A directory of 85 games produces dozens of numbers, but only a handful actually explain anything. This piece takes the extremes — the largest, the busiest, the most saved, the quietest — and works through what each one does and does not tell you. Every figure below was read from the same snapshot, so the comparisons are internally consistent.
| Extreme | Game | Value |
|---|---|---|
| Largest all-time visits | Brookhaven RP | 89,056,490,507 |
| Most people playing now | Steal An Egg | 1,035,679 |
| Most favourites | BARRY'S PRISON RUN! (OBBY) | 105,265,479 |
| Fewest people playing now | MeepCity | 107 |
| Visits across all tracked games | directory total | 695,639,738,874 |
| People playing across all tracked games | directory total | 4,699,280 |
| Favourites across all tracked games | directory total | 836,606,540 |
| Approval across all tracked games | directory total | 89.61% |
89,056,490,507 visits. That is the single biggest cumulative figure in the directory, and it belongs to a social roleplay game rather than a competitive one — which is the first lesson about visit counts. Visits accumulate with every entry, and a game people enter dozens of times a day for years will outrun a game people play once and love.
Note that Brookhaven RP also holds 435,046 concurrent players right now. It is not a relic; the history and the present are both enormous. That combination is rare and it is the reason this game sits at the top of almost every ranking you can build from this data.
Concentration is the sharper point. The five most-visited games account for 42.6 per cent of all visits across the 85 tracked titles. Visit counts are not distributed; they are dominated.
1,035,679 concurrent players on 9,950,444,946 all-time visits. This is the number that changes fastest and the one most worth checking directly rather than remembering. Concurrency is a single moment, and a game that leads it tonight may sit fifth tomorrow afternoon.
The ten busiest games by concurrent players hold 67.4 per cent of everyone currently playing across the directory. That is a slightly flatter distribution than visits, which makes sense: a game can own an enormous historical visit count while its live population shrinks, but it cannot own the live board without people actually showing up.
105,265,479 favourites, which is 12.6 per cent of all favourites in the directory — from a single game. That game holds 12,886 concurrent players on 4,770,548,047 all-time visits.
This is the most interesting gap in the whole dataset. Favourites are cumulative and never decrease, so they reward age and they reward a game that people used to play. A title can lead this column while sitting well down the concurrency board, and that combination is a specific, readable signal: enormous historical attachment, modest present occupancy. the MeepCity case study is the extreme version of this pattern.
107 concurrent players against 16,279,228,236 all-time visits. The quietest game in the directory is also one with a very large history, which is exactly the pairing the visits-per-player ratio exists to expose.
Do not read this as failure. A game at the bottom of the concurrency board at one moment may be at the bottom because of time zone, because of a school night, or because its audience is small and devoted. What it does tell you is concrete: if you join right now, you will very likely be joining an empty server.
695,639,738,874 visits and 4,699,280 concurrent players across 85 games, with 836,606,540 favourites and an approval rate of 89.61 per cent. Those four aggregates are the denominator behind every ranking on this site, and they belong in your head before any individual figure.
The approval rate in particular is easy to skip and hard to replace. 232,714,569 upvotes against 26,983,439 downvotes is the closest thing to a platform-wide peer review that exists publicly, and it is the one aggregate that does not simply grow with time. Visits, favourites and total players all accumulate; approval can fall.
Of the four single-game extremes, the favourites figure is the one most likely to be misread, and it is worth saying why in detail. Visits and concurrency both describe behaviour: someone entered, someone is present. Favourites describe intent, recorded once and never revised. A player who favourited a game in one year and never returned still counts, and there is no mechanism by which that count goes down.
That asymmetry means the favourites leaderboard is partly a measure of age. It will always favour a game that has been available longer, all else equal, and it will always overstate current attachment relative to the concurrency column. This is not a flaw in the data — it is what the field records — but any ranking built on it needs the caveat attached rather than buried.
The practical test is simple. Take any game near the top of the favourites column and look at its concurrent figure. If the two are wildly out of proportion, you are looking at a game with a large past. That is a legitimate thing to be interested in, but it is not the same thing as a game with a large present, and the directory gives you both columns precisely so you can tell the difference.
The sample grew from 79 games to 85. That single fact reorders every column at once, because six additional games enter the totals, the medians and the quartile boundaries. the October 27 version of this piece was built on the smaller sample; the original five-number version on a smaller one still. Comparing the three shows the same extremes holding their positions while every absolute value moved — which is the honest way to read any ranking built on a growing dataset.
8 games in the directory have not yet returned a complete statistics set, so all of the above is drawn from the 85 that have. Nothing here is estimated; anything not returned by the source is simply absent.
the live player board is where to watch the concurrency extreme move in real time; the simulator genre page shows the category containing the busiest simulator; the Blox Fruits code list is the largest code list in the directory and a useful reference point for what a well-populated list actually looks like.
