As of October 2026｜About 248000 site records: aim training accounts for almost two thirds
0
October 17, 2026
Daniel LuFull-Stack Engineer | Content Creator
A breakdown of 248024 submissions, separating activity, record labels and repeated attempts.
Categories：Data & Insights
As of October 2026
Of 248024 records, 160969 belong to aim training: 64.90%. That is a participation pattern, not evidence that all tasks receive comparable attention.
What the table includes
This article uses a fixed snapshot read on 17 October 2026 at 03:54:35 Beijing time. Only human_benchmark_records is counted: 248024 records, dated from 23 December 2025 to 17 October 2026, with the latest at 01:23:55 that day. The legacy human_scores table's 6507 rows are not added. These are site submissions, not a random population sample.
The table also contains trading simulation, the 24 game and a few 2048 records. Its name does not make every row a physiological measurement.
Aim training leads by record count
| Task | Records | Share of main table | Record labels |
|---|---|---|---|
| Aim training | 160,969 | 64.90% | 37,432 |
| Colour sensitivity | 28,138 | 11.35% | 6,942 |
| Reaction time | 21,850 | 8.81% | 7,024 |
| Stroop task | 8,534 | 3.44% | 2,702 |
| Sequence memory | 7,450 | 3.00% | 3,475 |
| Grayscale task | 6,092 | 2.46% | 2,852 |
| Trading simulation | 5,384 | 2.17% | 2,109 |
| Colour-pattern task | 4,951 | 2.00% | 2,885 |
| Covered-number memory | 4,310 | 1.74% | 2,271 |
| 24 game | 332 | 0.13% | 133 |
| 2048 | 14 | 0.01% | 12 |
There are 46644 labels across the table. Aim training has one record without a usable label: it counts towards activity but not label-based grouping.
Records are grouped by account ID when present, otherwise by guest name, with the two namespaces kept separate. A label is not a verified person: people may change labels and labels may be shared. Counts from different tools overlap and cannot be added as a total population.
Repeated attempts change the weighting
Aim-training submissions per label have a median of two. The ten most active labels contribute 9761 records, or 6.06% of labelled aim records. This does not establish cheating or identify who operated a label; it shows that frequent submitters carry more weight in an all-record distribution.
| Task | Records from ten most active labels | Share of labelled records |
|---|---|---|
| Aim training | 9,761 | 6.06% |
| Colour sensitivity | 3,825 | 13.59% |
| Reaction time | 1,040 | 4.76% |
| Stroop task | 1,099 | 12.88% |
Monthly records are not traffic
| Complete month | Records |
|---|---|
| 2026-06 | 41,031 |
| 2026-07 | 50,678 |
| 2026-08 | 39,571 |
| 2026-09 | 20,358 |
Months use Beijing calendar dates. October is incomplete and is not compared with full months. Changes may reflect participation, reporting, task mix or repeated attempts; this table alone cannot separate them or measure website traffic.
Choose the denominator for the question
Use record counts for submission activity, labels for coverage, and both all-record and first-per-label distributions for scores. A tool can be useful without turning its result into a ranking of people.
🔗 Aim training
Practise locating targets and compare time with accuracy.
→
🔗 Reaction time
Click when the colour changes and record several attempts.
→
No score filter is used for participation counts. Label grouping excludes missing labels. Top ten means the ten largest submission counts within each task. No raw IDs, guest names or individual histories are published.