The Warzone Board

Every season, the same question tears through Discord: do I stay, or do I go? And every season, players answer it half-blind — reading their own server's mood, guessing at the neighbors. But the game already told you everything. The moment Preparation Day starts, Last War freezes a number for every warzone in your group — its Transfer Points — and that number silently sorts all 128 servers into three fates before anyone moves an inch.

We recorded the board on server 493 the hour the July window opened, read all 128 states off the screen, and then did the thing the in-game view can't: we cross-referenced every server against the actual mover data — who arrived, who fled, and which way the power swung. Here's what the board was hiding.

Three Fates: Ghost Towns, Frontier, Fortresses

A warzone's own Transfer Points — the average power of its recently-active commanders — decide its gate. Two thresholds split the group into thirds, and each third plays a completely different game:

The verdict is brutal and clean: the fortresses won. The hardest servers to enter gained +35.5B as a group while the ghost towns bled −39.6B. If it feels like the strong servers just keep getting stronger — the board agrees with you.

Can a Dead Server Come Back? We Have the Receipts.

It's the most-asked question in every transfer thread, and it usually gets a shrug. Not here. We took all 32 fully-open ghost towns — the servers the game had already written off — and tracked the net power each one gained or lost. The result is a perfect coin flip: 16 revived, 16 kept dying.

The Fortresses Ate Everyone's Lunch

Here's the part that stings if you're free-to-play. The Limited-Open servers — the ones you basically can't get into — were the biggest winners of the entire event. As a class they pulled +35.5B net, and every one of the top gainers is anchored by an alliance north of 30 billion power getting even heavier.

Meanwhile the fortress wall wasn't perfect: server 504 cracked, shedding −5.90B despite its locked gate. High walls keep people out — they don't stop the people inside from leaving.

The Board Sets the Odds — Not the Result

Plot all 128 servers by their opening score against how they finished, and the trend is real but loose: correlation +0.25. A stronger opening tilts the table your way, but plenty of dice land the other direction. Coordination beats position.

A Second Set of Eyes: Independently Confirmed

Here's where it gets serious. Everything above is built on LastWarHub's top-200 power snapshots. But a commander on server #422 — HipTin of [FniX] — has been hand-tracking something completely different for months: the raw count of heavy hitters on every server in the group (players over 200m, 300m, 400m, 500m power), snapshot after snapshot, going all the way back to Season 5.

Two totally separate methods. One counts power pools; the other counts heads. When we lined them up server-by-server, they told the same story:

That's about as clean as community data gets. And HipTin's head-count makes the story tangible in a way power numbers can't:

Heavy-hitter counts © HipTin [FniX], "Servers 421–548" community tracker — shared with WARDAWGG and used with thanks. Two independent datasets, one conclusion.

Convoy Warfare: How the Smart Money Moved

Nobody good moves alone. The heaviest plays of the window were coordinated convoys — whole alliance cores relocating down a single route — and the sharp ones aimed at ground they could actually take: mostly semi-open frontier servers with real seats, not locked fortresses.

And the biggest single magnet was a gathering — [9Z6] pulled 68 accounts from 16 different servers into semi-open 449 (13.29B of power), building a new home from scratch. [GRRM] did the same into ghost-town 430 (59 accounts from 8 servers), which is exactly why 430 tops the revival chart. Convoys don't just move power — they manufacture it where they land.

Follow the Whales: Who Actually Moved

Convoys are alliances — but who are the actual people? Here we only count a move we can double-source: LastWarHub has to show the player leaving server A for server B this window, and LastRank's live roster has to still show them sitting on B. That intersection is 1,678 confirmed moves of the 6,742 LastWarHub reported — the high-confidence core, and the only players we'll name.

🏴 The recruiters — who pulled in the most confirmed muscle

Group every confirmed mover by the flag they fly now, and the hardest-recruiting alliances line up exactly with the convoys above — [9Z6], [GRRM] and [LioN] again:

🐋 The heaviest confirmed moves

The biggest guns who verifiably packed up — each one both reported by LastWarHub and sitting on the destination right now. Our mover report crowned Goldenfighter the #1 mover (532 → 545, now [UJA]); here's the rest of the top tier:

Method & a correction: a "confirmed move" appears in LastWarHub's July from→to table and on the destination server in LastRank's live roster (captured 2026-07-12). An earlier version of this section inferred moves from each player's in-game ID (which encodes an original home server) — that over-counted, since a player's home isn't the same as the server they left this window (it flagged long-settled residents like IBN and WanWan as movers). Fixed to double-sourced moves only — thanks to HipTin [FniX] for catching it. Data from LastWarHub & LastRank.

The Full Board — All 128, Search Yours

Here's the whole thing: every warzone's opening state, its seat quota, how full it got, its #1 alliance, and the arrivals, departures and net power swing we tracked. Search your server number or alliance tag, filter by fate, sort by whatever you care about. This is your scouting table for next window. Read the In column with care on collapsing servers — for a server whose roster emptied it's inflated by the top-200 cutoff dropping (phantom low-power "arrivals"); departures and net power are the reliable signal there.

Home Turf: Server 493 & [Kx0]

Our own board, for the record. 493 opened Limited-Open at 21.7K Transfer Points — a fortress — anchored by [Kx0] / The Kings Order at 33.5B, the #1 alliance on the server. When the dust settled we ran 38 arrivals against 37 departures for +0.48B net. The seat card read 22/40 27/30 3/3 0/0 the moment we captured it — the white Follower seats already more than half gone.

Straight From the Screen

No guesswork, no wiki scraping — this is read directly off the live in-game Transfer board on server 493 the hour the window opened.

Method & Sources

• The board (open state + seats + fills): read frame-by-frame off an OBS recording of the live in-game Transfer screen on server 493, captured 2026-07-11 when the Stage 1 window opened. All 128 warzones, verified one by one — every seat fill's denominators cross-checked against the server's open state (zero mismatches).
• The #1 alliance per server + confirmed player moves: pulled live from LastRank's public leaderboards — alliance power rankings and top-50 player rankings for all 128 servers. A player move is only named here if LastWarHub's July table shows the from→to and LastRank's live roster still shows them on the destination (double-sourced). Data from LastRank.
• Arrival counts to collapsing servers are unreliable — LastWarHub's "in" for a server whose roster emptied (e.g. 532) is inflated by the top-200 cutoff dropping and surfacing weak existing players as phantom "arrivals" (532's were all 63–135M). For dying servers we cite departures + net power only, never arrivals. Flagged in-game by HipTin [FniX].
• The movement (arrivals, departures, net power): WARDAWGG's recomputation of LastWarHub's crowd-collected top-200 mover data — 24,536 players matched before/after the window. Every per-server number here is joined from those two datasets by a script; nothing is hand-typed.
• Independent cross-check: heavy-hitter head-counts (players over 200m/300m/400m/500m per server, before and after transfer) come from HipTin [FniX]'s community "Servers 421–548" tracker, shared with WARDAWGG. It's an entirely separate collection method from LastWarHub's power snapshots, and the two agree at +0.79 correlation across all 128 servers — which is why we trust the story.
• Snapshot, not a movie: the seat fills are a single moment from the recording — seats kept filling after. "Net power" reflects tracked top-200 players only; moves out of the group are invisible, and a sliver of "moved power" is natural growth between the ~2-day snapshots.
• Fully offline & fair play: observation of our own screen plus public leaderboards. No game-API interception, no automation. Game © FirstFun; alliance data © LastRank; mover collection © LastWarHub and its contributors.

Corrections & Verification

We publish data, so we hold ourselves to data standards: every number on this page has been re-derived from the raw sources by an independent automated audit (July 13) — a second implementation recomputed all ~40 published quantities from the raw mover table, live rosters and community tracker, a blind second pass re-read a sample of the in-game screens (8/8 matched), and every external claim was checked against its live source. The audit runs again before any update to this page ships. Corrections, listed plainly:

• July 13: "Arrivals" to collapsed servers removed (they were a top-200 cutoff artifact, not real immigrants — 532 had none), and the original player-movement section was rebuilt on double-sourced moves only after its ID-based method mislabeled long-settled residents as movers. Both flagged in-game by HipTin [FniX] — thank you.
• July 13 (audit): fortress-anchor wording tightened — the top fortress gainers' #1 alliances range 31.9–35.4B, so "every one anchored by a 33B+ alliance" became "north of 30B". Semi-open gainer count corrected 44 → 45 (a +0.002B server had been rounded to zero).

Spot an error? Tell us — in-game (mail [KxO] Wardawgg 독언니 on #493) or via rank.wardawgg.com. Corrections get named credit, like this one.

Board captured July 11, 2026 · analysis published July 12, 2026 · verified & corrected July 13, 2026 · ← the aftermath: full mover report