How to read country comparisons, rankings and Radar

Check the comparison basis first

Matching names alone are insufficient: source, unit and period type must agree. OrbisLens compares published values from the same source series and period. Different indicators can have different latest common periods.

Hypothetical example: 60% minus 50% is 10 percentage points. A relative difference using 50% as the denominator is 20%. Population growth values use %/year; subtracting two rates gives percentage points per year (pp/year). These calculations answer different questions; percent and percentage points are not interchangeable.

Radar converts raw values into relative positions

Each 0–100 axis places a raw value within an eligible same-source, same-unit, same-period country cohort. Higher values point toward 100, without implying better outcomes. Ties share a position, so tied extremes need not land exactly at 0 or 100.

Hypothetical example: in a three-country cohort with values 10, 20 and 30, the middle value has position 50. This is not a quality score. Positions may change with the cohort or values; always read the raw value and period too.

Do not aggregate areas or ranks

Radar shape and area depend on the chosen axes and their arrangement. Changing axes changes the shape, so area is not an overall score. Adding ranks for measures with different meanings and units does not establish a country's superiority.

A ranking denominator counts eligible countries for that indicator and period, and may differ from 195. Missing countries are neither last nor zero. Trends are unavailable with fewer than three usable observations.

How missing values and ties affect ranks

Ranks use published numeric values within one source series, unit, period type and reference period. Countries with missing or ineligible values are excluded, not filled with zero. A global ranking period needs values for at least 80% of active registry countries. A region filter then narrows that eligible cohort and recalculates ranks.

Highest sorts descending and Lowest ascending. Equal unrounded values share a competition rank, with a gap afterwards (for example, 1, 2, 2, 4). Radar uses a different calculation: for a cohort of N, with L lower values and E equal values, position is 100 × (L + (E − 1) / 2) / (N − 1). Tied extremes therefore need not be 0 or 100.

Evidence and official references

These examples read the published common-period values when this page is opened. Editorial dates are not data reference periods; values can change with source revisions.

Check the published data

The notes below use the currently published values. Years and sources may differ across indicators.

GDP per capita, current US$

Japan $35,951 / South Korea $36,227

Reference period: 2025 · World Bank · World Development Indicators

Raw difference, left minus right: -$276. The raw value for South Korea is higher than for Japan. This establishes neither superiority nor the cause of the difference. This is an economy-wide average, not a person's salary or a household's disposable income. The same average can coexist with different income distributions.