Change blindness is the well-known tendency to miss large visual changes when a brief blank mask interrupts the view, and the Change Blindness Test turns that effect into a five-round browser game where you study an abstract scene for 0.9 seconds, wait through a 0.18-second blank, then locate the single altered feature on a nine-zone grid. Each round is a complete Original → Flicker → Answer cycle with its own geometry, its own declared before-and-after feature, and its own correct answer zone, and the cycle is identical from round one through round five so you always know what kind of input the game is waiting for. Five rounds, five correct clicks or number presses, and exactly 1,000 points — the math is the math, and the schedule is published up front. The whole game runs locally in your browser tab; nothing uploads, nothing streams, and the only thing the page stores is the best score you have finished with on this device.
That framing already explains why a beginner can walk in without preparation. The game does not ask you to prove attention, recall, eyesight, or reaction speed. It does not pull from a random target set, does not adapt difficulty, and does not reward fast clicks with bonus points. You see the same five scenes in the same order every run, the timers are presentation-only, and the scoring rule is a flat 200 points per correct zone. Treat it as a small observation exercise with a transparent schedule, and the 1,000-point finish becomes something you can plan around rather than guess at.
What the Change Blindness Test Is and Isn't
The product is built from five original abstract SVG scenes — rings, squares, rays, stripes, orbit dots, and bridges drawn from a single stable geometry. Their dimensions are reserved so the layout never shifts between the Original frame and the Changed frame, which is what makes a single altered feature visible instead of a recomposed scene. The fixtures are part of the game itself; they are not photographs, scientific stimuli, diagnostic material, or an imported perception test. The interface adds no upload step, no camera, no eye tracker, no account, no third-party image, and no network request.
The page makes this clear in its own FAQ pair: the game is entertainment, not a medical or eyesight test, and it claims no measurement of cognition or attention. The same is true of the round structure. Every round follows the same state machine, every round has one declared change, and every round ends in either 200 points or a counted miss. A Boss Key, activated by double-pressing Escape, swaps the screen for a fake spreadsheet and back, but the underlying trial is not changed, and the answer zones remain exactly the same nine regions counted left to right and top to bottom.
The Round-by-Round State Machine
Each round passes through three phases in this exact order: Original, Flicker, Answer. The Original frame shows the abstract scene exactly as it was drawn and stays on screen for 900 milliseconds, or 0.9 seconds. A neutral blank mask then covers the same canvas area for 180 milliseconds, or 0.18 seconds, with no scene content visible. After the mask, the Changed frame appears with the same overall geometry and exactly one altered feature, and the nine answer zones become interactive.
Two shortcuts act on this state machine without breaking it. The "Show change now" button and the S key end the Original phase early, but they preserve the full 0.18-second mask before the Changed frame is shown. There is no path that lets you skip directly from Original to Answer. Invalid timing, answering during Original or Flicker, and out-of-range zone numbers are all harmless no-ops; the game simply waits for a legal input during the Answer phase. Once the Changed frame is visible, each third of the scene becomes one large semantic button numbered 1 through 9 from left to right and top to bottom, so a beginner can fall back on the number keys instead of trying to land a precise pointer click on a small target.
How to Play the Change Blindness Test
- Open the Change Blindness Test and let the first Original frame appear; study the abstract scene for the full 0.9 seconds, or press S to end the Original phase early.
- Wait through the 0.18-second blank mask without clicking anywhere — the mask is part of the round and the answer layer does not mount until it finishes.
- When the Changed frame and the nine zones appear, click the third of the canvas that contains the altered feature, or press the matching number from 1 through 9 on the keyboard.
- Read the live feedback: a correct zone adds 200 points and loads the next Original frame; a wrong zone is recorded but lets you keep trying until you find the right one in the same round.
- Repeat for all five rounds, and on the fifth correct zone the run closes with exactly 1,000 points and a final score summary.
Pointer users click directly on the altered region. Keyboard users press the number that corresponds to its third of the scene, counted left to right and top to bottom: zones 1, 2, and 3 form the top row; 4, 5, and 6 the middle row; 7, 8, and 9 the bottom row. Touch users on a phone tap the same nine regions. Every region is sized larger than a minimum touch target, focus is visible, and reduced-motion users do not depend on decorative movement to play.
What Changes in Each of the Five Rounds
The five scenes are fixed product fixtures, and the change in each one is declared up front rather than rolled randomly. The table below lists the literal feature, the round's before-and-after transformation, and where on the three-by-three grid the change sits so you know what to scan for on your first run.
| Round | Scene feature | Before → After | Where to look |
|---|---|---|---|
| 1 | Central ring | Ring becomes a square | Centre third of the canvas |
| 2 | Northwestern rays | Ray cluster is reduced | Top-left third of the canvas |
| 3 | Southeastern stripes | Stripe set is rotated | Bottom-right third of the canvas |
| 4 | Northeastern orbit | A single dot moves along its orbit | Top-right third of the canvas |
| 5 | Southwestern bridge | A gap opens in the bridge | Bottom-left third of the canvas |
Round 1 always lands in the centre third because the ring sits in the middle of the canvas, and rounds 2 through 5 each pull toward one named quadrant so the change stays visible without crowding the border. The independent logic test that ships with the page restates all five literal before states, after states, and correct zones rather than computing an answer from the production array, which is why the schedule behaves the same way on every restart.
Scoring, Misses, and the 1,000-Point Run
Correctness and scoring are deliberately simple. One correct zone adds exactly 200 points and advances to the next Original frame. Five correct zones therefore equal exactly 1,000 points, with no speed bonus, no streak multiplier, no hidden adjustment, and no random target. The short timers control how long each phase is shown; they do not change how many points a correct answer earns, so there is no reason to rush an answer at the cost of accuracy.
Mistakes are tracked by round and by zone. Picking a wrong zone records one miss, and picking the same wrong zone again in the same round is deduplicated and does not cost you the second error. Picking a different wrong zone registers the second distinct miss and closes the run in a terminal deadlock. After deadlock the game freezes on the final state, restart restores trial one, the score to zero, the error budget to two, and the Original timer to the full 900 milliseconds. The independent logic test repeats one wrong zone, triggers a second distinct miss, and verifies both the terminal freeze and the deterministic restart, which is why every run behaves the same way.
Beginner Habits That Make the Change Easier to Catch
The Original frame only stays for 0.9 seconds, so a quick mental checklist is more useful than trying to memorize the whole scene. Anchor your eyes on the centre first — round 1 always changes the central ring — then sweep the four named quadrants in the order the rounds introduce them: northwestern rays, southeastern stripes, northeastern orbit, southwestern bridge. By round four you already know which quadrant to look at before the mask finishes, which is most of the work.
Use the S key strategically rather than reflexively. Pressing S during the first 100–200 milliseconds of the Original frame ends the Original phase at that point, so you only see that small slice of the original scene, and the 0.18-second mask still runs before the Changed frame. If you press S too late you trade the flicker for nothing, and if you press S after the mask you will be answering into a no-op rather than advancing. If your first guess is wrong, do not panic; one wrong zone is recoverable, and a repeat of the same wrong zone does not consume the second error, so a beginner has real room to confirm by elimination.
Other Visual Challenges Worth Trying
Once you have cleared five rounds of change detection, other browser games on the site test nearby skills in a comparable format. Spot the Difference is another fixed visual challenge on the site. Pattern Memory Game is another fixed visual challenge on the site. Hidden Object Game is another fixed visual challenge on the site., while Visual Memory Test is another fixed visual challenge on the site.. Each of these games keeps its fixtures and scoring visible the same way the Change Blindness Test does, which makes them a natural next step once change detection feels easy.
You can return to the Change Blindness Test any time; the best completed score is the only thing the page stores in this browser, and restart always sends you back to trial one with the full 900-millisecond Original timer and a fresh pair of recoverable misses.