How Guess a Country Works

How the Daily Country Is Chosen

Every date maps to exactly one country, and the mapping is a pure function of the date. There is no random draw at the moment you load the page, no server-side state, and no editor making a selection by hand. The same process picks the answer on day one as on day ten thousand.

Concretely, the date string is passed through a small string hash to produce a number. That number is reduced modulo the length of a frozen, ordered list of countries, and the resulting position selects the answer. Hashing the date rather than the clock is the important part: it means the result cannot change between two people opening the same puzzle, and it cannot change later either.

The list is described as frozen because it is kept stable even if the playable country list is later extended or trimmed. If a published date happens to point at a country that is no longer playable, the game walks forward through that same frozen order to the next country which is. The result stays deterministic without anyone having to hand-write a schedule, and — this is the point — every answer already published keeps the country it was built with.

Why it works this way

A shared puzzle only means something if it is genuinely shared. Hashing the date gives three properties at once: every player worldwide receives the identical country, the answer is decided in advance and cannot be manipulated after the fact, and a past answer is a permanent record rather than a value that could be regenerated differently. Two players comparing notes on different days are always looking at the same underlying rule.

Because the answer is derived from the date rather than drawn at random, the sequence is as fair as a hash can be — but it is not designed to be a good sequence. Two difficult countries can land on consecutive days, and no one is steering it. The unlimited mode exists for the days when the daily draw is not to your taste.

How Puzzle Numbers Are Assigned

The first published day is puzzle #1, and each subsequent UTC day increments by one, so the numbering is simply a count of days since launch. A new puzzle appears at midnight UTC, which is the boundary used everywhere on the site — the countdown, the streak, the archive and the sitemap all key off the same instant. Players in different time zones therefore face the same puzzle for slightly different clock times, which is the intended behaviour rather than a bug.

How Distance Is Calculated

Distance is the great-circle distance between the two countries' stored reference coordinates, computed with the Haversine formula. This treats the Earth as a sphere of radius 6,371 kilometres, which is the standard model for this kind of puzzle and is accurate to well under one percent for the distances involved.

The result is rounded to the nearest whole kilometre. It is a straight-line surface distance, so it ignores elevation, terrain and any political border — it answers “how far apart are these two places” and nothing more.

How the Direction Arrow Is Calculated

The game computes the initial compass bearing from your guess towards the answer, expressed in degrees clockwise from true north. That bearing is then snapped to the nearest of eight compass points by dividing the circle into 45-degree segments and rounding to the closest one.

This is why you see N, NE, E, SE, S, SW, W or NW instead of a precise degree, and it is deliberate: eight directions are readable at a glance on a phone, which a raw bearing is not. The trade-off is that an arrow can point the same way for two countries at quite different bearings, so the arrow is a starting point for reasoning rather than a complete answer on its own. A target icon replaces the arrow when you have guessed correctly.

How the Proximity Score Is Calculated

The proximity percentage is a straight transformation of the distance, and it is worth being precise about what it is not: it is not a measure of how sensible your guess was, and it is not a running comparison against your other guesses. It answers one question only — how near the answer, in kilometres, compared to the far side of the world.

• An exact match returns 100.
• Otherwise the game treats 20,000 km as effectively zero proximity and scales linearly from there.
• The result is clamped into the 0–100 range and rounded, so it can never display a negative bar or exceed the maximum.

Because 20,000 km is treated as the floor, a guess on the far side of the planet from the answer reads as 0% rather than as a small negative number. In practice a bar that is barely moving tells you the answer is very far away, and a bar that is more than half full tells you you are in the right region. The colour bands on the guess card step through distinct levels at 35, 60 and 80 so that the bar is readable without comparing precise numbers.

How Hints Are Drawn

Hints come from the same dataset the game is played on, not from a separate fact file, so a hint can never contradict a clue. The Get Hint button reveals them in a fixed order: the capital and continent, then the population, then the first letter of the country's name together with one bordering country.

Hints are free and never consume a guess. Their real cost is informational rather than mechanical — each one you read narrows your own thinking as much as the answer's, and it is difficult to reconstruct the clean distance-only solve afterwards. The first-guess guide covers opening well enough that you may not need them.

Where the Data Comes From

The dataset contains 233 countries, each with a capital, a population, a continent, coordinates, a map outline and a list of neighbours. Outlines are real country shapes rather than placeholders, and capitals are populated for every entry, so the game never has to display a missing value in the middle of a clue.

This is a fixed snapshot that ships with the game, not a live feed. That is a deliberate choice: because the clues and the facts you are shown afterwards come from one source, they are always consistent with each other. It also means populations and borders will drift over time. Where a value is genuinely missing it is shown as not recorded rather than estimated, and figures should be read as recorded values rather than breaking news. The About page explains how corrections are handled.

How Unlimited Mode Differs

Unlimited mode is the one part of the game that is genuinely random. Instead of deriving a country from a date, it draws one on request from a pool narrowed by the difficulty you pick:

• Easy — From the 60 most populous countries.
• Medium — Any country in the dataset.
• Hard — Countries under 2 million people.

The response is explicitly marked as never cacheable, because a cached “random” country would hand every player the same puzzle and quietly break the mode's purpose. The same 6 guesses and the same clue rules apply as in the daily game; only the source of the country changes, and nothing done in unlimited mode affects your streak or statistics.

Determinism and Rendering

One implementation detail is worth stating because it explains a behaviour players sometimes notice. The server and the browser both derive the daily target through the same function rather than the client recomputing it independently. If the browser calculated the answer on its own it would work from a different list size and could arrive at a different country, which would break the page during hydration. Routing every answer through one shared function is what keeps the rendered puzzle stable.

What the Clues Cannot Tell You

Distance, bearing and proximity are enough to solve almost every puzzle by reasoning alone, and that is what makes the strategy guide work. They are not, however, a complete description of the answer. Two countries can sit at nearly the same distance and bearing from a given guess, and no clue in the system accounts for population, language, land area or political status. Where the clues are genuinely ambiguous, that is a property of the geography rather than a defect in the puzzle.