METHODOLOGY & DATA

Analysis standards & data policy

We distinguish verified observations from what remains unknown.

How do I explore my logs?

Choose Raid or Mythic+ on the home page, then search for your realm and character. We check the specialization in recent public logs and find top players under similar conditions. You can also open a WCL report URL. Play both fights together, then select a character or enemy to inspect casts, incoming hits and health. Align the first cooldown use or the start of a dungeon pull. Detailed charts and values are available in the Combat Analysis tab.

The spell icons below the map show active buffs, remaining cooldowns and charges. Select an icon to compare both characters for that spell. Variable Putrefy charge consumption, enhanced combo points, actual Pulverize hits and free recasts are distinguished using logged talents and triggers. A readiness range means the initial state or trigger target is uncertain; it is shown separately from actual use.

Spell icons wrap onto additional rows. DPS for damage and support roles and HPS for healers use the five seconds preceding playback; totals begin at combat start. Pets are included; overhealing and shield absorption are excluded from healing. Tank armor uses observations from the previous three seconds. Physical and magic mitigation below the map combine observed armor, Versatility, buffs and selected talents multiplicatively. Actual Sentinel stacks, Imbued Shield, Sanctuary, Devotion Aura and Shared Resolve are included without counting the same effect twice. Blessing of Dusk multiplies missing health by 20% and rounds down to whole percentage points. Its one-second game update can differ from the observation time. Physical mitigation covers direct damage affected by armor. It does not equal actual mitigation for every attack, including blocks, avoidance, absorbs, bleeds, unknown effects or boss mechanics. Combined values are shown for verified Mythic+ and Heroic raid coefficients; missing observations are not guessed.

What does the replay show?

We use positions, health and combat events from the official API. The first position observed in each 0.2-second interval is kept and hidden after three seconds without an update. Enemy instances and map floors remain distinct. Both views use the same scale on the same map. WCL terrain, class and enemy images are positioned using game map coordinates; maps without terrain data show positions only. Both characters’ actual buff uptime and casts appear together for each spell. Cooldown readiness is reconstructed from talents used in that fight, observed reductions, resource spending and shared charges. Signature abilities for 40 specializations are grouped into tank, damage, healing and support roles. Current rules cover logs from August 11 through October 31, 2026. Initial charges use ranges; unverified talent combinations and trigger conditions remain unknown. This is not a complete reconstruction of every talent build. Aligning cooldowns does not align boss mechanics or dungeon progress.

Side-by-side comparison lets you select another participant in the same fight. It does not imply matching specialization, gear or role. Damage includes absorbed damage and identifiable pet damage. Healing excludes overhealing and shield absorption. Pre-pull buffs may be missing, and intervals without an observed removal are not treated as confirmed uptime.

Terrain comes from Warcraft Logs, map coordinates from Wago game data and basic spell information from Wowhead. Data reference date: October 31, 2026. An observed buff without a removal event has no confirmed remaining duration.

How are comparison candidates selected?

Raids match specialization, boss, difficulty and WCL ranking partition. Mythic+ matches specialization, dungeon, key level, affixes and season. We search the first page of global DPS rankings for an item-level difference of at most 15 and a fight-duration difference of at most 25%, then show up to three candidates whose player and specialization are verified in the original log.

Logs with higher DPS than yours are prioritized, then sorted by closest fight duration. We do not inspect every page of global rankings. Hotfixes, external buffs and strategies may differ even within the same WCL partition. Mythic+ completion time is distinct from the full collected timeline. Healer candidates are also currently selected by DPS, not healing performance.

Collection scope & limitations

We verify the supplied public report through the official API and collect events for the selected participant. The separate automatic collection workflow uses rankings and character queries. We do not claim to search all public logs or obtain a player’s complete history. We do not scrape web pages or bypass access restrictions.

Report totals appear only after all event pages for the selected fight are collected. Cooldown cast times are shown, but we do not calculate wasted time or optimal use counts because not every talent, reset and reduction is verified. Major abilities use localized labels; names without a verified translation remain as provided by WCL.

Real data & empty results

Raid and Mythic+ searches and automatic comparisons use real public WCL logs. Missing records or ineligible candidates produce an empty result, never automatic sample data. Existing development sample routes remain labeled as simulated and are not linked in the main menu.

After a public domain is configured, introduction, raid and Mythic+ landing pages can appear in search. Character search results, individual analyses and sample pages are excluded from search indexing.

Retention & revalidation

Directly opened reports are cached locally, with public access and changes checked whenever they are opened. Loss of public access deletes the cache. Recent reports include only those verified within 24 hours. This local cache currently has no automatic retention cleanup. Character searches may reuse results for 24 hours and rankings for one hour; original access and fight conditions are checked again for analysis.

The separate automatic collection database retains data for 30 days by default. In-progress reports become due for revalidation every 15 minutes, completed reports every 24 hours. Overdue real data is hidden until checked again. Reports that lose public access are deleted in the next job. Operators can also delete individual reports or characters.

Sources & service status

PeekAWoW is an unofficial fan project, not a Blizzard or Warcraft Logs service. There are no ads or payments. Commercial operation requires a separate review of current terms and commercial API access requirements.

Observed differences do not establish causation. We do not generate definitive judgments about a player’s intent or skill.