If you sort this directory by favourites, one title sits at the top by a margin of nearly ten million: Barry's Prison Run, with 105,251,144. The next three are 95,512,940, 67,677,551 and 31,801,277. So the single most bookmarked game here is not rivals, not brookhaven-rp, and not murder-mystery-2 — all of which have larger live audiences right now. It is a prison-run obstacle course with 8,531 people in it at the moment these figures were synced.
That gap is the interesting part, and it is worth taking apart properly rather than treating it as a curiosity. Favourites and concurrent players are both real, both measured the same way for every game in this directory, and they answer completely different questions. Understanding why they diverge on this one game tells you more about how to read any game's numbers than another ranking would.
| Game | Favourites | Playing now | Lifetime visits | Approval |
|---|---|---|---|---|
| barrys-prison-run-obby | 105,251,144 | 8,531 | 4,752,486,705 | 60.65% |
| rivals | 95,512,940 | 88,973 | 18,875,400,363 | 93.66% |
| metro-life-city-rp | 67,677,551 | 5,324 | 2,387,183,422 | 86.77% |
| driving-empire | 31,801,277 | 17,185 | 3,376,634,494 | 93.00% |
| brookhaven-rp | 30,607,426 | 228,859 | 88,766,331,279 | 85.89% |
| murder-mystery-2 | 24,602,294 | 575,945 | 31,549,640,841 | 90.82% |
A favourite is a one-way, one-time, permanent action. A player clicks a button once and the count goes up by one; it does not go down when they stop playing, and it does not get withdrawn when they lose interest. Over a long enough period a favourites count therefore accumulates like sediment. It is a record of how many people, at some point, wanted a fast way back into this game.
Concurrent players is the opposite kind of measurement. It is a snapshot. It tells you how many sessions were open at the instant the directory synced. It has no memory, and it goes down as easily as it goes up. A game can hold an enormous favourites total and a small live population at the same time without any contradiction, because the first number is about history and the second is about right now.
For Barry's Prison Run the derived ratio makes the point cleanly. Divide favourites by the number of people playing and you get 12337 favourites for every person currently in a server. rivals, by the same arithmetic, carries 1074. murder-mystery-2, which has the second-largest live population in the whole directory at 575,945, manages 43. Those three numbers are not competing measures of quality — they are three different shapes of audience.
The third column refuses to settle the argument either. Barry's Prison Run has 4,752,486,705 lifetime visits, which is substantial but only about a fifth of rivals's 18,875,400,363 and a small fraction of brookhaven-rp's 88,766,331,279. So the game with the most bookmarks has been entered fewer times than several games with fewer bookmarks. That is only possible if a large share of those favourites were added by people who then did not return very often, or were added during a period of much higher activity than the current one.
The visits-per-player ratio separates those two stories. For this game it is 557084 visits for every person playing right now. For murder-mystery-2 it is 54779, and for steal-an-egg, the directory leader on concurrency with 776,312 playing, it is 11572. A high ratio means a lot of accumulated history against a thin present — which is exactly the profile you would expect from a game whose peak was earlier than its bookmark count suggests.
Alongside the three headline statistics the directory also stores upvotes and downvotes, and Barry's Prison Run scores 407,208 against 264,181 — an approval rate of 60.65% across 671,389 votes. That is the second-lowest rate of any game in the 68-game synced set. For comparison, rivals sits at 93.66% on 11,565,202 votes, driving-empire at 93.00%, and metro-life-city-rp at 86.77%.
This matters because the favourites count and the vote count were collected from different people at different times, and they disagree. A large group bookmarked the game; a smaller group that voted was markedly less enthusiastic. Neither figure is wrong. What you are looking at is a game with a big, mostly silent, historically-acquired audience and a visibly divided active one.
One honest limitation: this directory exposes four genre pages that are live, and Barry's Prison Run is not listed on any of them. The obby page currently lists only 2 titles, and this is not among them, so any attempt to describe its genre here would be guesswork. If you want to browse the obstacle-course category that does exist, start from the obby genre page and treat this game as an adjacent case rather than a member.
Every figure used here is stored per game in this directory, so you can redo the arithmetic on any title you like without waiting for anyone to publish it. Take favourites, divide by concurrent players, and you have the density figure. Take lifetime visits, divide by concurrent players, and you have the history-per-present figure. Neither is displayed anywhere on the site; both are one division away from numbers that are.
Worth knowing what the result is sensitive to, because it is a ratio with a small denominator. Barry's Prison Run sits at 8,531 concurrent. If that figure were 17062 instead — a plausible swing for a game of this size between one hour and the next — the density would halve to 6169 without a single favourite being added or removed. The numerator is effectively frozen; the denominator moves constantly. That asymmetry is why a single density reading is a snapshot of a relationship, not a property of the game, and why it is worth recomputing rather than citing across weeks.
The same caution applies in reverse at the top of the concurrency table. steal-an-egg carries 8,403,315 favourites against 776,312 playing, a density of 11. If its live figure dropped by half tomorrow the density would roughly double, and nothing about the game would have changed. Read these ratios as descriptions of the moment the sync ran.
The practical lesson is to treat favourites as a bookmarking measure and nothing more. It is genuinely useful for one question — "has a large number of people wanted a shortcut back to this?" — and genuinely useless for "is this busy tonight?". Use concurrent players for the second question and lifetime visits for "how many people have ever tried it?".
If you want to see this gap computed across the whole directory rather than for a single game, the favourites-per-player breakdown ranks every synced title on the derived measure, and the raw favourites top ten shows how different the ordering looks before you divide by anything. For a contrasting single-game profile where the two measures agree rather than diverge, the Rivals numbers piece is the clean case. This game's own code and guide pages carry whatever is currently listed for it.