Divide a game's lifetime visits by the number of people playing it right now and you get a rough measure of how much history each current player represents. It is not a quality score and it is not a popularity score. It is a ratio between two clocks, and reading it correctly means reading both halves. This is the full board as it stood on November 3 2026 across 89 directory games that carry a complete set of statistics.
| # | Game | Lifetime visits | Playing now | Visits per player | Live rank | Visits rank |
|---|---|---|---|---|---|---|
| 1 | MeepCity | 16,279,291,580 | 112 | 145,350,817 | 89 | 10 |
| 2 | Tower of Hell | 29,328,678,401 | 11,954 | 2,453,461 | 45 | 7 |
| 3 | PLS DONATE | 5,260,382,294 | 2,370 | 2,219,570 | 83 | 31 |
| 4 | Grow a Garden | 36,003,571,189 | 16,859 | 2,135,569 | 35 | 4 |
| 5 | Work at a Pizza Place | 5,622,559,341 | 2,892 | 1,944,176 | 80 | 30 |
| 6 | Arsenal | 6,325,514,646 | 3,628 | 1,743,526 | 74 | 26 |
| 7 | Break In (Story) | 3,025,016,058 | 1,819 | 1,663,010 | 85 | 46 |
| 8 | Jailbreak | 8,084,652,785 | 4,919 | 1,643,556 | 66 | 20 |
| 9 | Rainbow Friends | 5,166,908,768 | 3,305 | 1,563,361 | 78 | 33 |
| 10 | [LEGACY] Toilet Tower Defense | 6,202,752,005 | 4,325 | 1,434,162 | 71 | 27 |
The number that dominates the table is 145,350,817. MeepCity has 16,279,291,580 lifetime visits and 112 players online, and the division produces a figure 59 times larger than the runner-up. That is not a rounding artefact and it is not a data error: it is what happens when a title accumulates an enormous lifetime counter and then retains a very small live audience. MeepCity sits at rank 89 of 89 by concurrent players and rank 10 by lifetime visits, so the ratio is simply expressing the distance between those two positions.
Everything below rank one is a different kind of game. Tower of Hell at 2,453,461 is a title with 29,328,678,401 visits and 11,954 players — a large audience by any measure, just one where the history is proportionally smaller. You can see how it is categorised on the obby games page, and its own list is on the tower of hell codes page.
Ranks three through ten cluster tightly. PLS DONATE at 2,219,570 and [LEGACY] Toilet Tower Defense at 1,434,162 differ by a factor of about 1.5, which is small compared with the gap between rank one and rank two. Most of the directory sits in this band, and the median entry is a dusty trip [SEASON] at 242,307.
The five highest-ratio games have a median of 16,279,291,580 lifetime visits and a median of 2,892 concurrent players. The five lowest have a median of 61,456,271 visits and a median of 30,375 players. Both bands therefore contain games with large visit counts — the difference is entirely in the denominator. A high ratio is produced by a small live population, not by a large visit count.
Two entries make this concrete. Grow a Garden carries 36,003,571,189 visits, one of the largest totals in the directory, and still lands at 2,135,569 because 16,859 people are playing it. Anime Dice carries 156,846,962 visits and lands at 3,106 with 50,497 playing. The visit counts differ by a factor of roughly 230; the ratios differ by a factor of 688, and the reason is entirely the live column.
| # | Game | Lifetime visits | Playing now | Visits per player | Live rank | Visits rank |
|---|---|---|---|---|---|---|
| 89 | [] Adopt Me! | 728 | 65,311 | 0 | 13 | 89 |
| 88 | Anime Dice | 156,846,962 | 50,497 | 3,106 | 15 | 80 |
| 87 | Huss Valley | 232,318,801 | 30,375 | 7,648 | 22 | 78 |
| 86 | Build An Ant Empire | 30,869,252 | 3,767 | 8,194 | 73 | 86 |
| 85 | Cold War [VIETNAM] | 61,456,271 | 6,126 | 10,032 | 61 | 83 |
The lowest ratio in the directory belongs to [] Adopt Me!, at 0.01 visits per concurrent player on 728 visits and 65,311 players. Read literally that says the lifetime counter is smaller than the crowd currently inside it, which is unusual and worth flagging rather than explaining away. The same title also carries an internal inconsistency elsewhere in its record, so treat it as an outlier to be aware of, not a pattern to generalise from.
Second from bottom, Anime Dice at 3,106, is the more instructive case: a game where the lifetime counter and the live population are close enough that the ratio collapses toward the hundreds. Games in this band are either very new or have a live audience that has outrun the recorded history.
First, as a retention signal. A game with a low ratio and a big live population, like Steal An Egg, is holding people in sessions right now. A game with a high ratio is mostly memory.
Second, as a discovery signal. High-ratio games are frequently worth visiting precisely because the lifetime count says a very large number of people found something there, even if the current crowd is thin. Jailbreak at 1,643,556 with 8,084,652,785 visits is in that category, and so is Arsenal at 1,743,526 — you can check the shooter side of the directory through the arsenal codes page and the jailbreak codes page.
Third, as a warning about mixing columns. The ratio is only meaningful when both of its inputs come from the same moment. Visits are cumulative and players are instantaneous, so the ratio drifts upward on its own as time passes and nothing else changes. Comparing a game's ratio today against its ratio six months from now measures elapsed time as much as it measures the game.
It says nothing about session length. Two games with identical ratios can have completely different player behaviour — one may have short, frequent sessions and the other long, rare ones. It says nothing about how many distinct people are involved, because visits count entries rather than accounts. And it says nothing about quality: a game can sit at the top of this board because it was enormous in 2019 and quiet now.
The directory publishes both columns separately for exactly this reason. If you want the underlying figures without the division, the original ratio explainer sets out how each column is collected, and the stored-versus-live gap piece covers what happens when the two clocks disagree.
One boundary condition worth stating: 4 of the 93 tracked titles are missing at least one statistic and are excluded from every row above rather than being filled in with an estimate. When those titles sync, the board reorders. It did so between November 1 and November 3, when the sample grew and every rank below shifted accordingly.