In the run-up to the 2006 Independent Games Festival, which is held at Game Developers Conference 2006 in San Jose from March 20-24, 2006, Gamasutra is showcasing a number of the IGF finalists in different categories. As part of a series of Gamasutra Education-exclusive articles, we profile the 2006 IGF Student Showcase winners by interviewing them about their award-winning titles, which will be playable at the IGF Pavilion at GDC this March.
Today Gamasutra interviews Michael Chrien, lead designer of NERO, which was developed at the Digital Media Collaboratory/Innovation Creativity and Capital Institute at the University of Texas at Austin. An excerpt from the game's official website follows:
"Although it resembles some RTS games, unlike most RTS games NERO consists of two distinct phases of play. In the first phase individual players deploy robots in a ’sandbox’ and train them to the desired tactical doctrine. Once a collection of robots has been trained, a second phase of play allows players to pit their robots in a battle against robots trained by some other player, to see how well their training regimens prepared their robots for battle. The training phase is the most innovative aspect of game play in NERO, and is also the most interesting from the perspective of AI research."
Gamasutra: What's the concept behind your IGF Student Showcase winning game, and give us an outline of the team that's behind it?
The concept of NERO is that the player is able to teach an army of robots different strategies by setting up various training scenarios and rewarding different behaviors. Initially the player’s team of robots has no intelligence. They’ll wander around and shoot randomly. The player is able to teach his or her team of robots different strategies by altering the robots’ environment and rewarding or punishing certain behavior traits (such approaching the enemy, or grouping together). Once the player is satisfied with the team, they are able to battle their army against opponents over the internet or LAN.
The team is currently composed of 5 programmers who are undergrads at the University of Texas at Austin, a graduate student who oversees research aspects of the project, and a producer who manages meetings. Over the course of the project, we’ve had about 30 students from UT as well as the Austin Community College system, who have worked on art assets, design aspects, scripting, and programming.
The project would not have been possible without Kenneth Stanley’s
graduate work on the real time version of “NeuroEvolution of Augmenting
Topologies” (rtNEAT for short). Ken now serves as an advisor to the
project. The current lead programmer is John Sheblak, and our graduate
student advisor is Igor Karpov. Our current programmers are Gary Tsang
and Dustin Stewart-Silverman, while our current art team consists of
Julianna Budding. Our producer is Aliza Gold. My name is Michael Chrien and I’m currently the lead designer (former lead programmer). Other notable project members include Philip Flesher, Ryan Cornelius, Bobby Bryant, Jonathan Perry, Aaron Thibault, and our faculty advisor Risto Miikkulainen.
|A snapshot of NERO|
GS: Tell us a little bit about the school and school program which were behind the game's genesis? Was this part of a course or final project? What kind of degree program did it count towards?
The game was not created as a part of a course, nor did it count towards anyone’s degree program. The game was created in order to research and promote the area of using neural networks as artificial intelligence in video games.
The initial game concept was proposed by Kenneth Stanley during a
breakout session of the “Artificial Intelligence in Games” conference
held at the Digital Media Collaboratory (DMC). The DMC lab is part of
the IC2 Institute at the University of Texas, Austin, and is responsible
for discovering, funding, and managing unique research projects for
commercial applications. The lab saw value in Ken’s idea, and decided to fund and organize the project. Reaching into Austin’s vast student population, Aliza Gold, the game’s producer, recruited student volunteers who were interested in art or programming to work on various aspects of the game.
The lab provided students with a place to work and the hardware and
software they needed to create the game, as well as the project
structure to make such a project work. Weekly meetings were held for
each team, art and programming, as well as a leads meeting, in which the
leads for each team would meet with the producer to discuss the broad
project plan for the game. From the beginning, the project was to be
run like a full time games studio, and structured itself accordingly. Students worked on the project as volunteers; however some of the students were able to get research credit through our faculty advisor in exchange for their participation on the game. Also, students had the ability to be promoted to a leads position and would have more say in
the large scale design and planning of the game. To date, NERO is the first large scale academic project to successfully model itself off of the game studio model using a team of student volunteers. It is also unique in the way it blended academic research
and commercial game creation into one project.
GS: How long did development on the game take and what tools did you use to create it?
The game has been in development for about 2 years now. However, this
is an on going project, and the development team is currently working on
a new version of the game. The game uses Garage Games’ Torque Engine as
the basis for the game. It was compiled under Microsoft Visual Studio,
and WinCVS was used for version control. Art assets were created using
3D Studio Max and Adobe Photoshop, while levels were designed using
Torque’s mission editor.
GS: What was the all-time best and all-time worst moment that you encountered during the game's creation?
The all-time best and all-time worst moment for me had to be when we
first released the game to the public. It was stressful for me as lead
programmer getting the game into a releasable state, wrangling up my
programming team, and arguing with the leads over last minute release
issues. Eventually we finally packaged it up and hosted it on the
Computer Science department’s servers at the university. To get people
to download the game, the team began spreading the word by contacting
newspapers and internet websites. We managed to get a mention on a
little known tech news site, Slashdot. Within about 12 hours of the
story’s posting to Slashdot, the number of downloads was rising past
25,000. Unfortunately, ironically, and somewhat humorously, the CS
department’s servers couldn’t handle the strain, and they had to
shut down the subdomain we were using to host the file. Evidently the
traffic was so high, that it not only knocked the CS department’s
webpage offline, but it started to affect the entire university network.
The best moment was when I went to the main University of Texas
website, and it loaded a lot slower than usual.
GS: Do you (yet) have any success stories or positive experience based on showing the student game to people in the game industry (praise, actually getting a job in the biz, etc)?
NERO was actually featured at the “Experimental Gameplay Workshop” at last year’s GDC, and we got some positive feedback from people who saw the presentation. Industry people sometimes swing by the lab every now and then and are usually awestruck by the game and AI in the game. Richard Garriott was especially interested in the AI during early development of the game. Philip Flesher, a former lead programmer, now works at NCSoft in Austin, and former artist Blake Lowery went on to co-found his own game studio, Cellien Studios. As for the current team, there’s no doubt in our mind NERO will be an excellent resume builder.
Also, to date, NERO has fostered six research papers, two of which have been recognized at various academic conferences including a best paper award at the IEEE 2005 “Symposium on Computational Intelligence and Games” and a mention at the 2005 “Artificial Intelligence and Interactive Digital Entertainment Conference”.
|Another snapshot of NERO|
GS: What are the most important things that student games should be showing off, in terms of both getting high marks in your courses and impressing potential employers?
I think that knowing all about the basic elements which constitute a
modern video game, such as the game loop, user input processing, victory
condition design, graphics rendering, etc, and how all of these ideas
and properties work together to create a final game should get high
marks in both courses and impressing potential employers. If the game
demonstrates beautiful graphics, but it has poor game design and has
loose controls, is it really a game? A game should show off the entire game, not just one aspect of a game. I think any game which grasps this idea will get high marks in coursework and will impress potential employers.
GS: Have you tried any of the other Student Showcase finalists? If so, which ones did you especially appreciate, and why?
I played Ballistic and loved the simplicity of the controls and the simple, yet addictive game design. I loved the hand drawn artistic style and puzzle gameplay of Orblitz. I think my favorite game was Cloud. I loved the art design, the game design, and the music. Cloud is one of those rare games that appeals to all audiences whether they are actually playing the game, or just watching it.
GS: Name one thing that people probably don't know about your game.
It has more world wide appeal than I thought it would. There have been news articles written about it in Russian, Spanish, Hebrew, and other foreign languages. We have gotten emails with translated versions of the documentation. Also, Ken recently came across some lecture slides from a University in Israel that are basically about NERO. The impact that the game has made not only with English speakers, but also with those people cannot even read or understand any text in the game is really pretty mind blowing.
GS: Have you any other messages for your fellow Student Showcase winners?
Congratulations to everyone. I know that a lot of hard work was put into the games that won, so enjoy your time in the spotlight. My team and I are looking forward to meeting you guys at this year’s Game Developer’s Conference.