vizcn

Which model delivers the most score per dollar (or token), and at what effort level?

when not to use: Needs >=2 runs per model to draw frontier lines; single-point-per-model data reads better as rank-bars.

What made up the flow over time, and when were the waves?

when not to use: When readers must read exact values off the chart — the B-spline deliberately does not pass through the data points (raw values live only in the tooltip).

Who leads on score, with what confidence spread, and what does that score cost?

when not to use: Wide-figure genre (min-width ~700px) — not for narrow containers or mobile cards; for a single metric without spread or economy columns, rank-bars is lighter.

Who clears the bar, with what spread, and at what cost?

when not to use: Single-metric rankings without spread or metric columns — use rank-bars; past ~12 rows the board outgrows a screen and labels stop scanning — use leaderboard-table.

How does the ranking reorder between two contexts?

when not to use: More than two contexts — use lines; when only the gap per item matters — use dumbbell-range.

How do two measurements of the same subjects compare?

when not to use: More than two measurements per subject — use a table or bench-matrix; two rivals over time — use delta-bars.

What do each subject's outcomes consist of?

when not to use: Reading exact segment values (no common baseline) — use a table; time series — use stacked-activity.

What makes up the whole, in one strip?

when not to use: Hierarchies or many small categories — use treemap; comparing composition across subjects — use verdict-stack.

How is the pool scattered across two axes?

when not to use: Few points with runs per entity and an efficiency story — use frontier-board; exact value reading — use a table.

How far apart are the two values per row, and who sits where?

when not to use: Not for more than 2 points per row (use a range strip or box plot) and not for time series — use lines.

How do these models compare across many benchmarks at once?

when not to use: Trends over time or more than ~6 models — use a time-series form or a ranking board instead of a wide table.

What makes up each day's volume, and how does it breathe day to day?

when not to use: Comparing exact values of individual segments across days (inner segments have no common baseline) or fewer than 5 days.

Who is ahead today, and by how much?

when not to use: Not for more than 2 rivals or for absolute magnitudes — use race lines or stacked forms instead.

Is this thing alive — how did its activity move over the last year?

when not to use: Not for exact value reading or short ranges (<12 weeks) — use a column chart instead.

Who leads this ranking, and by how much?

when not to use: Not for time series, part-to-whole composition, or more than ~10 rows — use a table or sparklines instead.

How do these few values stack up side by side?

when not to use: Not for many items (>8) or time series — use rank bars or lines instead.

How is the whole split up, and who dominates?

when not to use: Not for close values (areas are hard to compare precisely) or deep hierarchies (flat only) — use rank bars or a nested treemap library instead.