Wynncraft's provinces are divided into territories that guilds capture and hold. Territory warfare has existed since the Guild Update of December 2014^[1], in two mechanical epochs.

The two war systems

In the original system (2014–2021), wars spawned waves of mobs on dedicated war servers and territories generated XP for their holders^[2]. Version 1.20 'Gavel Reborn' (20 January 2021)^[3] rebuilt it: attacking now costs guild emeralds and is resolved by defeating the territory's war tower; territories produce resources moved along trade routes, can be taxed (including a separate ally tax), and support a headquarters whose strength scales with connected holdings^[2][4]. Patch 1.20.3 (5 July 2021) added guild seasons — Season Rating earned by holding and attacking territories, with resets and rewards every season^[5][6] — which turned holding land into a scheduled competition.

In-game screenshot of a standard guild war arena, the instanced combat space a player enters when attacking a territory's war tower under the post-1.20 system.

Reading the exchange record

The territory_exchanges log has run continuously since 3 January 2018^[7]. Raw exchange volume is not evidence of war: two phenomena generate enormous exchange counts in peacetime.

• FFA hotspots — clusters that churn perpetually regardless of politics, such as Detlas, the Hive, Lava Lake, Factory Entrance and the Emerald Trail, joined in 2020–21 by the Silent Expanse^[8]. Fighting concentrated on these territories is noise, not campaign^[9].
• Ally territory-trading — from the 1.20 era, allied guilds mass-trade territories for XP and SR, producing near-symmetric exchange counts (Avicia and Emorians once traded roughly 5,000 captures each way while allied)^[10].

Annotated territory map of the Jungle FFA introduced in guild Season 17, published in Titan Times vol. 13 (22 March 2024): map squares labelled with the holding guild's tag - TAq, ANO, BFS, AVO - and numbered in attack order.^[11]

A real war instead shows spread and asymmetry: sustained fighting across 100+ territories with a low share on FFA hotspots. Territory-exchange figures in these articles therefore count captures outside the FFA hotspots, and bloc rosters inferred from co-belligerence are inference, never recorded fact^[12]. In the Goose roster episode, a map-inferred roster initially contradicted the written record^[13].