This site does not publish an unsupported population average. A useful comparison starts with your own completed attempts under the same grid, device, input, display, and error conditions.
| Grid | Cells | Targets versus 5×5 | Layout note |
|---|---|---|---|
| 3×3 | 9 | 16 fewer | The site's shortest standard numeric sequence |
| 4×4 | 16 | 9 fewer | Seven more targets than 3×3 |
| 5×5 | 25 | Fixed reference size | Five rows and 25 selections |
| 6×6 | 36 | 11 more | Eleven more targets than 5×5—a 44% increase |
| 7×7 | 49 | 24 more | Nearly twice the 25 cells of 5×5 |
| 8×8 | 64 | 39 more | Exactly four times the 16 cells of 4×4 |
| 9×9 | 81 | 56 more | Nine times the cell count of 3×3 |
| 10×10 | 100 | 75 more | One hundred targets across ten rows and columns |
Use the guide when you need a paper reference for task length and comparison conditions. It explains observed task results without introducing universal norms or ability bands.
Use this paper record when you want a handwritten log or need to record offline attempts. It does not read, write, or synchronize the browser Progress Tracker.
No verified dataset in this project establishes one population average across ages, devices, table designs, and rules. Publishing a precise cutoff without a sampling method, participant definition, task protocol, and uncertainty would create a number that looks authoritative without a sound basis.
Cell count rises from 9 on 3×3 to 100 on 10×10. Each added target extends the required sequence, while extra rows can reduce target size in the same viewport. Total time from two sizes therefore answers two different task questions.
A mouse pointer travels and remains visible at its last position. A finger can cover the target and move the device. Keyboard selection follows focus or number-key behavior. Keep input methods separate in the Progress Tracker instead of assuming one is always faster.
The same CSS layout can occupy different physical sizes on a phone, tablet, laptop, or monitor. Portrait versus landscape, page padding, and 100%, 125%, or 150% zoom change target geometry and visual travel. Record those settings when a strict comparison matters.
Completion time alone omits wrong selections. If one mode counts or penalizes errors differently, its result is not like-for-like with another rule. Review time and mistakes together, and discard interrupted attempts rather than forcing them into a trend.
A personal best is the shortest saved completion and can be an unusual round. The median is the middle completed time after sorting, so it is less dominated by one extreme result. Review a median from several matching attempts beside the best and average mistakes.
The Progress Tracker filters saved browser records by grid, mode, completion, and input, then shows median, recent average, best, and errors. The Score Calculator transparently applies a site-defined error penalty to one entered result; it is not a population or ability score.
Seconds per cell can describe total time divided by cell count, but it does not make layouts identical: target density, sequence length, and input travel still differ. Keep a separate history for each size and use a neighboring grid only when you intentionally want a different task.
No verified site-user or population dataset supports fixed numerical bands, so the page uses factual task properties and your own matching history instead.
Not directly. The grids contain different numbers of cells and use different layouts.
Compare several completed attempts with the same grid, rules, device, input method, orientation, and zoom, then review median and errors together.