future years of
the jam. We hope to expand the jam’s resource pool to include
tutorials on basic Computational Creativity techniques
and perhaps an invited talk from a Computational Creativity
practitioner in a future event.
Issues relating to framing and context, as in (Charnley,
Pease, and Colton 2012), arise in several entries including
games like The Inquisitor which generate text as part
of gameplay (for example, to provide dialogue and scenesetting
for the murder mystery). This text is partly game
content but also acts as contextual information that justifies
decisions made by the generative system in producing other
content. We have argued in the past that framing for game
content generation is a broader concept than simply being
‘wall text’, and can extend to text that appears in-game to
help the player understand and contextualise generated scenarios
and systems (Cook 2014). Many of these games are
beginning to explore these ideas, and we hope to see this
trend continue in the future of the event.
Also related to framing, several entries play with the problem
of communicating the logic, internal representation or
behaviour of the generative system. Both Diversitizer and
Meadows present the player with a natural environment populated
with various flora. The locations of each plant, as well
as its properties, are governed by procedural systems and
vary each time the game is started. The player can gain an
understanding of these parameters and the expressive range
of the generator by observing the environment and repeatedly
generating new worlds, even though the software does
not communicate any information to the player through text.
In this way, discovering the decisions made by the software
become part of the purpose of interacting with the artefact,
which is an interesting kind of implicit framing that is not
often discussed in Computational Creativity discourse.
Many entries to the jam have clear ways in which they
could be extended using techniques from Computational
Creativity if the developer wished. Identifying simple ways
in which common game ideas can be extended is important
both in planning ‘code camp’ events for Computational Creativity,
and for giving compelling examples at events like
PROCJAM to show developers steps they can take to begin
exploring the field. Many generative systems use parameters
selected by the developer, such as Infinity Explorer
which generates 3D worlds for the player to fly around in an
airship. Encouraging developers to build their systems such
that they can select parameters either based on external, contextual
factors as in (Colton, Goodwin, and Veale 2012) or
by evaluating its own output as in (Smith and Mateas 2011)
is a good way to begin to move some of these generative
systems in new directions.
The idea of the software evaluating its own work seems
to be one of the most accessible ideas from Computational
Creativity that generative software developers can start experimenting
with. Generative systems tend to be developed
in such a way that they are guaranteed never to produce bad
output: in other words, they rely on reorganising hand-made
elements that the developer knows in advance will produce
reliable content. This is an effective method for game development
as it ensures the player will not be disappointed,
however the culture of experimentation that we tried to encourage
makes PROCJAM an ideal place for people to try
out ideas that are less robust but perhaps more interesting
and experimental.
Discussion
PROCJAM has relevance to the Computational Creativity
community because it represents the founding of an interdisciplinary
community of generative programmers who we
hope will, over time, be introduced to and begin experimenting
with ideas and approaches from Computational Creativity
too. The results of the jam and the details about its organisation
are important in their own right; that said, we
believe that PROCJAM also holds interesting potential for
future events that could strengthen and broaden the appeal
and reach of our field.
Computational Creativity is a relatively young academic
field that is still laying some of its foundations (Colton and
Wiggins 2012). At the same time, many of the aims it has
and the technologies and techniques it uses are highly relevant
to movements in digital art, videogames and web culture
as it stands today. In much the same way that outreach
events must target academics in related fields, we should
also look to engage these non-academic communities, to
share solutions, and to encourage the adoption of our ideas.
We often use the term ‘mere generation’ as a way of describing
purely generative software, but we must also bear in
mind that a lot of exciting and interesting work is being done
in generative software communities, and we should seek to
engage with these communities, learn from them, and try to
convince them that ideas from Computational Creativity are
exciting and interesting, too.
The format of an event like PROCJAM, particularly with
some of the changes we made that we discussed earlier,
make it ideal for informally bringing together several communities
at once, making new connections and allowing
them to demonstrate their working practices and techniques
to one another. It also serves as a small-scale and selfcontained
event to set for students who may be interested
in the field; PROCJAM was a credit assignment for one
university class in particular, and we understand that feedback
from the students was extremely positive. With encouragement
and additional resources about Computational
Creativity, future versions of PROCJAM (or perhaps a separate
Computational Creativity jam) could serve as informal,
global workshops that introduce people to the area in a practical
way.
Despite their short length and highly applied nature, jams
can serve a similar purpose for researchers as they do for
programmers. There already exist examples of published research
work which started off as a jam submission, in which
an idea was quickly prototyped and then later developed after
the jam (Cook and Colton 2014). PROCJAM also played
host to jam games which were implemented to demonstrate
an existing research tool or technique in a more concrete
Proceedings of the Sixth International Conference on Computational Creativity June 2015 201
way (Cerny 2014). We hope that subsequent PROCJAMs
will see more researchers from this community take part to
produce games or tools that demonstrate their work to game
developers.
PROCJAM also leaves a legacy of code and ideas that persists
after the jam has ended and gives the event lasting value
in the months when it is not running. PROCJAM’s 138 submissions
offer ideas and inspiration, and in some cases code
samples and open source projects. Entrants to the jam have
already collaborated with one another, expressed intentions
to develop their entries into full games, and in one case an
academic issued an open call to the PROCJAM community
for PhD applications, which was taken up by one entrant.
PROCJAM had multiple features written about it in industry
magazines and several others on major websites1011, and
many of the games created for the jam were individually featured
and written about as well. The jam’s slogan, Make
Something That Makes Something, has reappeared in other
events relating to procedural content generation too12. All of
this shows that the jam is more than just the week in which
it is held – it has a larger impact by creating a community of
people that we hope will thrive.
To build upon this, we are planning for PROCJAM 2015
to have more resources ready online before the jam begins,
aiming to encourage newcomers to writing generative
games, tools and software. Through talking to entrants,
we’ve identified several ways in which we can make the
event easier to enter for people. We are hiring an artist to
produce some public domain art assets for people to use,
specifically designed to be easily recombined and mashed
up in procedural systems. We’re also talking to developers
and researchers with the aim of producing some short tutorials
that demonstrate simple generative techniques. These
resources will persist beyond the jam itself, hopefully making
it easier for people to begin learning about generative
techniques at any time of year. We intend to include ideas
from Computational Creativity among these resources, in
the hope that it will encourage people to think of these ideas
as being as essential as any algorithm for making things that
make things.
Next year we hope to run some analysis on the entrants
to the jam, primarily through optional surveys. This will
help us get an idea of the jam’s makeup, and people’s motivations
for entering the event. We are concerned that the
lack of evaluation will make the jam complicated for people
to curate and explore afterwards, and also acknowledge
that some people will be interested in being rated by their
peers. We are still reviewing our decision to remove rating
altogether from the jam - we may make alterations in 2015
to improve filtering and curation, although it is still unlikely
we will implement global ratings that declare overall winners,
as we felt the lack of rating contributed a lot to the
jam’s informal atmosphere.
10http://tinyurl.com/procjampcgamer
11http://tinyurl.com/procjameurogamer
12http://tinyurl.com/aigameslecture
Conclusions
In this paper we reported on the first procedural generation
jam, or PROCJAM. The event was designed to create
a new community around generative techniques for games
and other software, with an emphasis experimentation, sharing
of ideas and introducing new people to writing generative
code. We described the changes we made to the classical
game jam format to encourage more participants and make
the event more accessible. We believe we were successful
in this regard, but we also know that there is a lot of work
left to be done in maximising the event’s impact and accessibility,
which we hope to address in future years. We then
showed some illustrative examples from the 138 entries received,
and discussed the potential for jams to impact communities
close to Computational Creativity and potentially
nurture relationships and collaborations between them.
Acknowledgements
Thanks to the reviewers which helped improve the paper,
particularly the introductory segment on procedural generation.
We would like to thank Azalea Raad for her help organising
the event as well as the speakers, press and entrants
who supported the jam. We wish to thank PROSECCO for
kindly sponsoring the day of talks that launched the jam.
Thanks also to Sophie Houlden, whose Fishing Jam inspired
several of the changes to PROCJAM’s format. This work
was sponsored by EPSRC grant EP/L00206X.
<references_biblio/>
References
Betts, T. 2014. An Investigation of the Digital Sublime in
Video Game Production. Ph.D. Dissertation, University of
Huddersfield.
Bytten Games. 2014. Lenna’s inception.
http://lennasinception.com/.
Cerny, M. 2014. Machine understanding for interactive storytelling.
http://martin-cerny-ai.itch.io/amuse.
Charnley, J.; Pease, A.; and Colton, S. 2012. On the notion
of framing in computational creativity. In Proceedings of the
Third International Conference on Computational Creativity,
77–81.
Colton, S., and Wiggins, G. A. 2012. Computational creativity:
The final frontier? In ECAI 2012 - 20th European
Conference on Artificial Intelligence., 21–26.
Colton, S.; Goodwin, J.; and Veale, T. 2012. Full-face poetry
generation. In Proceedings of the Third International
Conference on Computational Creativity, 95–102.
Cook, M., and Colton, S. 2014. A rogue dream: Automatically
generating meaningful content for games. In Proceedings
of the First Experimental AI in Games Workshop,
AIIDE 2014.
Cook, M. 2014. If only we could talk to the generators.
Talk at the Procedural Content Generation Summit at ITU
Copenhagen.
Gray, K.; Gabler, K.; Shodhan, S.; and Kucic, M.
2005. How to prototype a game in under 7 days.
http://tinyurl.com/7dayprototype.
Proceedings of the Sixth International Conference on Computational Creativity June 2015 202
Hecker, C. 2012. The depth jam.
http://chrishecker.com/The Depth Jam.
Interactive Data Visualization, I. 2002. Speedtree.
http://www.speedtree.com/.
Liapis, A.; Yannakakis, G.; and Togelius, J. 2012. Cocreating
game content using an adaptive model of user taste.
In Proceedings of the Third International Conference on
Computational Creativity.
Liu, H., and Singh, P. 2004. Conceptnet: A practical
commonsense reasoning toolkit. BT Technology Journal
22:211–226.
Llano, M. T.; Hepworth, R.; Colton, S.; Charnley, J.; and
Gow, J. 2014. Automating fictional ideation using conceptnet.
Proceedings of the AISB14 Symposium on Computational
Creativity.
O’Connor, A. 2014. Time to admire a secret habitat.
http://www.rockpapershotgun.com/2014/12/17/secrethabitat-art-gallery-simulator.
Smith, A. M., and Mateas, M. 2011. Knowledge-level creativity
in game design. In Ventura, D.; Gervas, P.; Harrell, ´
D. F.; Maher, M. L.; Pease, A.; and Wiggins, G., eds., Proceedings
of the Second International Conference on Computational
Creativity, 16–21.
Togelius, J.; Yannakakis, G. N.; Stanley, K. O.; and Browne,
C. 2011. Search-based procedural content generation: A
taxonomy and survey. IEEE Trans. Comput. Intellig. and AI
in Games.
Toy, M.; Wichman, G.; Arnold, K.; and Lane, J. 1980.
Rogue.
Yu, D. 2009. Spelunky.
Zook, A., and Riedl, M. O. 2013. Game conceptualization
and development processes in the global game jam. Proceedings
of the Foundations of Digital Games (FDG’13).
Zucconi, A. 2014. 0rbitalis. www.orbitalis.org.
Proceedings of the Sixth International Conference on Computational Creativity June 2015 203