shape

Spatial recall bends position, angle, and scale. This test flashes a shape, then asks you to rebuild it.

The score compares outline drift, rotation, scale, and center position.

Solo or multiplayer?

Shape memory direct answer

Shape Memory Test

Shape Memory Test measures short-term spatial recall. You see a generated shape, then recreate its position, scale, and rotation.

The report compares target and submitted geometry with raster overlap plus drift measures. Shape similarity needs more than one number because position, scale, and rotation errors can look different.

This is not a motor-control, visual-spatial disorder, or clinical drawing assessment. Pointer precision, screen size, viewport scale, and touch input can change performance.

Evidence at a glance

5 shapes per standard run

64x64 raster overlap comparison grid

IoU Jaccard-style overlap basis

4 tracked errors: overlap, center, scale, rotation

Target source

Seed-generated circles, polygons, and stars with controlled position, rotation, and scale.

Measurement

Raster overlap / Jaccard-style comparison plus center, scale, and symmetry-aware rotation drift.

Report use

Outline overlap, center drift, rotation drift, scale drift, and round-by-round consistency.

Boundary

Not a clinical visual-spatial test, motor test, or drawing-skill certificate.

Evidence

What authority supports this test?

The app compares target and submitted geometry with a 64x64 raster overlap score, then applies symmetry-aware rotation and scale penalties.

The cited overlap references support Jaccard / IoU-style binary image comparison language. They do not certify the memory task.

Target fairness

How same-seed comparison stays fair

Bot Challenge is labeled as a system rival and uses deterministic comparison pressure.

Friend Challenge and Live Match use the same seed, difficulty, and round count, so both sides recreate the same target geometry.

Limits

What can affect the result?

Pointer precision, touch-screen size, motor control, viewport size, and visual attention can affect results.

Report meaning

What the shape report should explain

A useful shape report explains whether the miss came from where the shape was placed, how large it was, or how it was rotated.

The report should avoid overclaiming because pointer device and viewport size can strongly affect a geometry recreation task.

Result analysis

How to read a shape memory result

1 Use overlap first

Overlap explains how much the submitted shape shares area with the target, which is a better summary than center position alone.

2 Split the geometry

Position drift, scale drift, and rotation drift should be separated because the same total score can come from different visual mistakes.

3 Account for input limits

Pointer precision, touch screens, and viewport size can change results, so repeated same-device runs are the fairest comparison.

Sources

Measurement references

Jaccard / IoU overlap for binary image comparison ch.mathworks.com/help/images/ref/jaccard.html IoU and Jaccard index in image segmentation pmc.ncbi.nlm.nih.gov/articles/PMC5977596/

FAQ

Shape memory FAQ

Why use overlap?

Overlap captures how much the submitted shape shares area with the target, which is more useful than checking only center position.

Why include rotation and scale?

Two shapes can overlap poorly for different reasons, so the report separates rotation, scale, and center drift.

Can touch screens change scores?

Yes. Screen size, touch precision, pointer device, and viewport scale can affect shape placement.

Can this diagnose spatial ability?

No. It is a browser-based recall task, not a clinical assessment.