How WoW Classic Guilds Actually Run Loot: Data from 33 Guilds

Between June 1, 2026 and August 31, 2026, 33 active guilds used LootList+ to manage 454 raid events and 5,432 loot awards across 451 raiders with approved lists. We looked at aggregated, anonymized activity to see how these guilds balance wishlist rank, attendance, bad-luck protection, and officer judgment. No player or guild names are included in the dataset.

This is product usage data, not a survey of every WoW guild. It shows how guilds using a transparent, list-based system behave in practice.

Raiders keep long, ranked wishlists, not short top-five picks

The median approved loot list held 18.0 items, measured across 581 approved lists in the window.

If you are used to thinking of a raider's list as "their five or six BiS items," this is bigger than that. A typical approved list ranks well past the obvious wishlist items, which means loot decisions for an officer are rarely a two-item choice. Knowing the fuller list matters for items well outside anyone's top pick.

This counts only the items still on a list at the end of the window, not everything a raider ever typed in. A raider who ranked forty items and later trimmed the list down keeps the length they ended up with, not the length they started with, so this number reflects a maintained list, not a first draft.

Most guilds don't leave attendance scoring on the default settings

84.8% of the 33 active guilds measured had changed at least one attendance-scoring setting away from what LootList+ ships by default.

Attendance scoring is on for every guild out of the box, so the interesting question isn't whether guilds track attendance, it's whether the default weighting fits how a specific guild actually runs raids. Most guilds we measured went in and adjusted it, which suggests the shipped defaults are a reasonable starting point for a new guild, not a setting most officers should assume is already tuned for them.

This does not measure whether a guild tracks attendance. Attendance scoring is on by default for every guild, with a nonzero bonus, so a simple presence check would say nearly everyone tracks it and would tell you nothing useful. What this measures instead is whether a guild's attendance configuration differs from the shipped defaults on at least one setting, which is a narrower and more honest question about active tuning, not passive presence.

Bad-luck protection is common, but far from universal

45.5% of the 33 active guilds measured had bad-luck protection turned on.

Roughly half of active guilds run without bad-luck protection at all, so if your guild has been debating whether to turn it on, you are not choosing between "everyone does this" and "no one does this." Either choice puts you alongside a large, normal group of other guilds.

This only tells you whether the setting is switched on, not how any guild has it tuned, and it says nothing about how often bad-luck protection actually changed who won an item during the window. A guild with the setting on and a guild with it off could both be running loot fairly; this number describes a configuration choice, not an outcome.

About three in ten drops go to someone whose list already had it at the top

29.5% of the 2,298 awarded items with a determinable prior rank went to a raider whose list already ranked that item in the top priority bracket.

A meaningful share of loot decisions land exactly where the list said they should: at the top. That is a useful gut-check for an officer weighing a close call between two raiders. It does not mean every award goes to the top of someone's list, so this number is a baseline for "how often does the list agree with the outcome," not a claim that the system always hands the item to the top-ranked raider.

This share only covers awards where the winner had a usable snapshot of their list from before the raid, and where the awarded item could be found on that snapshot with a determinable rank. An award without a usable prior snapshot or without a determinable rank is left out of both the numerator and the denominator here, it is not counted as a miss. "Top priority bracket" is a fixed rank range built into LootList+ itself, the same for every guild; it is not a setting any guild configures.

Methodology

This report uses a fixed calendar window, June 1, 2026 through August 31, 2026, not a window that moves forward with the calendar. The dates are locked in place, and the data page states when its numbers were generated. A list that an officer reviews again after the window closes moves out of the window under the definition above, so a fresh run reflects the data as it stands at run time.

A guild counts as active in this window if it had at least one raid event that was not marked skipped, or at least one loot award, or at least five approved loot lists. A raid marked skipped in the scheduler does not count toward activity. A loot list counts as approved as of whichever timestamp exists: an officer's review, or, if no review was ever logged, the raider's own submission time.

Raiders are counted as distinct characters holding an approved loot list in the window. This is a character count, not a person count: a player who raids on two characters, each with an approved list, is counted twice.

Where a guild had qualifying activity in more than one expansion during the window, it is counted in every expansion it was active in, not just one. That means an expansion-by-expansion breakdown can add up to more than the whole, by design, and the report says so wherever such a breakdown appears.

Every number in this report represents at least 10 guilds. Where a segment would fall under that floor on its own, it is folded into an "Other" row rather than published on its own, so no individual guild can be identified by process of elimination.

Percentages and medians in this report are rounded to one decimal place, rounding half up.

Every number on this page comes from a saved SQL query committed to our GitHub repository. You can read the exact queries, and run them yourself, in the saved-queries directory: https://github.com/alexandermayes/loot-list-plus/tree/main/scripts/analytics/queries/wow-classic-loot-systems-2026.

This report does not publish four measures the original plan recommended. A breakdown of active guilds by expansion is withheld because even after merging small segments into an "Other" row, that merged row itself falls under our guild-privacy floor, so it cannot be published without risking identifying a specific guild. Median time from guild creation to a qualified setup, and median time from a qualified setup to activation, are both withheld because the timestamps needed to measure them only started being recorded shortly before this window closed, leaving almost no guild in the window with a reliable timestamp for either milestone. A self-reported measure of officer time saved per week is not published because no survey of that kind has ever been run; the original plan calls for self-reported figures to be kept separate from behavioral data, and we have none to report. No other candidate metric was declined: every measure that could be computed accurately and safely within this window was published above.

• Distribution of expansions active guilds raided in: withheld: after merging every segment under the 10-guild privacy floor into an "Other" bucket, that merged bucket itself holds fewer than 10 guilds, so no segment here can be published without risking identifying a specific guild
• Median time from guild creation to qualified setup: guild_funnel_milestones shipped via migration 20260827000000 on 2026-08-27, four days before this report's window closes (2026-08-31) and roughly three months after it opens (2026-06-01). The evaluator (utils/analytics/funnel.ts) only stamps qualified_at with the time it happens to re-run and observe the condition newly true, not the historical time the condition became true, so any guild that qualified before 2026-08-27 carries a qualified_at timestamp reflecting an incidental later mutation rather than the true milestone date. Measured: only 2 of the 33 active guilds in this window were created on or after 2026-08-27 (funnel-cohort-coverage), well below the 10-guild floor, so even a cohort restricted to post-instrumentation guilds could not be published.
• Median time from qualified setup to activation: guild_funnel_milestones shipped via migration 20260827000000 on 2026-08-27, four days before this report's window closes (2026-08-31) and roughly three months after it opens (2026-06-01). The evaluator (utils/analytics/funnel.ts) only stamps both qualified_at and activated_at with the time it happens to re-run and observe each condition newly true, not the historical time each condition became true, so any guild that qualified or activated before 2026-08-27 carries timestamps reflecting an incidental later mutation rather than the true milestone dates. Measured: only 2 of the 33 active guilds in this window were created on or after 2026-08-27 (funnel-cohort-coverage), well below the 10-guild floor, so even a cohort restricted to post-instrumentation guilds could not be published.
• Self-reported weekly officer time before and after LootList+: No survey instrument exists in this product, no self-reported officer-time data has ever been collected, and the sprint plan requires self-reported figures to be reported separately from behavioral data, so there is nothing to report.

Get the data