Here is a claim that sounds like provocation but is simply a mathematical fact: every flat map of the world is wrong. Not slightly imprecise — genuinely, unavoidably distorted. And this is not a failure of cartography. It is a property of geometry that no amount of cleverness can escape.
Understanding why turns map-reading from passive absorption into something more useful: the ability to see what a map is emphasising, and what it is quietly sacrificing.
The orange peel problem
Peel an orange and try to press the skin flat on a table. It will not go. The peel tears, or it wrinkles, or you stretch parts of it thin. There is no technique that flattens it perfectly, because a sphere’s surface simply cannot become a plane without deformation.
The mathematics behind this was settled by Carl Friedrich Gauss in the nineteenth century. Curved and flat surfaces have a measurable, intrinsic difference that cannot be removed by bending or stretching. A sphere and a plane are, in this precise sense, incompatible.
The Earth is a sphere (near enough). A map is flat. Therefore every map distorts. The only question is what it distorts, and by how much.
The four things a map can get wrong
Cartographers think in terms of four properties. A projection can preserve some of them, never all:
- Area — whether regions have correctly proportioned sizes relative to each other.
- Shape — whether local outlines and angles look right.
- Distance — whether measured separations are accurate.
- Direction — whether bearings between points are true.
A projection that nails area will bend shapes. One that keeps shapes faithful will inflate some areas. This is a genuine trade-off, not a solvable puzzle — which is exactly why hundreds of projections exist. Each is a different answer to the question “what am I willing to lose?”
Mercator: the famous one
The Mercator projection, published by Gerardus Mercator in 1569, is the map most people picture when they think “world map”. It is also the one most criticised — often unfairly.
What it does brilliantly
Mercator preserves angles. This means a straight line drawn on a Mercator map represents a constant compass bearing. For a sixteenth-century navigator, this was transformative: plot a straight line, hold that bearing, arrive. No continuous recalculation across a curved surface.
It was designed for sea navigation, and for sea navigation it is superb. It is still used in marine charts today, and it underpins most online slippy maps — partly because preserved angles mean shapes look correct when you zoom into a city.
What it sacrifices
Area — and dramatically. To keep angles true, Mercator stretches land increasingly as you move away from the equator. Near the poles the stretching becomes extreme, which is why Mercator maps typically cut off before reaching them: at the poles themselves, the projection goes to infinity.
The consequences are startling once you notice them:
- Greenland appears comparable in size to Africa. Africa is roughly fourteen times larger.
- Alaska looks similar in size to Brazil. Brazil is about five times larger.
- Scandinavia looks vast relative to India. India is considerably larger.
- Antarctica becomes an enormous smear along the bottom edge.
Because the distortion grows with latitude, and because Europe and North America sit at high latitudes while much of Africa and South America straddle the equator, the effect systematically enlarges some regions relative to others. Whether Mercator’s original intent had anything to do with this is doubtful — he was solving a navigation problem — but the visual result has shaped a lot of people’s mental image of the world.
The main families
| Projection | Preserves | Distorts | Good for |
|---|---|---|---|
| Mercator | Angles, local shape | Area, badly at high latitudes | Navigation, web maps |
| Gall–Peters | Area | Shape — noticeably stretched | Comparing land areas honestly |
| Robinson | Nothing exactly | Everything a little | General-purpose world maps |
| Winkel tripel | Nothing exactly; balanced | Everything slightly | Atlases and reference maps |
| Azimuthal equidistant | Distance and direction from one centre point | Everything else | Radio range, flight planning |
| Goode homolosine | Area | Continuity — the map is interrupted | Thematic global data |
Robinson and Winkel tripel are worth noting as a distinct philosophy: rather than preserving one property perfectly, they distribute error so that nothing looks badly wrong. They are compromises by design, which is why atlases favour them.
Goode homolosine takes a different route: it keeps areas true and pays the price by cutting the map open across the oceans, producing that distinctive interrupted shape.
Beyond geometry: the choices maps make
Projection is only one decision. Every world map involves others that shape perception just as much:
- What sits at the centre. There is no geometric reason for Europe to be in the middle. Maps published in the Americas often centre the Americas; East Asian maps often centre the Pacific. Each looks “normal” to its audience and slightly odd to everyone else.
- Which way is up. North-up is convention, not physics. Space has no up. South-up maps are perfectly valid and deeply disorienting, which tells you how much of map-reading is habit.
- What gets included. Which cities are named, which borders are drawn, which disputed territories appear and how — these are editorial decisions, and sometimes political ones.
How to read maps better
- Ask what the map is for. A navigation chart and a population map have different jobs and should use different projections. Judge a projection against its purpose.
- Be suspicious of area comparisons. If a map is making a point about size, check whether it is equal-area. If it is not, the argument may be an artefact.
- Use a globe as your reference. A globe is the only undistorted representation. When something looks surprising on a flat map, check it against one.
- Notice the centre and the edges. Whatever sits at the centre gets emphasis. Whatever is cut at the edge gets diminished.
The point
There is no “correct” world map, and looking for one misunderstands the problem. There are only maps that are appropriate or inappropriate for a given question. Mercator is the right tool for crossing an ocean and the wrong tool for comparing continents. Gall–Peters is honest about area and clumsy about shape. Robinson is wrong about everything and useful for almost anything.
The skill worth having is not knowing which map is right. It is knowing what each one is trading away.
If you would like to go further into cartography, spatial thinking and how geographic data gets represented, the geography courses on Cursa pick up from exactly here.
