Bringing Simulations and Games to Scale
This chapter considers the potential to scale up the use of simulations and games for science learning. The first section provides an overview of current market penetration of games in formal and informal learning contexts and identifies barriers to increased distribution and sales. The second section discusses alternative future pathways to scale. Although the chapter primarily focuses on games, the scaling issues are relevant to simulations as well. The chapter ends with conclusions.
BARRIERS TO SCALE
Increasing the uptake of games for science learning is a complex problem affected by a variety of barriers to use in both the formal context of the science classroom and the informal context of the home, science museum, or after-school club. Some barriers, such as the lack of viable business models and inadequate attention to consumer testing, limit development and sales of games in both formal and informal learning contexts. At the same time, there are barriers to marketing educational games that are unique to formal education. Educational markets for games are fundamentally different from broader public markets. It is important to keep in mind that blockbuster sales of commercial games establish a bar that has never been achieved by any educational software product. For example, World of Warcraft—Wrath of the Lich King sold 2.8 million copies within 24 hours of its November 2008 release.
The Lack of Proven Business Models
Mayo (2009b) argues that the primary barrier to wider use of science games is the lack of a successful business model.
One business model, in which academic developers aim to commercialize a game, generally fails for one of two reasons, in Mayo’s (2009b) view. The first reason is that grants provided for game development generally do not include funding for commercial “hardening” (enhancing security, consumer testing, refining), marketing, and distribution. Second, even if the funders do support these activities, most academic developers lack the skills and knowledge, personnel, and financial resources to harden and market the game. In addition, academic reward systems typically do not encourage faculty members to commercialize educational games.
A representative of the commercial game industry (Gershenfeld, 2009) agreed with Mayo that most academic game developers lack the expertise needed to commercialize games. He argued that educational games have not sold well because academic developers have not designed them from the beginning to successfully meet market demand, as commercial publishers do. Publishers have staff and expertise to support the entire life cycle of a game, including marketing, distribution, and business development (see Box 6-1).
Another business model has also failed to gain traction in Mayo’s (2009b) view. In this model, a large commercial gaming company with knowledge, investment capital, and marketing expertise would develop and market games for science learning. However, the typical business model of entertainment companies—an enormous up-front investment in game development, including high-quality graphics, followed by millions of sales to individuals within a few months of release—is not aligned with educational markets. Entertainment companies are not familiar with educational markets or how best to market to them, and they may not view these markets as potentially profitable. Uncertain about the potential sales revenue of educational games, these companies have made few efforts to develop educational games and have not established distribution channels to market them, either to schools or to the public.
A variation of this model would tap the knowledge and marketing expertise of textbook publishers as a way to develop and distribute science games. However, these companies’ systems for selling print books—including their sales incentives and outreach to state textbook adoption committees—are poorly suited to marketing learning games. Textbook publishers generally focus on selling textbook editions that may remain unchanged for up to six years, but computer operating systems and software are revised frequently, so an educational game requires ongoing maintenance and upgrading. For all these reasons, efforts to market serious games through commercial textbook publishing companies have faltered.
Designing Games for Consumer Acceptance
Some observers attribute the limited sales of educational games to date to the lack of a commercial-quality example or market leader (Mayo, 2009b). Most educational games are produced for less than $1 million, while commercial games often cost $10-$100 million. A Sony Corporation executive (Hight, 2009) observed that, in the world of commercial gaming, graphics are very important. In 2009, half of his 135-person team working on the game God of War 3 was devoted to creating detailed three-dimensional graphics (the total project budget was over $40 million).
Mayo (2009b) argued that such large investments in graphics may not be necessary for consumer acceptance of educational games. She noted that Whyville has attracted 5 million regular players, although it cost only $30,000 to develop and incorporates simple two-dimensional graphics.* The Sony representative (Hight, 2009) agreed, noting that commercial publishers look for a variety of other attributes—besides expensive, detailed graphics—when considering the potential audience appeal of a game. He said that a coherent artistic vision throughout the game is very important, as illustrated by the small, web-based game flOw, created by a university student as a master of fine arts project. Hight invested less than $500,000 to purchase and market the game, which is sold on line through the PlayStation Network. He observed that game distribution channels are beginning to move beyond a handful of large retailers, which will accept only a few new game titles each year due to their limited shelf space. Games are increasingly marketed directly to consumers on the web—a trend that facilitates sales of inexpensive games (including educational games) in niche markets. At the same time, new authoring tools are reducing the costs of graphics design (Mayo, 2009b).
A key element in design for consumer acceptance is to repeatedly test the game’s acceptance by the target audience (Gershenfeld, 2009). Hight (2009) noted that Sony game development teams invite young people (the target audience) to play games in a special room, where their facial expressions and the content on the screen are recorded. Experts thoroughly observe the players as they navigate through every stage of the game, taking notes on what the players do and do not understand and when the players are enjoying themselves. Extensive testing is important because potential customers can be very quickly turned off (within 15 seconds) by a weak interface. This extensive consumer testing during the development process is likely to be as important with educational games as it has proven to be with purely commercial games.
Marketing Barriers in K-12 Education
In interviews, executives of companies engaged in developing and marketing educational games identified major barriers to marketing educational games to schools and school districts (Mayo, 2009b). Although most of the games discussed in these interviews do not focus specifically on science learning, the barriers identified are directly relevant to science games.
The executives pointed to a lack of distribution channels as the primary barrier to successfully marketing games in K-12 education. They emphasized the challenges to reach a point of purchase, noting that it is difficult, labor-intensive work to market games to schools and school districts. This work has included a variety of marketing approaches. Officials of two companies reported marketing games directly to teachers, approaching them through teacher conferences and websites. Teachers have purchased both individual games and classroom site licenses, using classroom supplies budgets and their own personal funds. However, one company found it more profitable to target school districts, marketing to curriculum coordinators and instructional designers with access to state and federal funding sources. Although the company experienced lengthy waits before licenses were purchased, the licenses were profitable and tended to be renewed for many years.
Another approach is to bundle a game with teacher professional development. One company has partially supported an educational game through sales of professional development classes, providing the game as part of the total package. In another approach, Numedeon, Inc. marketed the game Whyville directly to students at home, encouraging them to play the game and engage their class. In this case, no purchase was necessary, as the game is free to all users. Finally, the Kauffman Foundation has distributed educational games to schools by encouraging game developers to place older or demonstration versions of their games on state-financed laptops distributed to middle school students in Maine and Michigan. The developers obtained free exposure and potential sales for commercial variations of the same games.
The executives observed that, even if this primary barrier can be overcome and distribution channels are successfully established, several other barriers may limit the use of games in schools (Mayo, 2009b). First, as noted in Chapter 3, teacher professional development is essential for effective use of games, and companies are beginning to address this barrier by providing professional development in a variety of online and in-person formats. In addition, there may be barriers to installation and use of third-party software on school systems’ computer networks. For example, playing the game should not require video cards, because most student and school administrative office computers have either low-grade video cards or none at all. Similarly, the game should require only modest amounts of random access memory (RAM). Because delivering games on the Internet helps to address these
barriers, several experts believe that this approach is promising for scaling up educational games. However, access to the Internet from classroom computers may be constrained by limited bandwidth. In addition, the lack of a computer for each student in many schools limits the potential of games to support individualized learning.
Acceptance of educational games in schools may also be constrained by time and organizational limits. One response to these barriers is to design games that present educational content in short time increments of no more than 40 minutes (the typical class period). Some games present content in less than 10 minutes, allowing the teacher to flexibly integrate them into daily lesson plans. However, in this approach, students have no opportunity for the kinds of extended game play in which they may engage with recreational games—the very kinds of extended game play that have great potential to enhance science learning.
Concerns about protecting individual privacy can also pose a barrier if the game software requests self-identifying information. One solution is to avoid designing such requests into the software, and another is to involve the teacher in entering student contact information and storing it securely. Although delivering games on the Internet can reduce technology hurdles, it also raises privacy and security concerns. These concerns have been addressed in a variety of ways, including placing the game on a dedicated server that only students and teachers can access, preventing navigation to sites other than those related to the game, running background checks on all adults requesting access before allowing them to enter the students’ virtual space, and using other types of controls.
Finally, funding limits represent another barrier to increased use of educational games in schools. Inadequate funding can limit the ability of state or school district technology coordinators to purchase site licenses for games, to update computer hardware and software, to enhance Internet access in classrooms, or to provide teacher professional development. This barrier has become more significant, as the current economic downturn has resulted in major cuts to state and local education budgets.
All of these barriers to greater uptake of games in K-12 education, including the primary barrier of a lack of distribution networks, are in various stages of being addressed. Nevertheless, these barriers greatly limit the use of games. In 2009, educational game companies reported having sold only about 200-300 school site licenses for each game, reaching less than 1 percent of the 99,000 public schools in the United States.
Marketing Barriers in Higher Education
Markets for educational games in higher education have more in common with general consumer markets than with K-12 markets. In higher education,
as among the general public, an individual can make a final purchasing decision. A professor has greater freedom than a schoolteacher to dictate what textbooks, games, or other curriculum materials will be used in the course and to direct the campus bookstore to buy these materials. The barriers to increased use of games in schools—such as short time segments, state education standards, and technological constraints—are much smaller in higher education. Nevertheless, science learning with games remains rare in college and university classrooms. The exceptions tend to be classes taught by the professors who are also developers of educational games.
Marketing Games to the Public
The general consumer market is much larger than the K-12 and higher education markets, and distribution is much easier, as shown by sales figures for games that have been sold in both markets. For example, Software Kids has sold site licenses for Time Engineers to about 300 schools. However, when the company bundled the game with other software in a “Middle School Success” packet offered to the public through stores, it sold 80,000 units. Likewise, Muzzy Lane Software has sold site licenses for Making History to only about 250 schools, but was able to sell 40,000-50,000 copies of the consumer version when selling directly to the public.
Parents are the primary purchasers of educational software aimed at younger children, and, as shown by the sales figures above, they continue to play a role in purchases of games targeted to middle school. Parents constitute an important initial target market for scaling up the use of games for science learning. Parent interest in games—expressed through game purchases, observing their children at play, and playing the games with their children—could both increase science learning in the informal context of the home and also encourage greater use of these games in schools. However, parents seeking to advance their children’s educational success may want to know more about the effectiveness of a particular game or simulation in supporting science learning before purchasing it. Mayo (2009a, p. 81) observes that “the ability to distinguish between a high- and low-quality product will be essential to the growth and credibility of game-based learning as a field.”
By late elementary school, children increasingly make their own decisions about what games to purchase (see Chapter 4). One way to overcome the problem that middle and high school students may avoid a “brainy” game is to sell the game through hardware that is typically purchased by parents. For example, Numedeon partnered with Dell to include Whyville, preinstalled on all Dell computers sold at Walmart. Often, the hardware company provides the game developer with a modest payment for each computer (or other hardware unit) sold, which can add up quickly. The game developer can later sell upgrades and add-ons to those hardware purchasers who become
interested in the game. Another way to address this problem is through corporate sponsorship, with advertisements and placement of brands within games. In this approach, the corporate sponsor provides some immediate revenue to the developer of an educational game even if sales to preteens and adolescents are slow; it could buy time to implement other marketing strategies to reach this group.
Adults make up a large segment of the general public market, potentially providing a source of sustainable revenue to developers of educational games (Mayo, 2009b). For example, many adult history and strategy buffs have purchased Making History. As noted in Chapter 1, WolfQuest has attracted adults as well as young people. Adult players of Nintendo DS, a popular handheld gaming device, often purchase educational and self-improvement software; adult gamers comprise one of the fastest growing market segments for Nintendo.
Distribution of games to the general public is facilitated by the presence of “turnkey publishers,” who will carry out all manufacturing and marketing-related tasks, such as packaging, obtaining a rating from the Entertainment Software Rating Board, advertising, bundling with related products, and negotiating sales agreements with retail outlets. However, the game developer who uses this distribution channel loses both control of the product and a share of the profits to the publisher. A game developer may also hire a distributor, which does no marketing or advertising but can inject the game into a network of stores with which it has agreements.
The effectiveness of the Internet as a distribution mechanism depends on the website hosting a game. If it is not well known, the game may be invisible to most consumers. However, it may still be possible to increase awareness of a new game by constant, aggressive efforts to submit it to game review sites, game award contests, product review columns, and appropriate social networking sites.
The company executives interviewed by Mayo (2009b) reported few barriers to consumer acceptance among the general public. In fact, they noted that the public’s interest in learning generally enhances acceptance of educational games.
Marketing educational games to the public is constrained by far fewer barriers than exist in K-12 education. Distribution is facilitated through publishers and pure distributors, and consumer acceptance is in line with other learning products. All other factors being equal, games designed for science leaning should reach scale first and foremost in the public market. However, few educational games have been actively and professionally marketed to the public, and none has been professionally marketed in higher education. This is due partly to the lack of a commercial-grade product to bring to market, which is related to the lack of funding to support the required final hardening, consumer acceptance testing, and refining. It is also due
to academic developers’ lack of understanding of the complete life cycle of game development, marketing, and maintenance.
ALTERNATIVE PATHWAYS TO SCALE
The committee identified two overall models for bringing science learning games and simulations to scale: (1) a traditional “top-down” model of sales and distribution of games or simulations and their supporting systems to schools and school districts, and (2) a “disruptive innovation” model (Christensen, Horn, and Johnson, 2008). In the disruptive innovation model, widespread use of simulations and games for informal science learning by individuals and families would demonstrate a dramatic improvement over traditional science education, leading school systems to greatly increase their adoption of simulations and games. Success in this second model, elements of which could be emerging, could prove to be a way to enable wider use of games in the first model.
Within the disruptive innovation model, there are a number of promising pathways toward scaling up the use of simulations and games. One is represented by the growing number of small commercial publishers of educational games. Other pathways include nonprofit organizations taking on more of the roles of game publishers and a decentralized “commons” approach that encourages collaborative development and dissemination of games and simulations. The following section describes these three pathways, followed by a sketch of the possibilities for a traditional, top-down model of scaling up games through school systems.
Small Commercial Publishers
Mayo (2009b) observed that a business model of modest up-front investment in game development followed by long-term returns appears to be working for a new group of small-scale educational game developers, such as Muzzy Lane Software, 360Ed, Tabula Digita, Numedeon, and Software Kids. These companies have sold tens of thousands of copies of educational games. Unlike commercial games, which may be popular for only a few months, academic games should sell for years, as the scientific principles and concepts underlying the game remain unchanged. Although the content of an academic game need not change, the game will require ongoing support to keep pace with changes in its supporting hardware and software.
Nonprofit Organizations as Game Publishers
Gershenfeld (2009) proposed that science learning games could be scaled up if game development funders—foundations, nonprofit organizations, univer-
sities, government agencies—took on the roles of commercial game publishers. Publishers are responsible for the entire game life cycle, including marketing, distribution, business development, and ongoing support for the games. Lacking expertise in these areas, funders have invested millions of dollars in educational games that have reached only a handful of players—because they were not fun to play or were not effectively marketed or distributed.
Nonprofits could carry out a rigorous screening process to decide which games to fund and at what level—just as commercial publishers do. When considering a potential game concept, nonprofits would ask such questions as:
Who is the target audience (e.g., consumer, school system, library) and the purchaser (e.g., child, parent, teacher, department head)?
What is the desired learning goal or impact (e.g., science learning goals, a role in the core curriculum, a supplement)?
What evidence is there of market demand? Answering this question may require testing the game concept in target markets.
What is the best game platform to reach the target audience? This involves considering technology options (alternative video consoles, handheld devices, personal computers, etc.) for the target audience.
What is the business model? Will the game be sold as a product (e.g., by retail, by download) or as a service (subscription, micro-payment, etc.)?
What are the financial requirements and expectations? This will include considering how best to balance the potential financial and social/educational returns and deciding on an appropriate budget for the project.
What is the most effective team to develop the game? An understanding of who the audience is, the platforms, and the business models is necessary to select the best development team.
Is there a well-thought-out development plan with natural funding milestones?
Who is the most effective team to market the game?
What is the methodology and plan for assessment? This involves ongoing review of the project and repeated testing with target purchasers to ensure it is on track.
What is the overall threshold to approve the game project? This includes deciding who is on the “greenlight” committee and carefully defining the necessary milestones and approval process.
Nonprofit organizations will need to develop new knowledge and skills to answer these questions, Gershenfeld (2009) observed, and they will also
need to learn from their successes and failures. Creating partnerships with individuals, teams, and organizations can help them to build the needed knowledge and skills. In this pathway, nonprofits would use a similar business model to that of commercial publishers—developing and marketing a few blockbuster science learning games.
A Decentralized Approach
Osterweil (2009) advocates a decentralized pathway to scaling up educational games by building on the burgeoning independent games movement. He notes that a typical commercial game has only a few weeks either to recoup its investment in retail outlets or to find itself consigned to the remainders bin. To achieve this rapid payback, commercial games require very large marketing budgets, which may equal their development costs. Independent games, in contrast, are often distributed online, an environment much more conducive to targeted marketing and niche sales. Because of their small size, they can be created in a fraction of the time and cost required for a large commercial game. Many different groups and individuals, including students and industry professionals working in their spare time, are creating a variety of independent games, some of high quality. These developments contrast with the current, centralized approach to developing games for science learning, in which foundations and other funders have invested heavily in a few academics and small firms, who in turn produce a few large educational games.
The current trend toward web-based delivery of games would facilitate this decentralized pathway to scale for several reasons. First, web delivery is more effective for reaching small, niche markets. It allows consumers to download free demonstrations or make incremental purchases, a form of marketing that favors the independent developer without a large advertising budget to build demand. Second, web-delivered games can reach K-12 students and schools, overcoming some of the hardware and software barriers described above. Third, languages for creating web-delivered games are becoming increasingly sophisticated; some such languages can be used to create more than flat, simple two-dimensional games. A web-based market for science learning games could serve as a laboratory for diverse approaches, allowing best practices to emerge, rather than be preordained by a few experts.
Osterweil suggests that funders could create market conditions that would facilitate this decentralized pathway by supporting the creation of shared web platforms for development and distribution of educational games. Current examples of such platforms—the iPhone app store and the Android market—provide models for creating a new platform specifically to support science learning games. Each has inspired creative development of myriad applications by providing an easy development platform and lowering barriers to entering the marketplace. Another example is BrainPOP, a privately
held company that has created a site with videos on a wide range of school topics, indexed by grade and subject area and keyed to state educational standards for easy use by teachers. Thousands of schools have purchased annual subscriptions to access these materials.
A Top-Down Pathway to Scale
Zelman (2009) suggests that top-down educational policies can facilitate widespread adoption of simulations and games for science learning, overcoming the marketing barriers in K-12 education discussed above. She describes public education as a system, with classroom instruction at the center. Four related elements affect classroom instruction: (1) local, state, and federal accountability policies; (2) student, family, and community support; (3) educator professional development; and (4) state fiscal policies and educational technology plans.
Current developments throughout this system present new opportunities for scaling up the use of games and simulations for science learning. At the national level, states are joining to develop common core educational standards, including science standards that are expected to be higher, clearer, and fewer than current science standards. At the same time, the U.S. Department of Education is supporting consortia of states in developing shared assessments. These new standards and assessments may incorporate the broad range of science learning goals that games and simulations are well suited to advance.
Zelman argues that states and school districts are becoming more interested in technology as one route to improving the effectiveness of instruction and enhancing student performance on assessments. To foster this interest, the U.S. Department of Education (2010) recently published a draft National Education Technology Plan outlining local, state, and federal technology policies in the areas of learning, assessment, teaching, infrastructure, and productivity. Since 2002, the Department has provided grants to states to assist them in purchasing learning technology. As part of the grants program, the states are required to create educational technology development plans.
Zelman (2009) identified several state policies that might encourage wider use of simulations and games, including revising curriculum purchasing procedures that currently focus on textbooks to facilitate statewide software and hardware purchases. She advocates focusing state educational technology plans on the goals of ensuring statewide availability of computer hardware and software and broadband access, eliminating firewalls while maintaining security, and assisting in the distribution and marketing of educational games. She argues that such policies would increase science learning, not only at school, but also through family gaming at home. Such technology policies would also facilitate the development of common educational data
standards across the 50 states, making statewide performance data highly accessible, including for teachers, along with digital learning objects and online mentoring and professional development.
At the level of the individual school, Zelman suggests designating some schools as gaming schools and laboratories. At these schools, data would be gathered to evaluate the effectiveness of games in helping students achieve educational standards and the types of assessment data that can be gained from games. Other topics that could be explored in these schools include identifying the types of knowledge teachers require to use games effectively to support science learning; the financial costs of hardware, software, and teacher training, and how to budget for these costs; the roles of games in supporting informal learning after school, at home, and with peers; and the potential for collaborating with public radio and television stations.
Increasing the uptake of games for science learning is affected by a variety of barriers. Some of these barriers slow development and sales of games in both formal and informal learning contexts, while others are unique to the formal contexts of K-12 and higher education.
Conclusion: Several barriers slow large-scale development and use of games and simulations for science learning in K-12 and higher education. There is not yet a coherent market for either games or simulations in schools that is analogous to the textbook market. Increased use of games and simulations in schools and universities will require clear alignment with curriculum and professional development support for teachers or faculty members. These issues are dealt with primarily at the local level in highly decentralized structures, posing a serious barrier to scaling up the use of games and simulations. If districts, schools, and universities express interest, this will encourage the development and use of these new learning technologies.
The committee identified two basic models for scaling up the use of games and simulations for science learning.
Conclusion: There appear to be two basic possible models for reaching scale: (1) a traditional top-down model of sales and distribution of games or simulations and their supporting systems to schools and school districts and (2) a model of sales and distribution to parents, students, and individuals for informal learning. Success in the second model, elements of which could be emerging, could prove to be a way to enable access to the first model.
The committee explored alternative pathways for reaching scale. Pathways within the second model include the small commercial game or simulation publisher, the “nonprofit” publisher with foundation or government agency funding, and a decentralized approach that would support collaborative game development and distribution. A few small commercial publishers have successfully marketed educational games to parents and children. Parents could potentially constitute a large market for increased sales of games and simulations designed for science learning.
Conclusion: Parents of K-6 students concerned about their children’s educational progress could constitute a large and important initial market for increased sales and use of science learning simulations and games. However, parents may have questions about the educational value of various simulations and games, and these questions could potentially be addressed through the creation of a respected, independent, third-party system to evaluate and certify educational effectiveness.
The availability and quality of computer hardware and software systems greatly influence the extent to which individuals access and use simulations and games for science learning, in both formal and informal learning environments. Computer technology continues to change rapidly, requiring ongoing support for simulations and games.
Conclusion: Simulations and games for science learning require a sustained approach. Because a game or simulation needs to be updated and improved on an ongoing basis, it is not enough to simply develop and launch a standalone game or simulation. An ongoing development, research, and support effort is required for dissemination at scale.
A large number of stakeholders—including commercial entertainment companies, academic researchers, state and local education officials, game developers, and teachers—play a role in the use of simulations and games for science learning. Bringing these stakeholders together in partnerships could help bring research and development of simulations to scale.
Conclusion: Partnerships that include industry developers, academic researchers, designers, learning scientists, and educational practitioners could play an important role in scaling up research and development of games and simulations.