I want to discuss the experience of doing value sketches before full-color paintings, because it's advice I've given many times and I want to articulate specifically why it works.

The problem it solves: color is a seductive element. It's easy to get absorbed in beautiful color combinations while the underlying value structure is failing. Color can mask bad values temporarily but the failure eventually becomes visible and is expensive to fix.

The value sketch is a commitment device. Once I've established the value structure in a five-minute thumbnail, I have a contract with myself about where the lights, darks, and mid-tones belong. When I move to color, I'm solving color within the value constraints rather than letting color decisions accidentally modify values.

The secondary benefit: value sketches are faster to iterate than color paintings. I can try three value compositions in fifteen minutes. The composition problem gets solved before I've invested hours in a direction that might not work.