Review locally saved Schulte Table attempts without creating an account. Filter by date, grid size, mode, completion, and input method; then inspect time, mistakes, personal bests, medians, recent averages, and consistency. Trends remain separated by size so a longer grid is not mistaken for slower performance on a shorter one.
Time, best, and trend use one comparable group. Older records and tools without measurement metadata remain available to view and export.
No matching records yet. Complete a grid, visual-search drill, daily challenge, or guided session.
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.
Each unified record contains an anonymous stable ID, ISO completion time, grid size, cell count, practice mode, duration, mistakes, completion state, and an optional broad input method. It does not contain a name, email, IP address, or device fingerprint. Challenge records may include a local challenge identifier.
Choose a comparable conditions group after filtering. The default uses the newest group with complete measurement details. Best is its shortest completed raw duration; average uses its completed times, median is the middle sorted time, and recent average uses its latest seven. Attempt counts describe all filtered records. Selecting all records leaves combined time statistics empty rather than mixing tasks.
Classic records specify size, mode, first-correct timing, elapsed time including hidden tabs, count-only errors, input, display profile, and random or same-grid replay. Speed Test has a separate source group. Older records and tools without these details remain listed and exportable, but do not enter strict bests or time trends. A display profile describes the local viewport and appearance settings, not a calibrated device measurement.
A fast attempt with several wrong selections is not equivalent to an accurate attempt. The tracker shows average mistakes and an accuracy proxy: the share of completed attempts with zero recorded mistakes. This is a plain practice summary, not a sensory, cognitive, or medical assessment.
Consistency is the coefficient of variation for completed durations: population standard deviation divided by the average. A smaller percentage means the filtered times are closer together. It describes only the spread of your saved practice times and should not be interpreted as psychological or medical stability.
CSV is convenient for a spreadsheet; JSON preserves the full schema for backup and re-import. Import validates every item and removes duplicate IDs. Delete filtered records or all records only through the confirmation dialog. Export a backup first if the local history matters to you.
Screen size, pointer type, keyboard use, interruptions, browser load, grid mode, and familiarity can change a result. Compare like with like and look across several rounds. Hiding statistical outliers changes only the chart view; it never silently deletes the underlying records.
On the first Tracker visit, a one-time migration can copy supported older histories without removing the originals. A missing legacy mistake count becomes zero and an unparseable date becomes the Unix epoch; neither proves an observed error-free attempt on that date. The migration is not a continuous sync with every page. New standard-grid times may need manual entry through the Score Calculator. Input-method filtering appears when that field is present, but device model, zoom, display orientation, and timer-trigger protocol are not stored. Keep those conditions in separate notes. Untimed number-letter or large-print records can have duration zero; zero does not mean an instantaneous timed round. The Tracker has no timer-state filter. Keep those records separate using the exported configuration context and your notes. Mixed modes may still appear together under All, so select the relevant mode as well as size before interpreting statistics.
It is stored under a versioned localStorage key in this browser and is not synced to an account.
It is the population standard deviation of completed times divided by their average, displayed as a percentage.
No. Valid imported records are merged with current records, and matching IDs are kept only once.
You can inspect them as separate groups, but total time should not be treated as directly comparable across sizes.