Postcard From SGS 2005: Computer Gaming to Enhance Computer Science Curriculum

This interesting lecture, presented on the first day of Serious Games Summit 2005, was led by Bryan Barnett, lead program manager, External Research Programs, Microsoft Research, who made an enterprising start to the lecture by insisting on speaking from the floor, instead of the speaker platform, mentioning that speaking from on high "offends the sense of pedagogical informality from my teaching days." He also mentioned that John Nordlinger of Microsoft Research, the specific program manager for this initiative, had been called away to New Orleans as part of the National Guard, so Barnett, who is Nordlinger's boss, was presenting on his behalf.

Re-Invigorating The Field

Barnett started by laying out the fundamental question - can games help to reinvigorate Computer Science and ensure the future health of the discipline and the technology industry? He noted that he considers the lack of people interested in entering the CS field and the drop-out rate in North America to be a massive problem, but firstly took a step back, and explained why a gigantic company such as Microsoft and its research arm actually cares about the issue.

Microsoft Research, Barnett explained, is based around issues on the long-term future of computing, essentially, to "...making sure that Microsoft is around in long-term." There's some transfer of research to products, but the organization, which has 700+ researchers and 50 areas of research, is particularly aimed at solving fundamental problems. It has facilities in Redmond, WA, Mountain View, CA - Bangalore, India, Beijing, China and Cambridge, UK. The Indian and Chinese offices are relatively new, Barnett noted, and he points out that, in an industry that lives on brainpower and innovation, "...we have to go where the talent is." Beijing already has 150 employees and is invested in Chinese universities.

The division also liases with external research in the U.S., Latin America, and India - Barnett's job is to forge and maintain relationships with academic research groups, and focus on areas of research "that are not part of the core computer science research mission." This is why video game-related learning has particularly come up. The organization operates through Request For Proposals to fund academic research groups, and also funds PhD and faculty fellowships, workshops and sponsors conferences.

Barnett also referenced some of Microsoft Research's chief areas, which includes the eScience project, accumulating scientific data in a central location, Sensor Networks that transmit data and are distributed in forest canopies, and digital inclusion, involving technology for under-served communities in third world, in cellphone technologies and other areas. So, in some ways, Barnett suggested, some might see the future of computing education as a little off-topic for organization. But, it was underlined, it's a critical problem, especially in North America, where Microsoft is headquartered - Barnett pulled out his i-mate Jam PDA, and pointed out that the only part that's American made is Microsoft's software. Barnett pointed to the tremendous gains elsewhere in the world, and said: "If you're Microsoft, that's a fairly frightening prospect" - the company needs graduates to function competitively.

Getting Started With Games

So, how do games fit into this? Well, Barnett pointed out to the audience, which included a number of university professors: "We all know of [computer science] students, particularly young men, who get started gaming." In fact, the majority of students have experience of being able to change parameters or other attributes in games. Thus, it's believed that game-related learning may be a way to stave off the precipitous decline in entry to computer science departments - overall enrolments are now down near a level last seen in the 1970s, and the amount of women attracted to the discipline is "less than dismal," according to Barnett. Worse than this, there is also a high attrition level, with 10 to 20 percent of students dropping out each year.

Microsoft and Barnett are particularly focusing on introductory courses, since there are plenty of first-year challenges, and many focus groups feel that initial courses have too much math, too much strange syntax, and impel people too quickly into confusing programming environments. For those who haven't written a program, it feels very abstract, and Barnett even quipped of his company's own product: "Visual Studio for a [first-year] student is simply torture by another name." When you write executable code and the machine does it, it's a very empowered sense.

Thus, the software giant has already been helping to fund projects such as Alice, which is a "3D authoring system" extended by Carnegie Mellon University, and has seen improved grades and a much higher chance of continuing with computer science (88% of students, compared to 47% of students) if it is taken prior to the CS1 introductory programming courses. It tries to introduce students to the idea of programming without forcing them to contend with the difficult challenges of raw syntax, and consists of an authoring environment in which it's easy to create three-dimensional objects and endow them with physical characteristics.

In addition, Barnett referenced the also partially Microsoft-funded M.U.P.P.E.T.S. project at RIT, which is a multi-user virtual world that was described as "EverQuest-like", in which students can run around and see their friends, but which also includes a Java compiler inside it, so that students can code, and then instantiate a new object inside the world to show everyone in the multi-user environment. In addition, C# support is also forthcoming for the innovative introductory program, whose acronym stands for "The Multi-User Programming Pedagogy for Enhancing Traditional Study."

Inside The Terrarium

Finally, Barnett mentioned a game created within Microsoft that's very useful for teaching, named "Terrarium." Although the entire source code is not yet released, the game, which was originally created to showcase some of the technology emerging in the .NET platform by the .NET Tools languages group, has turned out to be a natural vehicle for teaching computing. The game includes a client application, linked in a peer to peer network with other clients, and players must design an insect and place them in a world with plants and other bugs to see who "wins" over time - a survival of the fittest test.

To "design" an insect at the lowest level, players can simply open up the insect class and change the attributes to give the bug different speed and strength values - much like deciding stats for an RPG. But it can get much more complex than this, and it's hoped that this method, without programming a game from scratch, can help students understand object-oriented programming. For example, students operating a client can escalate the complexity and write their own algorithms, and make the behavior of particular bug as complex as they like - effectively programming autonomous AI to outwit their fellow students.

Even further than this, Microsoft has given the server source code to some student teams, who have given the bugs hearing and radar, and even added soil fertility, so that when bugs die, the plants in the game would grow faster in those mortality-rich areas - a neat idea.

Conclusion: A Simple Plan

To conclude, Barnett explained that these above projects were all important, but Microsoft wants to go further, and explicitly explore games as a pedagogical vehicle for CS1 (the introductory university computer science course.) It intends to identify partners, an engine, and create game-based labs for existing courses (based on Microsoft technologies, of course - though that's not the entire point of the exercise!), and evaluate by teaching two versions of the same course to see if students really are interested and retained better by including more game-based learning in traditional Computer Science courses. More information will be available on the Microsoft Research external project site in due course, and finished courses will be available with the MSDNAA site.

In the question and answer session, Barnett agreed that finding the right mix of impressive graphics but easy changeability was tricky - he agreed with a questioner that getting introductory students to program directly into DirectX would not be sensible, but is looking at game frameworks to facilitate learning that is well-crafted for beginners, commenting: "We're not going to have Halo 2 in CS1." But he is definitely concerned that courses might have less impact because the graphics are relatively underwhelming. Microsoft's XNA development framework was also mentioned as a theoretically possible basis for these courses, meaning that students might even be able to develop on PC and/or Xbox 360 - but again, someone would need to code a straightforward training framework for the students to work on. But, whatever the eventual solution, it's heartening for the game industry in general that a company such as Microsoft is recognizing video games' possibly vital role in helping to stimulate the growth of computer science in America and elsewhere.

______________________________________________________