I had played EarthGuessr maybe a dozen times before this experiment — enough to know the game existed, not enough to be any good at it. I am not a geography specialist. I work in software. The last formal geography class I took was in school, more than a decade ago. I cannot consistently name the capitals of the Central Asian republics.
The experiment was simple: one EarthGuessr round per day for 30 days. No looking anything up between rounds. No reading geography articles to study up. Just the round, the guess, the result, and one quick reflection. The question I wanted to answer was whether a daily two-minute geography habit, with zero supplementary effort, would actually change anything.
Week 1: Humbling
The first week was rough. On day one I scored 1,400 points out of 5,000 — a guess that was on the wrong continent. On day three I scored 800 because I confidently placed a Vietnamese rice paddy in the Philippines. The default failure mode in the first week is overconfidence. You see a piece of evidence, jump to a hypothesis, and discover you have been fooled by superficial similarity. Tropical green is not tropical green — there are at least a dozen visually similar tropical landscapes spread across three continents.
By day five I had stopped guessing immediately. I started looking. The first noticeable change was visual: I was spending more time on each round, scanning the edges of the image for clues I would previously have missed. Average score by end of week one: 2,600 points. Not good, but the trajectory was up.
Week 2: Patterns Emerging
Week two was when things started to click. Specific visual signatures began to lodge in my brain: the bright green of irrigated agriculture in arid climates, the herringbone of deforestation along Amazon roads, the rectangular field grid of the American Midwest versus the irregular patchwork of European farming. I started to recognise specific landscape types as belonging to specific continents, even before placing them within continents.
On day 11 I had my first genuinely good round — a 4,600 point guess in central Kazakhstan. I had no idea I knew what central Kazakhstan looked like. Turns out I had picked up clues without realising it: the dryness, the specific colour of the soil, the absence of trees, the geometric Soviet-era field patterns. The guess was almost unconscious. The reflection after the round was: "I think my brain is doing something I am not in conscious control of."
Week 3: A Switch Flips
Week three was the surprise. Average score climbed to 3,400 — a jump I would not have predicted from the first two weeks' trajectory. But the more interesting change was happening outside the rounds. I noticed I was paying more attention to maps in news articles. I caught myself zooming in on satellite imagery on weather websites for no particular reason. I read a travel article and found I had a much clearer mental map of the country being described than I would have a month earlier.
I also started to recognise countries by the shape of their coastlines in casual contexts — a thumbnail on a news site, a passing image in someone's social feed. Geographic literacy was becoming ambient rather than active. I was not consciously studying. I was just looking, and the looking was teaching me.
Week 4: The Surprises
The final week brought a series of small surprises. I correctly identified a round as central Norway based on the specific look of the spruce forest and the fjord geometry — knowledge I did not realise I had. I started to make educated guesses about the latitude of a round before I looked at any other clues, just based on the angle and quality of the sunlight. I noticed myself being able to estimate, within rough bounds, the rainfall climate of a place from the vegetation. Average score for the final week: 3,800.
There was also a less measurable but more interesting change. The world felt slightly different to inhabit. I was more curious about the places news articles mentioned. I caught myself thinking about regions of the planet I had previously regarded as featureless blocks on a map — "Central Asia", "sub-Saharan Africa" — as actual places, with specific landscapes, specific climates, specific peoples. The map had filled in slightly.
What Stuck After
It has been three months since the experiment ended. I have not played daily since — I now play perhaps three rounds a week, when I have a couple of minutes between meetings. The surprising thing is how much of what I learned has stayed. I am still noticeably more geographically literate than I was 90 days ago. I still recognise country shapes I would not have recognised before. I still pay more attention to maps. The skill plateau is real, and it does not seem to require ongoing daily practice to maintain.
If I had to summarise what 30 days of one round a day actually produced, it would be this: the world became more visually distinct. The places I used to lump together — "vague Asian country," "vague African country" — now have specific visual identities. I can place a satellite image roughly correctly more often than I can place a name on a map. The skill is more visual and more spatial than I expected, and it transferred into other parts of my life in ways I would not have predicted.
Two minutes a day for thirty days. About an hour of total play time across the month. If you had told me at the start that an hour of total effort would produce a visible shift in how I look at the world, I would not have believed it. I do now. If you are curious whether the daily geography habit is worth trying, the only honest answer I have after the experiment is: yes, even more than I expected.