Whole-grid anomaly search
Inspect the whole field without a target preview, select its single structural exception, and review time, first-click accuracy, errors, streaks, and difference categories.
Infer the majority rule, then select its only structural exception.
Solved
0
Incorrect clicks
0
Questions
0/5
Choose a mode, then start.
Each board contains one defined exception among repeated project-owned vector symbols. The generator changes rotation, a stroke, left–right structure, mark count, or one shape detail. It does not rely on a system emoji or color alone.
Completed summaries stay in this browser. The task describes performance on these boards and settings; it is a recreational visual-attention challenge, not a vision, intelligence, or clinical test.
Unlike target-preview search, the majority of symbols establishes the rule. Scan for repeated orientation, line count, symmetry, and small marks, then test the item that breaks one of those regularities.
A first-click result records whether the first selection was the exception. Later correct selections still complete the puzzle, while every earlier selection remains in the incorrect-click count.
Rotation changes direction; missing-stroke removes part of a line; mirror reverses an asymmetric feature; count changes the number of marks; detail replaces one local shape. These are separate generator rules rather than recolored copies of one puzzle.
Practice can reveal the answer region. Standard and advanced sessions do not permanently highlight the correct cell after an error, so an incorrect click does not solve the board by elimination immediately.
Grid size adds items to inspect, but screen size, cell size, symbol complexity, difference subtlety, time limits, and hints also change the task. A 10×10 board is therefore not a calibrated extension of a 4×4 board.
Compare personal sessions only when mode, grid, hint rule, display scale, and input method are similar. Twelve-by-twelve is restricted to screens that can keep each symbol legible.
First-click accuracy is solved puzzles whose first selection was correct divided by solved puzzles. Median solve time uses the middle solved-puzzle time, reducing the influence of one unusually long search.
Category medians can show which generated difference took longer in this session, but a small category sample is descriptive only. The page does not combine those values into a visual IQ or brain score.
Display scaling, contrast, pointer precision, and physical screen size affect visibility. Keyboard users can move through the grid with arrow keys and select with Enter or Space; focus styling never discloses the answer.
A dense visual grid is not fully equivalent through a screen reader. The interface states that limit honestly, keeps controls and result summaries readable, and does not make color the only correct/error cue.
Visual Search provides a target first, while Symbol Search asks whether target shapes occur in a candidate row. Odd One Out provides no isolated target sample: the viewer infers the majority rule and finds its only exception.
Change Blindness compares two scene versions, and Pattern Matching compares two complete patterns. Those tasks use different evidence and state transitions even when they share visual details.
These questions stay tied to the visible content on this page.
Continue with tools that pair naturally with this page.
Find one target in symbol, letter, or number boards and review completion time, errors, difficulty, and local history.
Start
Scan symbol rows, decide whether a target appears, and review correct answers, errors, pace, and response consistency.
Start
Compare visual patterns, identify matches or differences, and review accuracy, response time, error types, and difficulty progression.
Start
Compare alternating scenes, find the changed detail, and review detection time, accuracy, hints used, and puzzle difficulty.
Start