Deliberately explore the most unlikely combinations

The combinations intuition skips are precisely the ones the method is designed to find.

Why it works

After filtering for feasibility, most practitioners choose from combinations near their existing ideas — which converts the method back into brainstorming with extra steps. The morphological method’s unique value is that it includes the combinations at the periphery of the matrix: high-value but low-prior-probability combinations that never surface under normal ideation because they feel counterintuitive. Actively selecting and exploring these outlier combinations is the step that produces genuinely novel output.

How to do it

1. After filtering, identify the 3–5 combinations you would not have generated intuitively.
2. Spend time with each: what would it look like if this combination were fully developed? What problem does it solve that the intuitive combinations do not?
3. Present the strongest unlikely combination alongside the strongest intuitive one when evaluating options.

Evidence

Research on cognitive fixation shows that solution quality improves when people are shown examples outside their initial frame; the morphological matrix serves as an externalized anti-fixation device that surfaces those examples systematically. Jansson and Smith found that even being shown an example measurably pulls subsequent designs toward it, which is the mechanism that makes the periphery of the matrix — the cells intuition never reaches — the valuable ones. (mechanistic)

Fixation research supports the principle; the specific claim about unlikely matrix cells as the most valuable is a logical extrapolation rather than a directly studied effect.

Sources

• Jansson & Smith (1991), design fixation, Design Studies — documents how initial examples constrain subsequent generation and how counter-examples help
• Jansson, D. G., & Smith, S. M. (1991). Design fixation. Design Studies, 12(1), 3–11.

Common mistake

Selecting from the matrix the way one would select from a brainstorm list — clustering toward familiar and plausible options, which makes the matrix a slow and complicated version of what brainstorming already does.