Target Practice

What you see

One trial is one wave. You press Start, and objects begin to enter from the left or right edge of a rectangular field, glide across in a straight line and leave by the opposite edge. Nothing stops, bounces or changes course. The Time counter above the field shows how many seconds the wave has left, and the line under the field counts the targets you have hit so far.

There are two kinds of object:

• Targets are blue circles with a solid inner ring. These are the only things you should click.
• Decoys are grey and change shape as levels rise. At levels 1 to 3 a decoy is a rounded square with a small solid square inside. At levels 4 to 6 it becomes a circle with a solid dot in the middle. From level 7 it is a circle with a dashed inner ring, which differs from a target only in the ring and the colour.

You aim with a mouse or finger and fire by clicking or touching the field. On a keyboard, the arrow keys move a crosshair four field units per press and Space fires. When a click hits a target, the target flashes and closes up. A click on a decoy leaves a brief red cross; a click on empty field leaves a grey one. Targets you never hit simply glide off the far edge.

How the wave is built

Each wave mixes solo travellers with crossing pairs. A solo enters at a random height and drifts across on a shallow diagonal. A crossing pair is one target and one decoy that spawn at the same instant on two parallel horizontal lines a fixed gap apart. In some pairs both travel the same way, side by side. In others they enter from opposite edges and pass each other in the middle of the field. The share of these converging pairs rises from 30% at level 2 to 75% at level 10. Pairs are the heart of the task: they force you to discriminate at the very moment two similar objects are close together.

A concurrency cap limits how many objects can be on the field at once, so the wave comes in a steady stream rather than a single burst. The whole wave lasts about 17 seconds at level 1, and around 8 to 9 seconds from level 5 upwards, because faster objects clear the field sooner.

What changes with level

Difficulty runs from level 1 to 10. The field is 100 units wide and 75 tall; speeds and sizes below are in those units.

Time to cross is for a straight horizontal pass; diagonal solos take slightly longer. Read down the table and four pressures stack up together: objects move more than three times faster, targets shrink to under 60% of their starting radius, decoys look more like targets, and the gap inside each crossing pair halves.

What it measures

The game is scored for selective attention, with a secondary weight on processing speed, and it sits in the Selective attention facet of the Attention dimension alongside Singularity. It is not a pure test of any single lab task. It is a composite of three well-studied abilities, and knowing which one is under strain at a given moment tells you what to fix.

Finding the target among lookalikes. Visual search research separates two situations. When a target differs from everything else by one simple feature, it tends to pop out, and adding more distractors barely slows you. When it can only be told apart by a combination of features, you have to check items one at a time, and every extra distractor costs time and accuracy. This distinction is the core of Treisman and Gelade's feature integration theory. Duncan and Humphreys later showed that search gets harder the more the non-targets resemble the target. The decoy ladder above is that principle turned into a difficulty dial: a square decoy differs in shape, mark and colour, while a dashed-ring circle differs only in ring style and colour.

Keeping hold of several moving things. Pylyshyn and Storm's classic tracking studies found that people could keep track of up to 5 objects moving among 10 identical ones, and did so far better than a one-at-a-time scanning account predicts. Up to six objects share the field at the top levels, around the number their participants could follow.

Aiming at a moving object. Fitts's law describes how pointing time grows as a target gets farther away and smaller. Studies of aiming at moving targets report that faster targets and narrower targets both hurt performance, and that people who move faster make more systematic errors. That is why shrinking radius and rising speed are separate difficulty axes here.

Running through all three is a restraint element. Every decoy is a lure you must not respond to, which is the logic of the go/no-go task: respond to one class of stimulus, withhold for another. pymetrics' own patent describes Go/No-Go as a test of attention and response control.

How it is scored here

Every click is recorded with its time and position on the field, and when the wave ends every click is checked against the objects' exact paths. Each click is judged like this:

1. The game finds every object on the field at that instant that has not already been taken, and measures the distance from your click to its centre.
2. An object counts as touched if your click lands within its radius plus 2.5 units. So the live zone is a little larger than the drawn circle.
3. If more than one object qualifies, the nearest centre wins.
4. If that object is a target, it is taken and counts as a hit. If it is a decoy, the click counts as a click on a decoy. If nothing qualifies, it counts as a click on empty space.

The wave score is then:

score = hits ÷ targets − 0.15 × clicks on decoys − 0.05 × clicks on empty space, kept between 0 and 1.

A wave is perfect when you hit every target and clicked no decoys. Clicks on empty space do not spoil a perfect wave; they only cost you when something else went wrong. Speed does not earn anything directly. A hit two seconds after a target appears counts the same as a hit after two hundred milliseconds, as long as it lands before the target leaves the field. Speed matters only because slow aiming lets targets escape.

Difficulty is adaptive. After a perfect wave you move up one level; after a wave that scores something but is not perfect you stay where you are; a wave that scores zero drops you a level. Your 0 to 100 score for the game compares your wave scores with what the level you work at predicts, and 50 means you are doing what that level predicts. Because harder levels predict lower raw scores, a 70% wave at level 8 counts for more than a 70% wave at level 2.

After a session, the report shows the share of moving targets you caught, the share of your clicks that went to decoys, the share that hit nothing, and how many targets got away.

Where you will meet it

No vendor publishes a game identical to this one, but the same family of demands turns up in several real assessments.

• Pilot selection in India. The Computerised Pilot Selection System, developed by DRDO's Defence Institute of Psychological Research and Aeronautical Development Establishment for Indian Air Force flying-branch screening, evaluates psychomotor skills, speed and accuracy of information processing, and coordination. Candidates face multitasking psychomotor tests rather than this exact wave, but the blend of tracking, aiming and selective response is the same family.
• Motion anticipation tests. Schuhfried's Vienna Test System includes the Time/Movement Anticipation test (ZBA), which assesses the ability to estimate the speed and direction of moving objects, and lists personnel selection in manufacturing and safety among its uses.
• Game-based assessments built on attention tasks. pymetrics' patent lists Go/No-Go and Flanker tasks among the tasks its system can use. The decoy-restraint part of Target Practice is a close cousin of that go/no-go logic, and the search part is practised more directly in Singularity.

The method

1. Learn the one feature that matters at your level

   Before you press Start, say to yourself what separates target from decoy at this level. Levels 1 to 3: round versus square. Levels 4 to 6: ring versus dot. Levels 7 to 10: solid ring versus dashed ring. Checking one known feature is far faster than asking "does this look right?" of every object.

2. Rest your eyes near the centre, not on an edge

   Objects enter from both sides. Watching the middle third of the field lets you see arrivals from either edge in peripheral vision, and most crossing pairs pass each other near the middle.

3. Take the object closest to its exit first

   When two targets are on the field, the one nearer its exit edge is the one about to escape. The newcomer has the whole width still to travel. Serving the most urgent target first saves more hits than serving the nearest one.

4. On pairs, confirm the ring before you fire, then aim at the far side

   In a crossing pair, find which of the two has the solid ring, then aim at its centre or slightly on the side away from its twin. The nearest centre wins, so a click drifting toward the gap can be awarded to the decoy.

5. Lead fast targets on touch screens

   Your finger lands a fraction of a second after you decide where to put it, and the object keeps moving. At high speed, aim a little ahead along its path.

6. When unsure, hold fire

   If you cannot confirm the feature before the object leaves, let it go. A missed target costs at most one target's worth; a click on a decoy costs more. A click on empty field is cheap, so a quick second attempt at a target you just missed is fine.

Worked example 1: the crowded pair

You are at level 10. A target and its decoy twin are crossing the middle of the field, their centres 8 units apart vertically. Click A lands 3 units above the target centre, on the side away from the decoy. Click B lands 3 units below the target centre, toward the decoy. Click C lands 5 units below the target centre, so 3 units above the decoy centre. How is each click scored?

Show the solution

• Click A is a hit. It is 3 units from the target and 11 units from the decoy. Only the target qualifies.
• Click B is a hit, narrowly. It sits 3 units from the target and 5 units from the decoy. Both are inside their 6.5-unit zones, so the nearest centre wins, and the target is nearer.
• Click C goes to the decoy. It is 5 units from the target and 3 units from the decoy. Both qualify and the decoy is nearer, so you lose 0.15.

The dividing line is the midline, 4 units from each centre. Anything on the target's side of it goes to the target; anything past it goes to the decoy. That is why the method says to aim at the centre or the far side: a click that drifts only 1 unit past the midline turns a hit into a penalty.

Worked example 2: click the doubtful one or let it go?

You are at level 4, so the wave has 8 targets. Late in the wave a crossing pair converges fast, you can see one solid ring and one circle with a dot, but you are not sure which is which. In version one you click the wrong object, then panic and throw two more clicks at empty field, and finish with 7 hits. In version two you let the pair go, keep your aim clean elsewhere, and finish with 6 hits and no stray clicks. Score both. Then work out how sure you need to be before a doubtful click is worth taking at this level.

Show the solution

Version one: 7 ÷ 8 − 0.15 × 1 − 0.05 × 2 = 0.875 − 0.15 − 0.10 = 0.625.

Version two: 6 ÷ 8 = 0.750.

Holding fire wins by 0.125, even with one fewer hit.

The break-even point. Suppose you think an object is a target with probability p. Clicking gains 1 ÷ 8 = 0.125 if you are right and loses 0.15 if you are wrong. The click is worth taking only when 0.125 × p is greater than 0.15 × (1 − p), which works out at p above about 0.55. With 10 targets at level 10, each hit is worth only 0.10, and you need to be above 0.6 sure. So a genuine coin-flip is always a pass. Spend the time confirming the ring instead.

Tips and traps

Levelling up

Moving up requires a perfect wave: every target taken and no decoy touched. That makes the climb as much about restraint as about speed. Around level 4 the decoys lose their square shape, and around level 7 they lose their solid centre. Each of these steps catches out candidates who were relying on a quick overall impression instead of a specific feature. Expect a plateau at each step while you retrain the one check that matters.

At the top levels the problems change character. Targets cross in about 1.2 seconds, up to six objects share the field, and five of the ten targets arrive inside tight pairs. Here the winning habits are a steady central gaze, strict urgency ordering, and single, deliberate clicks. If you stall at levels 7 to 8, check whether you are still reacting to movement first and checking the ring second.

Be realistic about transfer. A large meta-analysis by Bediou and colleagues found that action video game training improved some attention and perception measures by about a third of a standard deviation in intervention studies, with the caveat that publication bias probably inflates published effects. Practice here will make you better at this kind of task. It is not a general attention upgrade, so treat it as preparation for the format.

1. Week 1: accuracy first
   • Three sessions of 10 to 12 minutes at your natural level
   • Count your clicks on decoys after each wave and aim for zero before worrying about hits
   • Say the distinguishing feature aloud before each wave
2. Week 2: urgency and pairs
   • Three sessions; in each wave, consciously take the object nearest its exit first
   • On every crossing pair, find the solid ring before moving your aim
   • Alternate with one short Singularity session to sharpen pure search
3. Week 3: speed without spray
   • Three sessions at your highest stable level
   • Practise leading targets if you use touch
   • Review the report: if more than one click in ten hits nothing, slow your trigger finger slightly

Check yourself

5 practice questions

1. You are at level 5 (eight targets). You hit all eight targets, click no decoys, and make three clicks that hit nothing. What happens?
2. At level 10, which is more costly: letting one target escape, or clicking one decoy?
3. A click lands 5 units from a target centre and 3 units from a decoy centre, and both hit zones reach 6.5 units. How is it scored?

and 2 more

Create a free account to answerSign in

Sources

1. Treisman, A. M., & Gelade, G. (1980). A feature-integration theory of attention. Cognitive Psychology, 12(1), 97–136. doi:10.1016/0010-0285(80)90005-5
2. Duncan, J., & Humphreys, G. W. (1989). Visual search and stimulus similarity. Psychological Review, 96(3), 433–458. doi:10.1037/0033-295X.96.3.433
3. Pylyshyn, Z. W., & Storm, R. W. (1988). Tracking multiple independent targets: Evidence for a parallel tracking mechanism. Spatial Vision, 3(3), 179–197. doi:10.1163/156856888X00122
4. Fitts, P. M. (1954). The information capacity of the human motor system in controlling the amplitude of movement. Journal of Experimental Psychology, 47(6), 381–391. doi:10.1037/h0055392
5. Chiu, T. T., Lin, C. L., Young, K. Y., et al. (2011). A study of Fitts' law on goal-directed aiming task with moving targets. Perceptual and Motor Skills, 113(1), 339–352. PubMed 21987931
6. Bediou, B., Adams, D. M., Mayer, R. E., Tipton, E., Green, C. S., & Bavelier, D. (2018). Meta-analysis of action video game impact on perceptual, attentional, and cognitive skills. Psychological Bulletin, 144(1), 77–110. doi:10.1037/bul0000130
7. Polli, F., & Yoo, J. (Pymetrics Inc.). Systems and methods for data-driven identification of talent. US Patent 9,842,314. Google Patents
8. Defence Research and Development Organisation. Computerized Pilot Selection System (CPSS). drdo.gov.in
9. Schuhfried. ZBA: Time/Movement Anticipation. Vienna Test System Marketplace. marketplace.schuhfried.com

Put it into practice: Target PracticeSign in to try it