Geography games compared, by someone who built one

By ·

Read this first. I make one of the games below, so I am not a neutral reviewer and you should not treat me as one.

What I can offer instead is that I have spent a long time watching how people actually fail at geography, and I will tell you plainly which of these to use when mine is the wrong tool. There are several cases where it is.

All of these are free to start, play in a browser, and take a few minutes. They are not really competitors, because they train genuinely different skills. Pick by what you want to get better at.

The short version

If you want to…Play
Recognise country shapesWorldle
Reason your way to a countryGloble
Learn what borders whatTravle
Learn where countries are, from memoryBorderline (mine)
Drill a region hard, for an examSeterra
Read landscapes and street scenesGeoGuessr
Learn economies as a side effectTradle

The daily puzzles

Worldle

You get a country's outline and six guesses, with distance and direction after each one. It is the purest shape-recognition trainer in the genre, and the reason so many people can now identify Chad by silhouette. It has spawned variants for flags, capitals, neighbours and languages.

Best at: shapes. Weak spot: knowing a shape is not the same as knowing where it sits. Plenty of people identify the outline and then place it on the wrong continent.

Globle

Guess any country and the globe colours it by how close you are, warmer as you approach. It is the most forgiving of the dailies because a wild guess still teaches you something, and the deduction loop is genuinely satisfying.

Best at: spatial reasoning and relative distance. Weak spot: you can win by narrowing down without ever learning the answer's location precisely.

Travle

Get from one country to another by naming every country on the shortest land route. It is the only one of these that forces you to hold a whole region's adjacency in your head, which is a surprisingly rare skill.

Best at: borders and regional structure. Weak spot: brutal in parts of the world you do not already know; landlocked Africa can be a wall.

Tradle

Guess the country from a treemap of its exports. Half geography, half economics, and the fastest way to learn which countries live on one commodity.

Best at: economic geography. Weak spot: it is barely a location game, and you can be good at it while still not finding the country on a map.

The big two

GeoGuessr

Dropped into a street scene, work out where on earth you are. Nothing else compares for the sheer pleasure of reading a landscape, and at a high level it is a genuine skill: bollards, road markings, vegetation, the direction of the sun.

The thing to know is that most modes moved behind a paid subscription in 2024, and prices rose again at the start of 2026. As of September 2026 the free tier is one daily challenge of five rounds. That is still a good daily, but "GeoGuessr is free" is no longer accurate.

Best at: visual deduction and depth. Weak spot: cost, and it teaches you to read places rather than to know where they are.

Seterra

The workhorse. Hundreds of click-the-map quizzes by continent, region, capitals, rivers and flags, and the tool most teachers reach for. Now owned by GeoGuessr, with the core quizzes still free.

Best at: deliberate drilling when you need a region cold by Friday. Weak spot: it is a quiz engine rather than a game, so it relies on you supplying the motivation every time.

Where mine fits, and where it does not

Borderline names a country and asks you to tap it on an unlabelled 3D globe, five a day, same five for everyone. It is built around pure recall, and the misses feed a spaced repetition system that brings countries back at widening intervals until they stick.

I built it because I could identify a lot of country shapes and still could not point at Suriname, and none of the games above quite fixed that.

It is the wrong choice if:

What the data says about which skill matters

Because we record where every first tap lands, we can say something the marketing pages cannot: recognising a country and locating it are separate abilities, and most people have far more of the first than the second. Almost everyone reading this could pick El Salvador out of a list of names. Six in ten cannot put a finger on it.

That gap is the argument for playing more than one of these. Shapes from Worldle, adjacency from Travle, raw location from something like ours, and Seterra when you need to force a region into memory before a deadline. They stack better than they compete.

Try today's five countriesfree, no signup, about two minutes