Posts Tagged ‘serious games’

On Architectural Creativity (Lessons from Game Design)

July 26th, 2009 No comments

I’ve spent the past two years working in a very interesting work environment, surrounded by a very innovative and talented group of individuals. For the record, our mandate was nothing less than to completely transform learning from a dull, staid, painful and uncertain process to one that was engaging, immersive, fun and robust. I’ve worked on a good half dozen learning games in this period, where we essentially took learning objectives that would normally have been the focus of a PowerPoint deck, wrote these up in the form of a curriculum, and brought them to life in the shape of interactive learning simulations.

There were three distinct groups that supported the effort of designing and implementing these serious games:

  1. Designers and managers who ensures that the learning, game design and assessment/analytics were meaningful
  2. Art, sound and content creators who took the designs and reified them
  3. Software engineers who built the underlying interactions and processes, and stitched together the pieces of the games into the final product.

Obviously, I’m missing many people in the cast of thousands here, there is no mention of product/portfolio management, quality assurance, customer-service, sales etc. These are all support functions, and although they are critical to the success of the firm, they do not actually directly create the game. They may exercise creativity during the course of their work though.

I’ll talk a bit more about my personal role(s) in this process at some other point, however the key point I’d like to discuss in this article is creativity. I had numerous opportunities to observer creativity (and the lack thereof) in action while working on these games, many of which rivaled Hollywood blockbusters in their complexity and the ambition behind their vision.

What I realized recently, after two years of observing some occasionally bewildering behavior, is that creativity is all about how you deal with ambiguity. It can be a truly counter-intuitive process by which you first introduce ambiguity (by adding choices) and then eliminate it (to arrive at the final solution). As a part-time machine learning researcher, I can even point out a class of algorithms where computer software ‘practices’ creativity to learn and model sometime very useful ‘solutions’ in the form of patterns. These bootstrapping algorithms essentially start with a seed set of rules (or examples) and ‘invent’ patterns derived from new data available that have a decent likelihood of forming a rule in the desired classifier. However, the ‘creative’ process does not end there, as each of these ‘candidate’ pattern rules need to be evaluated against what is known (via the seed set or other identified ‘ground truths’) in a process that ruthlessly eliminates most of the patterns. Interestingly, human creatively works in a very similar way.

The individual most successful at the creativity game keeps the end-goal of the product firmly in their mind, while they explore the many options available to achieving this goal. Thus, the most creative individual will bring in ideas from other fields (which may not be readily apparent to others) and practice lateral thinking.

There are many celebrated successes of lateral thinking around us. Consider ‘object oriented design‘ in software engineering which is based on the successful application of component strategies in electronics and hardware. In hindsight, it makes absolute sense, build components that hide all the complexity within, and you can afford to reuse and repurpose software, and create sophisticated designs (by assembling mature components) that are much more ambitious than would be humanly possible if you were building an completely integrated software. This is a key reason why Intel’s CPUs, which were originally designed to abstract out common processing function for calculators and traffic light controllers (which were quite surprisingly Intel’s specialization in the early years), now form the ‘brains’ behind computers throughout the world. Some ‘creative’ engineer figured out the absurdness of reinventing the processing component for every new project, and eliminated a complete waste of resources by serendipitously architecting this as a separate module.

However, it is almost important to emphasize at this point that the creative process does need to be focused towards the end-goal as well. Architects (who are undoubtedly creative people who need to deliver working products) speak fondly of the phrase ‘form follows function’. Indeed, the most classical architect broke this down into three distinct areas, and gave us a very useful set of evaluation metrics. Vitruvius argued that any architecture (and the resulting product build) can be measured based on its utility, robustness and aesthetics (although Vitruvius referred to this as Commodity, Firmness, and Delight in the original text, which can be a bit confusing). The same metrics work wonderfully well for modern software, which can range in quality from a veritable tower of cards to the Taj Mahal.

Although it is true that the first solution you think of, is quite likely to be successful (thus proving the utility of early commitment in hostile environments), it is important to consider multiple paths to your end-goal if your role is that of a creative. Generally, you have the luxury of more time than firefighters and hostage negotiators, on whom the early commitment studies was based. It is also much simpler to make changes early in the process while dealing with paper prototypes than when you are stuck with the sunk cost represented by semi-built buildings or half-developed code. I would argue that a project where the entire focus changes mid-way, the initial model is nixed, and a ‘reset’ takes place is one that has been designed by an intellectually lazy architect, who takes pleasure in letting others ‘live out’ his or her experiments. It’s a dangerous practice that is insensitive to the people who are paying for the project, for the customers who need this product and the team that is living out the development of this solution and can literally burn-out from lack of perceived progress.

For creative individuals, there is an important tradeoff between keeping your options open, and delivering a useful end-product. You absolutely need to iterate through various architectural options while designing the final solution. However, when it comes to the final push to implement, it is absolutely critical to have a design already nailed down — the commitment should have already taken place. This design may not be detailed at the level of every piece of content or code required or outline all necessary activities, however there should be a clear acceptance criteria that specifies the form (in terms of utility provided, robustness/operating environment supported, and visual manifestation) and the underlying event processing model (which is the game designer’s equivalent of the Hollywood narrative script).

There should not be any doubts about the commodity, firmness and delight anticipated by the availability of the final product at this stage. Any deviation from this axiom reflects a lack of responsible behavior, or an an abdication of design ownership. In these scenarios, you may want to consider resurrecting the ancient pirates tradition of overboarding and make the architect walk the gangplank.