Quest audit against retail WildStar (2026-10-12)

A short public summary. The full audit is in docs/research/2026-10-12-quest-audit.md: 75 behaviours of the quest loop, each with its evidence, verdict and action.

Method

The quest loop (accept, objectives, progress, turn-in, rewards, experience) was compared with the retail record:

• Carbine's patch notes (kept verbatim on the Fandom wiki) and developer posts;
• the Fandom wiki's quest pages, which show the rewards of the launch build (319 pages for the two starting continents and their neighbours);
• the client's tables (Quest2, QuestObjective, XpPerLevel, Quest2Difficulty, Creature2Difficulty, UnitProperty2, LevelDifferentialAttribute, GameFormula) and the shipped UI (XPBar, QuestLog, QuestTracker, FloatText);
• player measurements (experience per kill) and the NexusForever source.

Each finding is labelled reference, derived or guess. Where retail left no evidence the value stays designed and is listed, with its basis, in docs/design/designed-values.md.

Results

What the evidence settled:

• Quest experience matches the client tables exactly, and the tables match retail. Four of the five launch-build rewards we could compare on the wiki are the table's values divided by 1.15 (or equal, below level 6), which is Carbine's "quests from level 6 to 49: +15% experience" (patch 02/03/2015). The level curve is the client's.
• Quest money is 20% to 50% lower on the launch wiki than the table gives (the exponent 1.70 fits where the table has 1.93, and the gap grows with the level). The table is kept; it is a build difference and not settled.
• Kill experience was on the wrong scale. The client's XpMultiplier column (Fodder 0.2, Minion 0.3, Grunt 0.5, Challenger 1.0, Superior 1.75, Prime 3.2) matches a Carbine developer's 2014 statement exactly, and players measured about 1% of the bar for a same-level kill; kill experience is now that.
• Fixed bugs: a 41st quest was accepted; exclusive quests (136 of the 427 the Exile 1-15 path offers) could both be taken; a crash during a turn-in could pay its rewards twice; creatures a script spawns (holdout waves, public event units) could not be hurt; several objectives (a unit pressed F on, an optional first step) could not complete.

Tests

Property tests over about 1,700 real quests (progress never exceeds its count, Achieved means every objective is done, turn-in is idempotent, abandoning restarts from nothing, the log limit holds), database tests that inject a failure at each write of a turn-in and then retry and relog, zone tests for each fix, and headless-client scenario runs for the Support Training quests.

The pace of day one

The headless day-one test (a new Exile character, quests, kills and loot from level 3) is the yardstick. Its pace was examined: see "The pace of day one" in the full audit. In short, the server's reward tables were right; a bug in script-spawned creatures cost the bot a quarter of its time and is fixed; the bot now earns about 640 experience a minute as an Esper (level 10 in 40 minutes), which puts level 15 at about 2 to 3 hours, in the range of retail play. The numbers of every run are in the day-one reports in this folder.