AFOQT Block Counting Practice Test (30 Questions, 5 Minutes)

Counting what is hidden

You are shown a pile of cubes drawn in isometric projection with one block marked, and you count how many other blocks touch it. Thirty questions in five minutes is about ten seconds each.

The visible blocks are not the problem. An isometric drawing hides everything behind and beneath the faces you can see, and the marked block usually touches at least one of them. The subtest is really asking whether you can reason about a structure from a partial view — quickly, thirty times.

Touching means sharing a face

Two blocks touch only when a whole face is in contact. Blocks that meet along an edge, or at a single corner, do not count. Diagonal neighbours are the most common source of an over-count, and because the answer options are clustered around the correct number, one extra block is enough to lose the mark.

A single block can therefore touch at most six others: four sides, one above, one below. If your count comes out higher than six, you have counted a diagonal.

A fixed order beats intuition

Ten seconds is enough for six checks if you never have to decide which check to do next. Use the same sequence every time and say each one to yourself:

1. Left — is there a block level with the marked one on its left?
2. Right.
3. Front.
4. Back — the one people forget, because it is hidden by definition.
5. Directly below — present unless the marked block sits on the ground.
6. Directly above — present unless the marked block is on top of its column.

Six checks, maximum six touching blocks. The value of a fixed order is not speed on any one item; it is that you stop spending time deciding where to look, and you stop forgetting the back.

Infer the hidden blocks from support

A cube cannot float. If you can see a block at the third level, there are blocks beneath it at levels one and two, whether or not they are drawn. Reading the pile as a structure that has to hold itself up tells you what is underneath without seeing it.

The same logic works sideways. A block that appears to be resting partly on something must have that something beneath it. Once you start reading piles as physical objects rather than as drawings, the hidden blocks stop being guesswork.

This is why our piles are generated from a height map — a grid of column heights. That representation makes a floating block impossible to express, so every structure you meet here is physically valid and the support logic always works. The number of touching blocks is then computed from the same height map that draws the picture, which means the answer and the figure cannot disagree.

Pacing and the answer spread

The options are numbers close together, so there is no partial credit for being nearly right and no way to eliminate by plausibility. Either your count is correct or it is not.

Given that, do not deliberate. Run the six checks, take the number, mark it, move. If you lose track halfway through — which happens — restart the sequence rather than trying to remember where you were, because a half-remembered count is worse than a fresh one and costs about the same time.

There is no penalty for a wrong answer, so fill everything in. At ten seconds an item you will feel the clock, and the last few questions are where blanks appear if you have not built the habit of committing.

Where it counts, and when it arrives

Block Counting contributes to the rated composites, so it matters most to Combat Systems Officer and Air Battle Manager applicants. Alongside Table Reading it forms the perceptual-speed part of those composites, and both improve almost purely through repetition rather than study — which makes them unusually good value for the time invested.

It also falls near the very end of the appointment, after Instrument Comprehension and before Aviation Information. Spatial reasoning at ten seconds a question is markedly harder when you are three and a half hours in, which is an argument for meeting it at least once inside a full-length exam rather than only fresh.

If you are not applying for a rated slot, this subtest may not feed a composite you need at all — check how the composites are built before spending time here.

Where the count usually goes wrong

Three errors account for most lost marks, and they are worth naming so you can catch yourself making them.

Counting a diagonal. A block touching at an edge or corner looks adjacent and is not. If your answer exceeds six, this is what happened.

Forgetting the hidden back face. The one direction you cannot see is the one the fixed sequence exists to protect. If you find yourself skipping straight from front to below, you dropped it.

Assuming the ground. A marked block partway up a column has one below it; a marked block on the base does not. Candidates who count “below” automatically over-count on ground-level blocks, and those are common.

Why a poor first attempt means very little

Almost nobody has done this task before. Counting how many blocks touch a marked block in a drawn pile is unlike school, work or any other test, and a first attempt measures unfamiliarity far more than ability.

What improves is a method and a rhythm, and both come from repetition rather than from understanding. The method is worth being deliberate about: work the same way round every time — the block above, the block below, then the neighbours — so that under time pressure the sequence runs without deciding. Candidates who count in a different order each time lose marks to double-counting rather than to the geometry.

This is also a subtest where short, frequent practice beats long sessions. Ten minutes three times a week builds the rhythm; an hour on a Sunday mostly builds fatigue.