A Four-Week Perceptle Training Plan

The four weeks at a glance

Ten to fifteen minutes a session, four sessions a week. The pass mark is deliberately the same shape every week: seven of ten rounds at or above the target grade. Anything less and the week repeats.

Week 1: Count and Track

Both Perceptle events ask for a single visual estimate with an answer you can see afterwards, which makes week 1 the easiest place to build the habit of reading the error line instead of the score.

Split each session in half: ten Count rounds, then ten Track rounds, and log the signed miss for both. By the fourth session the log should show a direction in at least one of them. If neither has a direction, the problem is attention rather than technique, and the fix is shorter sessions rather than more rounds.

The target is a B in both events, which on Count is a miss of 17% or less and on Track one of 51 pixels or less. A Perceptle Count score of 70 allows roughly one dot in six to be wrong, so hitting the target in week 1 is a realistic bar rather than a stretch.

Week 2: Angle and Split

Angle and Split both depend on a small set of anchors, which is why they share a week. Angle has its landmarks at 0, 45, 90 and 180 degrees; Split has half and a quarter of the line. Both events go wrong in the same way when the anchor is skipped: the estimate drifts towards the middle of the range.

Drill the anchors first and the targets second. Ten rounds of landmark placement with no scoring, then ten scored rounds. The pass mark is a B in both, which is 6.8 degrees on Angle and 4.2% of the line on Split.

Week 2 is also where the two Perceptle events diverge. Angle errors carry a direction and can be corrected with a constant offset; Split errors usually do not, so the only route is a better anchor. Expect Angle to move first.

Week 3: Tempo

Perceptle Tempo gets a week alone because it is the only event with no visual reference, and because its two failure modes need different drills. Drift is a clock that runs fast or slow and can be corrected. Jitter is an unsteady hand and cannot.

Play ten rounds per session and write down drift and jitter separately. If drift holds its sign across sessions, start every run deliberately late or early to cancel it. If jitter dominates, subdivide the beat and count out loud instead.

The target is an A, which is a combined error of 34 milliseconds or less. It is the highest bar in the plan and the one most likely to need a fifth week, because the correction depends on a clock you cannot see.

Week 4: full rounds only

The last week drops single-event practice entirely. Four full rounds per session, three sessions, and a target total of 80 or better in three of the four rounds.

The point of the week is not fitness. It is to find out which event collapses when the five run back to back, because that event is where the next four weeks should start. A Perceptle full round is the only place that question gets an answer.

If the total is lower than the average of your week 1 to 3 scores, the gap is the cost of switching tasks, and it is worth naming rather than training away. A round that scores 78 with a weak Tempo is a different result from a round that scores 78 with everything at 78.

How to tell whether a Perceptle week worked

Compare the median of ten rounds, not the best round. One good run is what a random question looks like when it suits you, and the median is the only figure that survives a different set of questions.

A week counts as worked when the median moves by one grade band and the signed error points the same way in at least seven of the ten rounds. A median that moves without a stable direction is noise that happened to land well.

If two consecutive weeks show no movement, change the drill rather than the volume. The usual cause is practising the same correction that was already absorbed, and more of it produces a flatter curve, not a higher one.