Common pixel judgment mistakes.
Similar-looking rectangles can miss for different reasons. Use these illustrative cases to distinguish scale, proportion, and direction before choosing a correction. They describe possible patterns, not measured habits of Pixactly players.
The same area can hide two wrong dimensions
A 180 × 120px target and a 240 × 90px drawing both cover 21,600 square pixels. Yet the drawing is 60px too wide and 30px too short: Pixactly scores it as 90px off. Judging the filled area alone misses both errors.
Solid: target 180 × 120Dashed: drawing 240 × 90
Before drawing, name width and height separately. After scoring, ask which edge extends beyond the target rather than whether the filled space looks approximately equal.
Correct proportion does not guarantee correct size
| Target | Drawing | What happened | Total off |
|---|---|---|---|
| 180 × 120px | 198 × 132px | Both sides 10% larger; ratio stays 3:2 | 30px |
| 180 × 120px | 198 × 108px | Width +10%; height −10%; ratio changes | 30px |
The first drawing preserves the target's proportion. Rechecking the ratio alone will not catch it; use a smaller size anchor for the whole shape. The second needs a shorter width and a taller height. Shrinking the whole rectangle would make its height error worse. The equal scores conceal two different corrections.
A large average error does not tell you the direction
Consider five illustrative signed width misses: +12, −12, +12, −12, and 0px. Their signed mean is zero, but their mean absolute error is 9.6px, displayed as 10px. The eye has not matched every target; the directions have cancelled in the signed average.
If your round notes alternate in this way, avoid deciding to make every rectangle narrower. Try choosing an endpoint before dragging and check whether the spread of misses becomes smaller. If notes repeatedly say “too wide,” a modest width reduction is a more relevant experiment. Record the direction during feedback because the final result reports absolute averages only.
A changed setup can look like a changed habit
Check your notes for a different device, browser zoom, window size, or input method before comparing sets. For example, a mouse set at one window size and a touch set on a phone involve different drawing conditions. Keep separate baselines instead of treating a larger score as evidence that the previous correction failed.
If a result surprises you, repeat the same process in the same setup. Watch whether a late pointer movement changes the edge you intended to place. The score measures the submitted rectangle, not the size you initially pictured.
Choose the next experiment from the observed miss
- Same ratio, both sides too large: test a smaller overall anchor.
- Too wide and too short: choose one axis to correct first and monitor the other.
- Alternating directions: test a more consistent endpoint or drawing pace.
- Only one unusually poor round: collect another set before assuming a repeated pattern.
The exercise guide turns those choices into a recording task. Keep the observation separate from your explanation: “four widths overshot” is visible evidence; “my horizontal reference is too large” is a hypothesis to test.