pt-BR surge and a handful of unstable breakouts

Two kinds of movers dominated the week: steady, multi-day risers you should watch, and very large jumps that did not display day-to-day growth or that the sheet marks as absent by the end. BAD TOWN! is the clearest trajectory: it rose from an average of 1 last week to 203 this week (+201), grew on 5/6 days, and shows an unmistakable upward daily shape (3→3→66→192→319→398→438). The server’s rank jump (from 29,048 to 190) identifies it as a genuine breakout rather than a marginal noise event. LUX TOWN and Streets Of Atlanta are the other examples of growth that held across most days: LUX TOWN averaged 273 (prev 158), grew on 4/6 days and maintained daily levels in the 238–299 range; Streets Of Atlanta averaged 248 (prev 80) and grew on 5/6 days with a steady 63→311 spread. Those servers have the behavioural signature of sustainable traction — repeated gains on multiple days rather than a single spike. By contrast, a set of big-number entries in the grew list are ambiguous. OASIS ROLEPLAY went from 0 to an avg 236 and grew on 5/6 days (142→287), but the fact sheet also lists it as last seen 2026-05-25 with an average of 247 in the went-dark table. Narco City appears the same way: it’s recorded as averaging 660 this week (prev 325), growing on 4/6 days, yet also appears in the went-dark list as last seen 2026-05-23 and averaging 647. Those contradictions make the apparent gains hard to trust until visibility is consistent across the next reporting period. Smaller but notable rises — ProdigyRP Public West 4.0 (+122, grew on 3/6 days) and ProdigyRP Allowlist 4.0 (+112, grew on 2/6 days) — show partial holding: both increased their weekly averages but only on a minority of days, so they look like recovering communities rather than runaway winners. notify-probe-withchild is another anomaly: the sheet lists a large week-on-week increase to avg 187 (prev 7) while showing a 0/6 figure for growth days and a two-point daily sample (309→167); that pattern reads as a data artefact unless it repeats next week. The practical distinction this week is clear: sustained rises are the ones that grew on most days; large single jumps or internally inconsistent records are not yet reliable.

The rank table this week is dramatic at the top and bottom. Nexord Life climbed to rank 1,202 from 32,742 — a gain of 31,540 positions — even though its reported average is modest (avg 22); that is the steepest climb on the sheet. BAD TOWN! is the second major climber, moving to rank 190 from 29,048 (up 28,858), matching its week-long upward daily pattern. Other large upward moves include PH PvP (rank 3,993 from 31,692, up 27,699) and Greek Fear RolePlay S06 (rank 747 from 27,067, up 26,320). These are not all large communities by average concurrency, but the scale of their rank jumps marks a sudden reappearance in the public index and — for several — a sustained presence across multiple days. Downwards movement is equally sharp: Destiny 18+ fell to rank 32,375 from 549 (down 31,826), and KUALA ELITE ROLEPLAY fell to rank 32,655 from 2,009 (down 30,646). Several servers with previously visible ranks — Balkan Hub, Kurdlabs, Slimeto Rz and others — sit much lower now, in some cases accompanied by large declines in average players. Rank movement this week therefore splits into two stories: big upward rediscoveries or re-indexing of small servers, and catastrophic falls for a handful of previously visible communities. For anyone watching discovery and referral traffic, the upward movers are worth tracking to see if their presence stabilises, while the big fallers are candidates for further investigation into outages or permanent closures.

New arrivals that actually survived the week are small but meaningful: ten servers first seen this week were still present on the last day. BRASILÂNDIA! was first seen 2026-05-26 and is recorded as present 6/7 days with avg 182 and peak 394; BERLIM ROLEPLAY (first seen 2026-05-28) was present 4/7 days with avg 166 and peak 283; Aspen University (first seen 2026-05-29) is present 3/7 days with avg 114 and peak 438. The new cohort is heavily tilted toward Portuguese-language projects (BRASILÂNDIA!, BERLIM ROLEPLAY, Aspen University, several others) and a handful in Spanish, Chinese and Turkish — a reminder that regional launches are the dominant source of fresh activity. The week also recorded ten servers that went dark after previously having measurable audiences. Notable disappearances in the went-dark table include Narco City (last seen 2026-05-23, was averaging 647 players), three entries labelled WE THE BEST FAKE PLAYER PROVIDER! with last seen 2026-05-25 (averaging 524, 371 and 371), and OASIS ROLEPLAY (last seen 2026-05-25, was averaging 247). Two further examples are IBIZA ROLEPLAY v3.2 (last seen 2026-05-26, averaging 194) and Lucid District Roleplay (last seen 2026-05-18, averaging 129). The fact sheet contains internal contradictions: BRASILÂNDIA!, OASIS ROLEPLAY and Narco City are reported among this week’s growing or newly seen servers and also appear in the went-dark list with specific last-seen dates. Those entries need reconciliation before anyone can treat their week-on-week numbers as solid evidence of retention or collapse; as presented, several high-growth entries are flagged as gone within the same week.

For owners and community managers the practical test next week is simple: did the multi-day growth pattern replicate? Servers that grew on 4–5 of the six tracked days (BAD TOWN!, Streets Of Atlanta, LUX TOWN) deserve monitoring for a second week of consistent daily gains — that would convert a breakout into a confirmed trajectory. Anything that posted large averages but grew on 0–2 days, or that appears in the went-dark list, should be treated as unconfirmed until it remains visible across the next reporting window. Watch three specific signals: (1) 'grew on N/6 days' continuing above 3/6; (2) rank movement holding rather than reversing; (3) presence on the final day of next week’s sheet. Also track regional share: pt-BR at 19.6% (25,802 players) is now the largest locale slice and a second week at that level would indicate a material shift in where attention is concentrated. Owners should remember the weekday shape — Wednesday (134,058) remains the low point and Sunday (162,545) the top — and test retention strategies across midweek rather than relying on weekend spikes. Finally, the dataset anomalies this week underline an operational point: a single large nightly peak is not proof of health; repeated multi-day growth and presence on the index are.