What Umbra is
GRADES WITH RECEIPTS.
A Mythic+ performance grading tool built for two things: helping you see exactly where your play needs work, and helping you find groupmates who take the game as seriously as you do. Every grade drills down to the specific events that produced it: kicks, avoidable damage, cooldown presses.
Why we built it
EVERYONE'S TIME IS PRECIOUS.
If you're a player who wants to push keys, you already know the wall: you run a dungeon with people who are gearing up to parse, not to time, and three hours later you've learned nothing about your own play. The tools that exist tell you what people achieved, not how they play. Umbra tells you the second.
We built this to be a self-improvement mirror first, and a vetting tool second. Look yourself up, read the breakdown, and you'll know exactly which habits are holding your keys back. Look up someone you're inviting, and you'll know if their grade matches their promises.
What we provide
EVERYTHING LINKS TO EVIDENCE
A role-aware letter grade, S+ to F-
One composite number per character, computed from six weighted categories tailored to how tanks, healers, and DPS actually contribute. Updates automatically as you run keys.
A category breakdown you can argue with
See exactly how much of your grade came from interrupts vs. survivability vs. damage. Every category links back to the specific numbers it's scored against. No opaque composites.
A per-dungeon breakdown
Your overall grade hides which dungeons you're actually struggling with. We compute a separate grade per dungeon from just that dungeon's runs, so weaknesses are obvious at a glance.
A pull-by-pull recap of every run
Click any run and see what happened, pull by pull: the kicks you hit, the avoidable damage you ate, the pulls you died in. The whole dungeon in a 30-second read.
A cast-by-cast rotation tab
See exactly what you pressed in a run: the first 15 casts of your opener, a frequency table grouped into Rotation / Cooldowns / Utility, and a timeline per pull. Compare your own sequence against what your spec is supposed to do. Every curated class and spec (including Midnight's Devourer DH) has its own classification data.
A free addon that surfaces grades in-game
Tooltips in the world and in Group Finder show grades inline: yours and everyone else's. No signup, no account, no payment. Download, install, it just works.
Fresh data when you ask for it
Run a key, upload via Archon (the official WCL uploader; the addon handles combat logging), and look yourself up on the site. Your profile has a 'Refresh my profile' button that pulls your recent logs from Warcraft Logs on demand, capped at once an hour per character. The in-game addon ships a new bundled grade database once a day, so tooltips and LFG badges stay current without you doing anything.
How to use it
USE IT ALONGSIDE EVERYTHING ELSE.
Umbra doesn't replace IO. It doesn't replace item level. Each tool answers a different question, and the three together give you a signal no single one can:
Item level
What did they gear up to?
How much pressure their kit can apply. Necessary but not sufficient.
IO score
What have they actually completed?
Proven ceiling and consistency at key levels. Doesn't show how they got there.
Umbra grade
How do they play under the hood?
Kicks, positioning, cooldown usage, survivability. The difference between a timed and a depleted key.
A player with high IO + low Umbra grade is a carry-passenger. A player with low IO + high Umbra grade is someone ready to push, they just haven't been given the keys. A player with both is the groupmate you want. That's the information asymmetry we're trying to close.
On the gatekeeping question
A RATING CAN'T STAY HONEST BY BEING HIDDEN.
Raider.IO was called toxic when it launched. Players argued it would turn every LFG into a gated meritocracy. By 2022 it was the default way to vet groups, because transparent data is less toxic than rumor, reputation, and guess.
Any rating system can be misused. We can't stop that. What we can do is make the system itself as honest and as transparent as possible, so the misuse is obvious when it happens and the data actually helps more than it hurts.
The right stance isn't "no ratings." It's "ratings you can argue with." If you disagree with your Umbra grade, you can open your run breakdown, see every event that drove it, cross-check against the source WCL log, and tell us what's wrong. Rumors don't let you do that. Pugs don't let you do that.
We'd rather ship a grade you can disprove than a vibe you can't.
Our commitments, in code, not words
- check_circleYour grade is never hidden from you. Every player page shows the composite, the category breakdown, the per-dungeon grades, and the per-run pull events. You can always see what produced your number.
- check_circleNo pay-to-upgrade tiers. The addon is free. The website is free. There will not be a Pro plan that shows you extra digits of your score.
- check_circleNo selling user data. The only thing we ingest is combat logs you already publish to Warcraft Logs. We don't collect anything else, and we don't sell what we do have.
- check_circleDungeon ability data is sourced from live logs, not guesses. Every 'avoidable damage' ability in our scoring comes from a cross-log sampler that reads the top 20 speed-runs per dungeon each week. No curated list, no editorial favoritism. If top players take damage from it and good players avoid it, we track it.
- check_circleOpen roadmap, open tradeoffs. We ship changelogs and we write up the scoring changes we make, including the ones that hurt our own characters. Category weights and benchmarks change when the data proves the old ones wrong.
Principles
HOW WE BUILD A FAIR GRADE
Role-aware weighting
Tanks, healers, and DPS are scored on different categories at different weights. A tank isn't dragged down by DPS; a healer isn't penalized for not kicking a spell only Resto Shaman and Holy Paladin can interrupt.
Spec-aware benchmarks
Fury Warrior and Marksmanship Hunter have very different natural cast rates. We use spec-specific benchmarks instead of one universal curve, so a clean MM Hunter isn't flagged as lazy and a slacking Fury isn't given a free pass.
Talent-honest cooldown tracking
We only score CDs every player of the spec has access to. If you didn't pick a talent-gated ability (Tree of Life, Flourish, etc.), we don't score you on something you couldn't press.
Evidence-backed categories
Every score is tied to events in the combat log: specific cast IDs, specific damage breakdowns, specific deaths. When you disagree with a grade, you can inspect the raw per-run numbers and see exactly what drove it.
Why Umbra
WHAT MAKES US DIFFERENT
Receipts, not rankings
Gear score shows what you brought. IO shows what you completed. Umbra shows how you actually played: the kicks you hit, the avoidable damage you ate, the cooldowns you held. Every grade drills down to the specific events that drove it.
Fair across roles
Most rating systems quietly weight toward DPS output. Our tank and healer grades reflect tanking and healing: survivability and utility carry real weight, not the ceremonial 5% other tools give them.
Per-dungeon + pull-by-pull
You don't just get an overall letter. You get a grade per dungeon (see where you actually struggle) and a pull-by-pull breakdown per run (scan the whole dungeon in 30 seconds). Nobody else does this.
Built for M+, not raids
Raid parsing tools are mature. M+ evaluation isn't. We start from M+-native assumptions: key-level weighting in every category, timed rate as context not penalty, per-dungeon avoidable-ability lists sourced from live top logs.
Deeper reading
FULL METHODOLOGY →
Role weights, category-by-category explanations, the keystone weighting curve, and the universal timing stat. Every input that produces a composite grade, written up end-to-end.
Read the methodologyarrow_forwardHow Umbra grows
THE MORE OF YOU USE IT, THE BETTER IT GETS.
Every grade on wowumbra.gg exists because a player uploaded their combat logs via Archon. We can't scrape WCL for everyone, and we don't want to. The only way a tooltip fills in for a new player is someone looking them up on the site, or them running keys with the addon and Archon active. So the database grows at exactly the rate players are willing to be seen.
That makes this a network-effect tool. The more of your guild, your Discord, and your pug circle install Umbra, the more LFG applicants arrive with a grade already attached, and the more vetting becomes a glance instead of an interview. There is no ad budget. There is no marketing team. There is you, a free addon, and a free site.
Four ways you can help Umbra grow
- extensionInstall the addon. Run your keys with Archon uploading. Your grade populates within the hour and every LFG tooltip you touch gets one more graded name.
- shareShare your /umbra panel. When your raid or guild asks how a key went, drop a screenshot of the panel. A concrete grade and category breakdown lands harder than "it was fine."
- linkLink your profile when someone questions your play. The grade is designed to be argued with. Make it public and the conversation moves from opinions to evidence. If your grade is wrong, we want to hear it.
- groupTell one person who would care. A friend who pushes keys. A guildmate trying to climb. A pug that surprised you. This is the only marketing that actually works for a free addon.
Roadmap
HONEST GRADES, THEN COACHING.
The grade is the beginning, not the end. Our next phase adds an AI coach that reads your per-run pull breakdown and gives spec-specific, dungeon-specific advice: which kicks you missed, which defensives you hoarded, which mechanics killed you. We won't ship the coach until the grades it's built on are rock solid, because a confident wrong answer is worse than no answer.
Look up a playerarrow_forward