Home/Puzzle Games/The Long Road

The Long Road

Twelve strangers, twelve problems, and twelve decisions the test cannot score

Works on phones and tablets · Free · No download · No sign-up

About The Long Road

The Long Road is an IQ test you walk. The Index scores every citizen and decides where they live, what they do and how far they may go. Yours is under appeal, the hearing is at dawn, and the road there is long and badly lit.

You hold a button and move down the street. Every so often somebody steps out of the dark and stops you. They have been scored badly too, and they have a problem — offered to you three ways, gentle, straight or steep. Solving it earns standing, and in this city standing is literally how far you may go: the road ahead runs exactly as far as you have earned. Then they ask you for some of it.

That second half is the part no test usually has. Nothing there is scored right or wrong, because there is no right answer to what you owe a stranger, and the game does not invent one. It records what you did — applied the same rule to everyone, gave away more than you could spare, or kept what you were going to need — and reports it back at dawn next to the number, without a verdict attached. Give everything away and your own appeal fails. That is the point rather than a bug.

Underneath the walk it is a real adaptive test, the same engine as Mindforge. Every item carries a difficulty from 1 to 10 and the next one is chosen at the level you are currently holding, so the number of right answers barely matters. Four right at level 8 beats seven right at level 3, and the game scores it that way. Twelve encounters, two problems each, twenty-four items in about twelve minutes.

Letting you choose the difficulty is the part that had to be got right. A short test plus items much harder than you plus a one-in-five guess flatters the estimate — so choosing the hard option every time would have been worth free points. That was measured across 20,000 simulated players and corrected, and the correction is stated on the results screen. Taking the steep road buys you standing. It does not buy you a score.

Difficulty is not only in the puzzle — it is in the wrong answers. An easy item offers options that are plainly off. A hard one offers options that are each a single minimal step from the right one: a count off by exactly one, a shape swapped for one that looks like it, a piece with one cell moved one square, a mirror image rather than a rotation. The clock scales with the item too, so a hard one can be worked out properly instead of rushed.

The problems are the ones real cognitive tests use, generated fresh every time rather than drawn from a bank. Progressive matrices where several attributes change at once, and where from the middle of the scale a rule may run down the columns rather than across the rows. Logic grids where the third pattern in each row is the first two combined by an operation you have to find. Odd-one-out figures, visual analogies, number series built from arithmetic, geometric, interleaved and recursive rules, balance scales, letter ciphers, mental rotation where the tempting answer is always a mirror, paper folding, shape fitting, syllogisms and conditional chains written with nonsense names, and knights-and-knaves puzzles where one group always lies.

The twelve are drawn in 3D and lit by the street they are standing in — they breathe, shift their weight and turn to look at you while you decide. Where a browser has no WebGL the whole road falls back to a painted version and the game plays exactly the same.

What you get at the end is your appeal outcome, the people who walked in with you, and an IQ-scale score with an honest account of where it came from. The percentile sits on a curve this game defines by simulation, and the game says so rather than implying a population rank. It is a good game that takes your reasoning seriously. It is not a clinical assessment, and a browser tab is not where anyone gets diagnosed with anything.

How to play The Long Road

Hold the walk button — or hold space, W or the up arrow — to move. Let go and you stop. You find your stride after a couple of seconds and the road starts to move properly.

Someone steps out of the fog roughly every hundred metres. Read what they say, then pick one of the three things they offer. The label tells you the kind of puzzle, the badge tells you how hard relative to what you are currently managing, and the number on the right tells you what it pays.

Answer by tapping an option or pressing its number key. There are two problems per stranger and the second is a step harder than the first. Getting either one right is enough for them to come with you.

If you miss one, Look at it again on the verdict card brings the puzzle back with the right answer marked. It runs after the item is scored, so looking changes nothing.

Then they tell you what they actually need, and you decide. Every option is spelled out in standing so you can see exactly what it costs you before you choose.

Walk on. The road ends after the twelfth, at the hearing. Then you get the score, the profile, what you did with what you earned, and how it went for all thirteen of you.

About this game

Genre
Puzzle Games
Released
22 September 2026
Developer
DevAI Games
Platforms
Desktop and mobile browsers
Price
Free — no sign-up
Last updated
22 September 2026

More puzzle games

Recently added

The Long Road rankings

Ranked by best run. Scores are reported by the browser that played the run, like every board on this site — it is a bit of fun rather than a competitive ladder. Today resets at midnight IST, the week resets on Monday.

Loading…

The Long Road — frequently asked questions

What do I get at the end?

Two things, kept apart on purpose. An IQ-scale score on the familiar mean-100 scale with a four-spoke profile of pattern logic, number sense, spatial sight and deduction — that is what the Index in the story measures, and it comes only from the puzzles. And then what it could not measure: what you did when twelve people asked you for something, how much standing you gave away, and how your own appeal went as a result.

What is the story?

The Index scores every citizen and decides where they live, what they do and how far they may go. Yours is under appeal and the hearing is at dawn. Everyone on the road has been scored badly too. Solving their problem earns standing, and standing is how far you may walk — then they ask you for some of it, and every yes shortens your own road.

Are the decisions scored?

No, and the game says so on the screen where you make them. There is no correct answer to what you owe a stranger, so the game does not invent one. It counts what you did — by the book, gave it away, kept your own — and reports it back without a verdict attached. Only the puzzles feed the IQ number.

How is this different from Mindforge?

Same engine, different shape. Mindforge is the sit-down test: six sections, twenty-eight adaptive items plus memory and speed, and a live ten-player race. The Long Road is twenty-four items across four sections wrapped in a walk, where you choose the difficulty of each one and then decide what to do with what you earned. About twelve minutes instead of eighteen, and because it asks fewer questions its number is looser. Both put a middling player at 100.

Does taking the steep option get me a higher score?

No, and that was measured rather than assumed. Harder items plus a one-in-five guess would otherwise flatter a short test, so the estimator subtracts that effect. Across 20,000 simulated players, always choosing steep and always choosing gentle land within one point of each other. Steep buys you standing, not score.

What does the percentile mean?

It places your score on a reference curve this game defines — the spread of scores it gives 20,000 simulated players — not on a normed sample of real people. The results screen says exactly that rather than dressing a modelled number up as a population rank.

Can I memorise the puzzles?

There is no bank to memorise. Every puzzle is generated at the moment you see it, from pools measured at hundreds to thousands of distinct items at every difficulty. The game also remembers the last 500 puzzles it has shown you — across this game and Mindforge both — and will not repeat one. The twelve people and what they ask are written, and their order is shuffled every walk except the last.

What happens if I get one wrong?

You earn less standing, and the road ahead is shorter. There is no failure state and no restart: you walk on either way. Getting one of a stranger's two problems right is enough for them to come with you.

How long does a walk take?

About twelve minutes. Each problem has a clock that scales with how hard it is — roughly 30 seconds on an easy one up to 90 on the hardest — so a hard one can be worked out on paper rather than rushed. Running the clock out counts as wrong.

Is the IQ figure a real IQ?

It is your result placed on the scale people recognise — mean 100, spread 15 — and nothing more. On simulated play its typical error is about eight points either way, which is what a twenty-four-item test looks like. A clinical assessment takes an hour or more, is given in person, and is interpreted by somebody qualified to do it. This will put you roughly in the right place; it will not settle an argument, and it diagnoses nothing.

Does it work on a phone?

Yes. Hold anywhere on the walk button to move, tap to answer. The street and the people are drawn in 3D where the browser has WebGL and fall back to a painted road where it does not, so the game plays either way.

Who are the twelve?

A night-bus conductor whose route has been called redundant, a twelve-year-old who was given a number at school today, a market trader whose licence was not renewed, a locksmith rated unskilled, a woman reassigned out of the city, a tiler whose hours were cut, a mechanic whose workshop is surplus, a security guard nine hours into a shift, a man whose residency is under review, a busker whose pitch was revoked, someone whose transport allocation was withdrawn, and the last human auditor, who is always last.

My score moved a lot between walks. Is it broken?

No — that is what a twelve-minute measurement looks like, and this one is shorter than the sit-down test. Sleep, mood, screen size and plain luck all move it. Two or three walks on different days tell you far more than any single number does.

Tell us what to fix

Found a bug, or think something should play differently? This goes straight to the developer — no account, no email address required.