DESIGN IN CYCLES - Don't Try to Solve Everything at Once



Students think design is linear: Idea > Plan > Section > Render. Professionals know it's a cycle: Draw > Look > Critique > Redraw. Hundreds of times.

Frederick says: The first idea is always bad. The tenth idea is better. The thirtieth is good. If you fall in love with your first idea, you will defend a bad idea for months.

The architectural thinking cycle has 4 steps: 1- Divergent - generate many ideas fast, no judgment. 2- Convergent - select best 2-3. 3- Develop - detail them. 4- Test - with models, with sun, with structure, with client. Then loop again.

Most junior architects skip step 4. They develop without testing. Then in site, structure fails, or sun glare makes space unusable. Cost: millions.

The tool that forces cycles is the physical model. Your brain lies. You think that corner works. Model tells truth in 10 seconds: It doesn't. That's why Frederick says: Make models, and make them quickly and roughly. A 10-minute cardboard model is worth 10 hours of SketchUp.

For technical office, this means: Don't finalize shop drawing in first version. Print it, look at it on wall, walk away, come back. Show to MEP, show to structure, show to site engineer. Each look is a cycle. Each cycle saves site problems.

Instagram post idea: Show 5 iterations of same corner - from bad first idea to good final. People love seeing process, not just final product. It proves you are thinker, not just renderer. Caption: "My first idea was trash. Here is iteration #7 that worked."

Remember: Architecture is not about having good ideas. It's about having process to turn bad ideas into good ones through cycles.

Comments