Friedrich Schulte’s grid of numbers has been around since the 1960s, and the modern version — a 5×5 table of digits you tap in order — has been a fixture of brain-training literature for almost as long. The exercise is genuinely useful. The marketing around it has, in the last ten years, drifted further and further from the science. This post is the honest version.

What the practice is

A grid of numbers (or letters, or symbols) appears on a flat surface. The user is asked to find them in sequence — 1, 2, 3, 4, 5 — as quickly as possible. Performance is typically measured in total time, and sometimes in time per cell. The grid sizes that show up most often in research and in apps like Schulte Table Game are 3×3, 4×4, and 5×5.

The task is a continuous performance test. You cannot pause to think. You cannot externalize the work. You have to keep your eyes on the centre of the grid and use peripheral vision to register where each number is. The smaller the grid, the more you can rely on foveal (central) vision. The larger the grid, the more the periphery has to do.

What it actually trains

Three things, in order of how well the evidence supports them:

  1. Peripheral visual attention. This is the headline claim and the one that has the most empirical support. Studies on the Schulte table and on similar grid-search tasks show measurable improvements in the speed and accuracy with which participants can locate targets outside their foveal field.

  2. Sustained attention. The continuous-performance nature of the task — no pauses, no feedback until the end of the grid — is similar to the kind of attentional control measured by continuous performance tests in clinical psychology. People who practice the Schulte table tend to get better at the table itself, and the transfer to other sustained-attention tasks is modest but real.

  3. Reading speed. This is the most-debated claim. The theory is that a wider useful field of vision means fewer saccades per line, which means faster reading. There is some evidence for this in educational contexts, but the effect size is small and the studies are not as rigorous as the marketing suggests.

What it does not train

It does not train fluid intelligence. It does not raise your IQ. It does not, on its own, make you a better driver or a better chess player or a better surgeon. The Schulte table is a perceptual training tool, not a general cognitive enhancer. The best evidence we have for transfer effects is to other perceptual tasks — the kind of near-transfer that shows up in any well-designed practice study.

The right way to practice

A few things that hold up across the literature:

  • Keep sessions short. Five to ten minutes is the sweet spot. Longer sessions show diminishing returns and, past a certain point, fatigue starts to undo the training effect.
  • Be consistent. Daily short sessions beat weekly long ones. The dose-response curve is not linear.
  • Track your time per cell, not your total time. A 5×5 grid takes longer than a 3×3 by design. The interesting metric is how long, on average, it takes you to find the next cell, regardless of grid size.
  • Mix grid sizes. If you only ever practice 5×5, you only ever get better at 5×5. The Schulte Table Game app rotates grid sizes for exactly this reason.

The Schulte table is one of the few cognitive exercises with a century of evidence behind it. It is worth doing. It is worth doing honestly. And it is worth knowing what the practice does and does not do for you.