visualization-strategy-and-critique

Start by defining the decision the view must support, then classify the data and required comparisons. Bars and dots suit category comparison, lines and small multiples suit time, scatterplots suit relationships, and maps only when geography is analytically meaningful.

Critique semantic truth first, then encoding, composition, and implementation. Check whether the title states a claim, the key comparison gets the strongest encoding, scales and baselines are trustworthy, and the figure survives as a static screenshot.

Use a Gantt chart when task spans, milestones, dependencies, critical path, or baselines are the main evidence. Use Kanban, tables, or dependency graphs when workflow state, lookup, or topology matters more than planned schedule bars.

Use scrollytelling when a linear author-led path helps readers understand staged change. Choose a stepper when discrete states, replay, or direct access matter, and small multiples when simultaneous comparison is clearer than sequential reveals.

Avoid 3D bars, tilted pies, dual axes without careful explanation, rainbow ramps for ordered magnitude, equal-weight KPI tile dashboards, decorative particles, and tooltips as the only place important values appear.