No—target discs do not stay visually marked during motion in the Multiple Object Tracking Test. The two accent rings appear only during Preview and are removed when Tracking begins. Once motion starts, all six discs have the same visible fill, stroke, size, shape, styling, and accessible description, so the player has to preserve each target's identity from its remembered position and path rather than following a persistent color or label.
Why the markers disappear at motion start
Preview is the only phase that identifies the targets with accent rings. When motion begins, the target-ring elements are unmounted rather than simply hidden behind another layer. No target-specific label, data attribute, special class, color, or number follows either disc. The moving field therefore contains six visually equivalent circles, and the task depends on remembering which two were highlighted before movement began.
Each motion frame is a complete rearrangement of the six disc identities. That guarantees exactly one disc in every location and prevents overlapping or missing objects. During Answer, the moving SVG is replaced by six numbered position buttons. Those numbers identify locations for the response; they do not reveal which disc originally occupied each location.
| Phase | What is visibly provided | What the player must do |
|---|---|---|
| Preview | Six discs on a two-by-three field, with accent rings on exactly two targets. | Memorize the two targets, then start motion. |
| Tracking | All six discs use the same visible styling, and the target rings are gone. | Follow both original targets through the fixed frames. |
| Answer | Six numbered position buttons replace the moving SVG. | Choose the two final positions occupied by the original targets. |
What still helps you track the two discs
The remaining information is positional continuity rather than a persistent visual marker. Each round contains either three or four fixed position frames, and every frame contains a complete permutation of the six identities. Because every location is occupied, there is no empty slot, overlap, or uniquely styled object that can substitute for the missing rings.
A practical manual approach is to anchor each target to its previous position, note where it moves in the next frame, and carry that identity forward. Tracking one target before confirming the other may also make the field easier to monitor. These are player strategies rather than assistance supplied by the game: during Tracking, the interface provides no target-specific visual cue that follows either disc.
Motion advances automatically every 650 milliseconds with a visible CSS transition. You can let the declared frames play, or choose Finish motion now. The shortcut lands directly on the same final frame and uses the same validator as the automatic route, so it reduces the wait without changing the endpoints or scoring rules.
How to play all five tracking rounds
- Memorize the targets during Preview. Identify the two ringed discs among the six and hold their starting positions in memory. Do not expect either ring to accompany the disc once motion starts.
- Start the motion. Press Space or choose Start motion. The rings are removed as Tracking begins, leaving six identically styled discs.
- Follow both original targets. Track each target from one fixed position frame to the next, using its previous location and movement to maintain its identity. You can wait for the sequence to finish or press Space during Tracking to use Finish motion now.
- Review the final frame. The Answer phase replaces the moving discs with six numbered position buttons. Determine which two locations are occupied by the discs that had the rings in Preview.
- Select exactly two positions. Press keys 1 through 6 to toggle positions, or select the numbered buttons with a pointer or touch. Your selection is an unordered pair, so the order in which you choose the two positions does not matter. Choosing an active position deselects it, an inactive position adds it until two are selected, and an attempted third selection does nothing.
- Submit the pair. Press Enter or choose Submit. The control stays disabled until exactly two positions are selected. A correct pair returns to Preview for the next target set.
Worked scoring example: 5 correct rounds × 200 points per round = 1,000 points.
When a miss is recorded
Changing a selection is free and never counts as an error. Only submitting a complete pair that differs from the correct answer records a miss, so you can reconsider either position before committing. The submitted pair is checked as an unordered set; selecting the right locations in the opposite order still counts as correct.
One submitted wrong pair is recoverable. Submitting the identical unordered pair again is recognized as a repeat and does not consume another error. To record a second distinct miss and close the run, change at least one selected position and submit another wrong pair. A correct pair awards 200 points and sends the game back to Preview for the next target set.
For a focused explanation of the difference between editing a response and answering it, see Do Selection Changes Count as Errors? Tracking Test.
There is no speed bonus, random path, adaptive motion, click penalty, or hidden multiplier. The ledger reports the current frame count, selected count, score, and misses, so the result reflects the game's exact answer-checking rules rather than reaction speed.
Controls for keyboard, pointer, and touch
| Action | Keyboard control | Pointer or touch control |
|---|---|---|
| Start from Preview | Space | Start motion |
| Finish during Tracking | Space | Finish motion now |
| Toggle final positions | Keys 1 through 6 | Select the numbered position buttons |
| Submit a complete pair | Enter | Submit |
Keyboard play covers the complete game, while pointer and touch users receive equivalent start, finish, position, submit, and restart controls. The controls use large targets and show visible selected states. Reduced-motion styling removes the animation while preserving each state, so the tracking logic does not depend on the CSS transition being visible.
Everything runs locally. There is no camera, eye tracker, upload, account, API, model, external asset, random service, or personal-data requirement. The only information that can remain in the browser is a shared best completed score. Restarting restores the first target preview, frame zero, zero score, and no selected positions or errors.
What the final score represents
Despite the word “test” in its title, this product is a deterministic entertainment game rather than a scientific, medical, vision, attention, or diagnostic assessment. Its paths, target pairs, frame intervals, scoring, and error rules are original fixed fixtures. No score is compared with population norms, and no result is presented as a measurement of a cognitive trait. A 1,000-point route means that five submitted pairs were correct under the game's rules; it does not establish an ability score.
The timer route and Finish motion now route end at the same declared positions and use the same validator, which keeps the challenge reproducible. Selection freedom also means the score reflects the final submitted pair rather than every position toggled along the way. The supported result is therefore a transparent game score, not a clinical interpretation.
In short, target marking is a setup cue rather than a tracking aid: the rings vanish at motion start and do not return. The supported workflow is to memorize the two previewed targets, follow their identities through the fixed frames, and submit the two final locations occupied by those original discs.
For a deeper look, see Are Shadow Matching Game Objects From External Images?.