The same dataset that's readable on a wide monitor turns into a smear on a phone screen

What it is

The same dataset has entirely different density ceilings on a 27-inch fullscreen display, a half-screen laptop, a phone in portrait, and a projector: shrink the canvas and previously separable points begin to touch, previously crisp lines smear into a mass. Density must be judged at the final presentation size — "looks fine at mock size" is not evidence.

Why it happens

The size dependence follows directly from the pixel budget: the minimum distinguishable mark size is roughly constant (bounded by visual acuity), so more canvas pixels carry more marks; halving the canvas quarters the mark ceiling (area scaling), and scaling the mark count linearly overflows it. Resolution acts more quietly: high-DPI screens let smaller marks stay legible, but projectors, low-grade displays, and print consume that headroom — a retina mock may retain only a fraction of its distinguishable marks once projected. Responsive charts add a runtime layer: the density must be recomputed as users resize windows, and an unrecomputed chart overflows.

Where it stops holding

Size and resolution do not compensate for each other: a high-DPI small screen is not a large canvas — physical size bounds acuity, and no pixel density rescues a thousand-point scatter on a five-inch display. Interactive zoom pushes the "exploration" ceiling higher, but the initial view is still judged at its initial size. Print is a one-way degradation: every interactive density on screen freezes on paper, so print versions get a fresh paper-based density estimate.

Applying it

• Record the target presentation size and device in the chart spec (fullscreen desktop, in-card, mobile, projection, A4 print) and run the density review at that size.
• Re-estimate density at key responsive breakpoints; switch narrow screens to aggregated or faceted variants where needed.
• Verification: run the identification test on the target device (or a scaled screenshot); passing only at mock size while failing at target size counts as a failure.

Related

• Same group: U5.06.1 One chart's mark count is limited by overlap and occlusion · U5.06.2 Past the density ceiling, aggregate, sample, or facet · U5.06.4 High-density charts keep overall shape and lose individual readability · U5.06.5 Solve density in data processing, not by shrinking marks
• Nearby: U11.04.4 In-chart text is sized by final presentation size, not the design file · U5.06.1 One chart's mark count is limited by overlap and occlusion
• Search terms: responsive visualization · target size · pixel density