The shipping Rock Band controller, the pad layout of which first found form in thumbnail sketches.
Rock Band, drum controller design. Early on, I did some thumbnails of various drum pad layouts, mostly to help me think visually about the various alternatives we were discussing.
This varied from things like the number of pads, asymmetrical vs symmetrical layouts, and different mappings of drums and cymbals to pads.
By doing quick drawings of the layouts which bubbled up in conversation, I was forced to identify more important criteria (for instance, pads should be readable as one row to aid usability).
This in turn fed into more design discussion. The solutions we gravitated toward at this early stage (four pads in a row, kick pedal mounted to the frame) shaped our first physical prototype, became requirements for the industrial design phase, and can be seen in the shipping drum controllers.
EyeToy: AntiGrav, level design thumbnails. On this project, quick sketches helped us explore big-picture ideas for each level and the main sections within. Ideas came from the whole team through Harmonix's internal newsgroups, and were incorporated into thumbnail sketches and resulting level designs.
A thumbnail sketch for EyeToy: AntiGrav's skyway level. Sections 1, 2, and 3 rough out the big picture of a journey which descends though clouds, rooftops, and down the side of a skyscraper into the city.
The ideas came together first in sketch form, informing the digital 2D maps and 3D level creation that followed. The sketches focused on the main lines through a level, dramatic changes of environment and direction, and most of all, created a sense of journey.
A moment from the rooftops section in the final version of Eyetoy: AntiGrav.
Time pressure limited the amount of exploration, but there was just enough for solid ideas to come together on paper, and the resulting levels were well received.
AiLive, game concepting. The thumbnail approach has helped us generate and evaluate new game concepts at AiLive. In this case, the thumbnail sketches weren't literally sketches, but words: a collection of very short game descriptions in the spirit of thumbnail sketches. Each game thumbnail attacked our new project's brief from a different angle. I'm happy to say a clear winner emerged, and while I can't say much about it yet, I will say it wasn't the first interesting idea we had; an important point we'll return to in a moment.
When creating and reviewing thumbnails, there are a couple of tricks to avoid pitfalls. First, you should never go with the first idea. This is easier said than done. It can be very tempting to develop the first half-decent idea that emerges. Jumping straight into a design or prototyping phase like this is a risky move. No amount of detailed thinking or iteration makes up for a bad early design call. Iteration is like walking up a mountain. You need to do it to reach the summit, but it helps if you choose the right mountain first. Or put more simply, you can't polish a turd.
Avoiding this pitfall is simple. Be diligent. Explore different ideas in thumbnail sketches before developing any single idea too far. Force yourself (or your team) to come up with several approaches to your problem, thumbnail them, and do a little analysis of the options on the table before making big design calls.
Second, make the criteria explicit. It's much easier to evaluate an idea objectively as strong or weak with a list of criteria. What qualities are required in the solution? What's desirable? What isn't desired? Exploring options via thumbnails gets you thinking about the pros and cons, and forces you to get a clear picture of your own criteria. The more explicit the needs, the more defined the solution space is, and the easier it is to generate good ideas. A lack of criteria can lead to idea churn, and can force you to run with an idea you're not sure about.
If there are enough criteria, it can make sense to throw together a quick table to make evaluation easier. For example, list your idea thumbnails (one per row) and your criteria (one per column). Find a simple metric by which to rate each column and complete the table. This helps clarify the big picture, and makes a good conversation point for collaborative decision-making.