World Player Statistics
The distribution of results across every MIKIRI run (each completed run counts once, so repeat plays by the same person are included; flagged runs excluded; refreshed hourly). Use it to see where your record sits. Samples come from browsers worldwide, so this is play data, not an academic dataset.
Updated: 2026-10-04 03:16 UTC
FLASH โ rapid recognition
The level reached maps to how many digits you read and for how many frames (1/60 s each). Level 1 is 3 digits for 0.5 s; level 13 is 5 digits for about 0.17 s; level 21 is 7 digits for about 0.07 s; level 28 is 10 digits for 0.05 s. Each step makes only one thing harder โ digit count or speed (levels that add a digit slow the display down) โ so the level where you stop tells you which of the two held you back.
View as table
| Level / round reached | runs |
|---|---|
| LV.1 | 6 |
| LV.2 | 9 |
| LV.3 | 6 |
| LV.4 | 4 |
| LV.5 | 18 |
| LV.6 | 10 |
| LV.7 | 11 |
| LV.8 | 7 |
| LV.9 | 12 |
| LV.10 | 58 |
| LV.11 | 27 |
| LV.12 | 33 |
| LV.13 | 18 |
| LV.14 | 40 |
| LV.15 | 24 |
| LV.16 | 9 |
| LV.17 | 25 |
| LV.18 | 8 |
| LV.19 | 6 |
| LV.20 | 2 |
| LV.22 | 1 |
MATCH โ visual search
Total time to keep picking the digit that matches the sample. Each wrong tap adds one second, so this distribution mixes speed and accuracy. One second is worth several answers, so cutting a single mistake usually saves more than rushing one answer.
View as table
| Clear time (5s bins) | runs |
|---|---|
| 15โ20s | 6 |
| 20โ25s | 9 |
| 25โ30s | 26 |
| 30โ35s | 28 |
| 35โ40s | 30 |
| 40โ45s | 32 |
| 45โ50s | 17 |
| 50โ55s | 13 |
| 55โ60s | 8 |
| 60โ65s | 1 |
| 65โ70s | 2 |
| 70โ75s | 1 |
| 75โ80s | 2 |
| 95โ100s | 2 |
| 340โ345s | 1 |
NEXT โ read-ahead reaction
Reaction time that includes movement. Unlike a simple press-a-button test, you move your finger to the next circle and aim, so values cluster noticeably slower. Do not read these as a slow simple reaction time โ it is a different task.
View as table
| Average reaction time (25ms bins) | runs |
|---|---|
| 300โ325ms | 1 |
| 325โ350ms | 1 |
| 350โ375ms | 1 |
| 375โ400ms | 5 |
| 400โ425ms | 10 |
| 425โ450ms | 10 |
| 450โ475ms | 7 |
| 475โ500ms | 20 |
| 500โ525ms | 11 |
| 525โ550ms | 10 |
| 550โ575ms | 7 |
| 575โ600ms | 8 |
| 600โ625ms | 8 |
| 625โ650ms | 13 |
| 650โ675ms | 10 |
| 675โ700ms | 3 |
| 700โ725ms | 5 |
| 725โ750ms | 5 |
| 775โ800ms | 4 |
| 800โ825ms | 3 |
| 825โ850ms | 6 |
| 925โ950ms | 1 |
| 950โ975ms | 1 |
| 975โ1000ms | 1 |
| 1000โ1025ms | 1 |
| 1025โ1050ms | 1 |
| 1100โ1125ms | 1 |
| 1750โ1775ms | 1 |
| 2025โ2050ms | 1 |
| 2500โ2525ms | 1 |
| 2975โ3000ms | 1 |
FOLLOW โ eye tracking
Rounds survived while tracking one highlighted dot among identical moving dots. Each round adds dots, speed and shuffle time. Once your eyes lose the target it cannot be recovered, so the peak of the distribution shows where attention tends to slip.
View as table
| Level / round reached | runs |
|---|---|
| R.1 | 4 |
| R.2 | 2 |
| R.3 | 6 |
| R.4 | 3 |
| R.5 | 8 |
| R.6 | 6 |
| R.7 | 12 |
| R.8 | 13 |
| R.9 | 13 |
| R.10 | 16 |
| R.11 | 16 |
| R.12 | 8 |
| R.13 | 4 |
| R.14 | 4 |
| R.15 | 1 |
| R.16 | 2 |
| R.19 | 1 |
| R.23 | 1 |
INTERCEPT โ trajectory sniping
Rounds survived by predicting where a vanished target is now. The multiplier grows the longer you wait, so the top records reflect judgement about when to shoot as well as accuracy. The gap between the median and the top 10% tends to be wide because of that risk-and-reward choice.
View as table
| Level / round reached | runs |
|---|---|
| R.1 | 22 |
| R.2 | 13 |
| R.3 | 17 |
| R.4 | 10 |
| R.5 | 8 |
| R.6 | 9 |
| R.7 | 10 |
| R.8 | 5 |
| R.9 | 5 |
| R.10 | 5 |
| R.11 | 1 |
| R.12 | 8 |
| R.13 | 1 |
| R.14 | 5 |
| R.15 | 1 |
| R.16 | 1 |
| R.17 | 1 |
| R.18 | 1 |
| R.19 | 2 |
| R.20 | 1 |
| R.22 | 2 |
| R.25 | 1 |
OFFBEAT โ rhythm perception
Rounds survived by finding the one dot whose beat is off. The offset shrinks every round. Records played with sound off are included, which is why many runs end early โ the format depends on hearing the rhythm.
View as table
| Level / round reached | runs |
|---|---|
| R.1 | 31 |
| R.2 | 16 |
| R.3 | 7 |
| R.4 | 4 |
| R.5 | 8 |
| R.6 | 4 |
| R.7 | 4 |
| R.8 | 4 |
| R.9 | 3 |
| R.10 | 1 |
| R.12 | 3 |
| R.14 | 1 |
| R.29 | 1 |
MIRAGE โ illusions
Rounds survived choosing the objectively correct option under an optical illusion. The real difference shrinks and the time limit tightens each round. Deciding what to compare (edges, centres, area) before looking reduces the pull of the illusion better than trusting the first impression.
View as table
| Level / round reached | runs |
|---|---|
| R.1 | 4 |
| R.2 | 6 |
| R.3 | 7 |
| R.4 | 2 |
| R.5 | 3 |
| R.6 | 8 |
| R.7 | 4 |
| R.8 | 2 |
| R.9 | 8 |
| R.10 | 6 |
| R.11 | 12 |
| R.12 | 6 |
| R.13 | 4 |
| R.14 | 9 |
| R.15 | 3 |
| R.16 | 6 |
| R.17 | 3 |
| R.18 | 6 |
| R.19 | 1 |
| R.20 | 3 |
| R.21 | 2 |
| R.22 | 1 |
| R.23 | 2 |
| R.26 | 1 |
| R.27 | 1 |
| R.30 | 1 |
| R.37 | 1 |
| R.42 | 1 |
| R.61 | 1 |
GATEFALL โ gate reading
Rounds survived reading the approaching gate stones. Forbidden red stones must be let through, so the record reflects the decision not to react as well as speed.
View as table
| Level / round reached | runs |
|---|---|
| R.2 | 3 |
| R.3 | 3 |
| R.4 | 4 |
| R.5 | 9 |
| R.6 | 12 |
| R.7 | 12 |
| R.8 | 6 |
| R.9 | 3 |
| R.11 | 2 |
| R.19 | 1 |
| R.21 | 3 |
| R.23 | 1 |
BLINK HUNT โ selective memory
Rounds survived remembering only the lights that match the rule. It is about filtering, not memorising everything; boards and rules get more complex each round.
View as table
| Level / round reached | runs |
|---|---|
| R.1 | 4 |
| R.2 | 2 |
| R.4 | 3 |
| R.5 | 2 |
| R.6 | 5 |
| R.7 | 4 |
| R.8 | 8 |
| R.9 | 16 |
| R.10 | 11 |
| R.11 | 8 |
| R.12 | 11 |
| R.13 | 4 |
| R.14 | 1 |
| R.15 | 3 |
| R.16 | 2 |
| R.17 | 1 |
| R.18 | 3 |
NAGARE โ moving numbers
Rounds survived reading a number as it crosses the screen. Each round either shortens the pass time or adds a digit, so the round where you stop shows which one held you back.
ZANZOU โ position memory
Rounds survived tapping, from 1 upward, where briefly shown numbers were. Each round makes either the count or the display time harder.
ANZAN โ flash mental math
Rounds survived adding up numbers shown one after another. Each round raises one thing: the count, the speed or the number of digits.
RENBAN โ number order
Rounds survived tapping 1, 2, 3โฆ before time runs out. From round 13 the numbers move after every tap, so memorising the layout stops working.
SENKO โ gold digits
Rounds survived catching the gold digits in a fast stream. From round 13 there are two golds, and the closer together they are, the easier the second is to miss.
KIZAMI โ rhythm copy
Rounds survived tapping back the model rhythm. Offbeats start at round 7, triplets at 11 and quintuplets at round 21 (demon).