Player data
On the Dot asks players to stop a hidden stopwatch at a target time between 5 and 15 seconds. From launch on 2 August to 5 October 2026, players completed 61,936 rounds and finished 8,366 five-round games. These are the patterns in those rounds.
Scores
Each round scores how close the stop came as a percentage of the target, and a game's score is the average of its five rounds. Half of all finished games scored 89.9% or more. The middle half landed between 78.9% and 94.4%.
Table
| Score | Games |
|---|---|
| Below 50% | 754 |
| 50.0 to 50.9% | 20 |
| 51.0 to 51.9% | 20 |
| 52.0 to 52.9% | 20 |
| 53.0 to 53.9% | 22 |
| 54.0 to 54.9% | 21 |
| 55.0 to 55.9% | 23 |
| 56.0 to 56.9% | 16 |
| 57.0 to 57.9% | 28 |
| 58.0 to 58.9% | 30 |
| 59.0 to 59.9% | 25 |
| 60.0 to 60.9% | 28 |
| 61.0 to 61.9% | 25 |
| 62.0 to 62.9% | 32 |
| 63.0 to 63.9% | 41 |
| 64.0 to 64.9% | 44 |
| 65.0 to 65.9% | 30 |
| 66.0 to 66.9% | 41 |
| 67.0 to 67.9% | 49 |
| 68.0 to 68.9% | 49 |
| 69.0 to 69.9% | 50 |
| 70.0 to 70.9% | 55 |
| 71.0 to 71.9% | 60 |
| 72.0 to 72.9% | 64 |
| 73.0 to 73.9% | 82 |
| 74.0 to 74.9% | 73 |
| 75.0 to 75.9% | 78 |
| 76.0 to 76.9% | 86 |
| 77.0 to 77.9% | 107 |
| 78.0 to 78.9% | 117 |
| 79.0 to 79.9% | 97 |
| 80.0 to 80.9% | 126 |
| 81.0 to 81.9% | 130 |
| 82.0 to 82.9% | 146 |
| 83.0 to 83.9% | 163 |
| 84.0 to 84.9% | 172 |
| 85.0 to 85.9% | 208 |
| 86.0 to 86.9% | 199 |
| 87.0 to 87.9% | 253 |
| 88.0 to 88.9% | 304 |
| 89.0 to 89.9% | 328 |
| 90.0 to 90.9% | 352 |
| 91.0 to 91.9% | 457 |
| 92.0 to 92.9% | 497 |
| 93.0 to 93.9% | 497 |
| 94.0 to 94.9% | 588 |
| 95.0 to 95.9% | 569 |
| 96.0 to 96.9% | 555 |
| 97.0 to 97.9% | 399 |
| 98.0 to 98.9% | 211 |
| 99.0 to 99.9% | 25 |
754 finished games (9%) scored below 50% and are left out of the chart. One round is a fifth of the score, so a single round stopped far too early can pull a game down by up to 20 points.
| Share of finished games | Score needed |
|---|---|
| Top half | 89.9% |
| Top quarter | 94.4% |
| Top 10% | 96.6% |
| Top 5% | 97.5% |
| Top 1% | 98.6% |
The best finished game so far scored 99.68%. Across all 61,936 rounds:
- 248 rounds, about one in 250, were stopped on the exact hundredth of a second.
- 8.3% landed within a tenth of a second of the target.
- 33.9% landed within half a second.
- 53.3% landed within one second.
The first round
The first round of a game is the least accurate, and its misses lean one way. In 65.9% of first rounds the player stopped too early. From round 2 on the split is close to even, between 50.3% and 51.5% early, and median accuracy rises from 84.0% in round 1 to 93.4% in round 5.
Table
| Round | Rounds played | Stopped early | Median accuracy |
|---|---|---|---|
| Round 1 | 17,500 | 65.9% | 84.0% |
| Round 2 | 13,757 | 50.9% | 89.1% |
| Round 3 | 11,382 | 51.5% | 91.8% |
| Round 4 | 9,584 | 50.8% | 92.8% |
| Round 5 | 8,403 | 50.3% | 93.4% |
Most of that early lean comes from new players. In the first round of a player's first game, 72.9% stopped early and the median stop came at 79.8% of the target, about 8 seconds into a 10-second target. 15.2% of those first attempts were stopped within 2.5 seconds, around the moment the numbers fade. Leaving out every stop before the halfway point, the median first attempt still came at 93.5% of the target.
Table
| First round of | Rounds | Median stop | Stopped early | Median accuracy |
|---|---|---|---|---|
| Player's first game | 10,710 | 79.8% of target | 72.9% | 74.0% |
| Games 2 to 4 | 4,774 | 98.6% of target | 55.0% | 90.2% |
| Game 5 or later | 2,480 | 99.5% of target | 53.1% | 94.9% |
Players who come back stop much closer to the target in round 1: at a median of 98.6% of the target in their games 2 to 4, and 99.5% from game 5 on.
Long and short targets
Targets run from 5 to 15 seconds. From round 2 on, the typical miss grows with the target, from 0.45 seconds on 5-second targets to 1.21 seconds on 14-second targets. As a share of the target, the typical miss stays between 7.7% and 8.8% at every length, and median accuracy stays between 91.3% and 92.3%.
In seconds
As a share of the target
Table
| Target | Rounds | Typical miss | Share of target | Median stop | Stopped early | Median accuracy |
|---|---|---|---|---|---|---|
| 5.00 to 5.99 s | 4,979 | 0.45 s | 8.1% | 100.2% | 48.9% | 91.9% |
| 6.00 to 6.99 s | 3,620 | 0.52 s | 8.0% | 101.0% | 45.7% | 92.0% |
| 7.00 to 7.99 s | 5,133 | 0.64 s | 8.6% | 100.4% | 48.1% | 91.5% |
| 8.00 to 8.99 s | 3,535 | 0.66 s | 7.8% | 99.8% | 50.2% | 92.2% |
| 9.00 to 9.99 s | 4,875 | 0.82 s | 8.6% | 100.0% | 49.8% | 91.4% |
| 10.00 to 10.99 s | 4,497 | 0.86 s | 8.3% | 99.9% | 50.2% | 91.7% |
| 11.00 to 11.99 s | 3,496 | 0.90 s | 7.7% | 99.0% | 53.7% | 92.3% |
| 12.00 to 12.99 s | 5,224 | 1.02 s | 8.2% | 99.0% | 53.8% | 91.8% |
| 13.00 to 13.99 s | 4,244 | 1.19 s | 8.8% | 98.7% | 55.0% | 91.3% |
| 14.00 to 14.99 s | 3,523 | 1.21 s | 8.3% | 98.7% | 54.9% | 91.7% |
Timing researchers call this the scalar property: the error in a timed interval grows in proportion to the interval. Sense of time covers the research. It is also why each round is scored as a percentage of its target. Under that rule a 14-second target costs no more points than a 5-second one, and the median scores above bear that out.
Long targets do pull stops a little early. From round 2 on, the median stop on 5- to 10-second targets falls between 99.8% and 101.0% of the target, and 45.7% to 50.2% of stops are early. On 11- to 14-second targets the median stop is between 98.7% and 99.0% of the target, and 53.7% to 55.0% of stops are early.
After a miss
A miss in one direction makes the next miss more likely to go the same way. After a round that ended more than 10% early, 69.4% of next rounds in the same game also ended early. After a round more than 10% late, 67.3% of next rounds were late again.
Table
| Previous round | Rounds | Next stopped early | Next stopped late |
|---|---|---|---|
| Over 10% early | 13,788 | 69.4% | 30.4% |
| 3 to 10% early | 6,282 | 54.2% | 45.1% |
| Within 3% | 9,193 | 46.7% | 52.6% |
| 3 to 10% late | 7,054 | 37.7% | 61.8% |
| Over 10% late | 7,593 | 32.3% | 67.3% |
These figures can't separate two explanations. Some players run consistently fast or slow, and a single round of feedback may not be enough to reset a player's sense of the beat. Either way, the data shows no sign of players overcorrecting.
Practice
Players improve quickly. Of the 1,440 players who finished at least two games, 78.7% scored higher in their second game than in their first. The median gain was 6.2 points, and the median score went from 84.5% in the first game to 92.7% in the second. Of the 168 who finished at least five games, 69.0% scored higher in their fifth game than their first, with a median gain of 3.6 points.
Phones and computers
Tapping and typing score about the same. Median round accuracy was 91.1% on phones (19,334 rounds) and 90.3% on desktop computers (41,074 rounds). Tablets were lower at 84.2%, from a smaller sample of 1,528 rounds.
Method
These figures come from the anonymous gameplay events the game records, described in the privacy policy. They cover hitthedot.app from launch on 2 August 2026 to the end of 5 October 2026, Pacific time, and include daily, practice and challenge games.
- Excluded: traffic flagged as automated, test builds on other web addresses, and rounds on the Hall of Fame's fixed set of five targets, which players can replay without limit.
- Players are counted by browser, so one person on two devices counts as two. The 61,936 rounds came from 10,710 browsers.
- "Typical" and "median" both mean the middle value.
- A game counts as finished when all five rounds were played. 17,492 games were started and 8,366 (47.8%) were finished.
- The rounds counter on the game's home page counts every round, including the exclusions above, so it runs higher than the totals here.