Why is my DPS lower in a slower Mythic+ key on Warcraft Logs?

The DPS column divides each player's total damage by the whole run's duration, from the first pull to the final boss. The denominator is the same for every player and it grows with every second the group spends walking, waiting or re-pulling. A slower run lowers everyone's DPS even if nobody played worse.

What is the difference between Parse % and Key % on Warcraft Logs?

Both are DPS percentiles for your spec in that dungeon. Parse % pools every key level, so at high keys it barely separates players. Key % is bracketed to your key level and is the one worth reading.

Which Mythic+ numbers are fair to compare between two runs of different length?

Damage per enemy forces point, boss-only DPS, and DPS over each player's own active time. All three take the run's total length out of the denominator, so a slow run does not deflate them.

Home
Guides
Mythic+

Why Your Mythic+ DPS Looks Lower in a Slow Key

September 5, 2026 · 6 min read

You finish a +13 two minutes over the timer. You open the log. Your DPS is 15% below your parse in the same dungeon last week, and so is everyone else's. Nobody played worse. What changed is the number Warcraft Logs is dividing by.

This guide explains what the Mythic+ DPS column actually measures, why Parse % stops meaning much at high keys, and which numbers survive a slow run so you can compare two keys honestly.

The DPS Column Divides by the Whole Run

On a raid boss, DPS is damage divided by the length of the fight. In a Mythic+ run, Warcraft Logs uses the same idea for the whole dungeon: each player's total damage divided by the time from the first pull to the last boss dying. That denominator is identical for every player in the group.

It also includes everything that is not combat. Walking between packs, waiting for a cooldown, the thirty seconds after a wipe, the pull someone pulled by accident and the group ran from. All of that is in the denominator and none of it is in the numerator. A run that takes 20% longer produces DPS numbers that are roughly 20% lower for all five players, before any of them presses a button differently.

What we measured

Two public Altar of Fangs +13 logs, one timed at 23:49 and one depleted at 30:49. The DPS denominator in each log equalled the fight window exactly, 1393 seconds and 1828 seconds, and was the same for every player in that run. The "Active" column is shown next to it but is not what the DPS divides by.

The practical consequence: the DPS column cannot tell you whether a slower group did less damage or simply spent more time not fighting. Those are different problems with different fixes, and the number blends them.

Parse % and Key % Are Both DPS Percentiles

Warcraft Logs shows two percentages next to Mythic+ DPS. Both rank your DPS against other players of your spec in that dungeon. The difference is who you are ranked against.

Column	Compared against	What it is good for
Parse %	Every run of that dungeon at every key level	Very little at high keys. A +13 player is pooled with +2 runs, so almost everyone at +13 sits near the top of the pool.
Key %	Runs at your key level only	The honest ranking. Two Arcane Mages in the logs above both read Parse 80 despite a 34% DPS gap; their Key % read 18 and 91.

Read Key %. Parse % is a leftover from the raid layout and at the key levels most people care about it barely discriminates.

Both percentiles still inherit the whole-run denominator. A depleted key drags your Key % down with it, because you are ranked against groups that finished faster.

The Timer and the Log Are Different Clocks

One more trap. The keystone timer that decides whether you get +1, +2 or +3 is Blizzard's clock, and it starts when the key is activated. Warcraft Logs' fight window starts at the first combat event. In our two logs the game's clock ran 22 and 37 seconds longer than the log's window.

That means timer maths (how far over you were, whether +2 was in reach) must use the keystone time, while any per-second rate must use the fight window. Warcraft Logs shows both; the score it shows next to the run is Blizzard's own Mythic+ rating, passed through unchanged, and it drops by roughly two key levels' worth the moment a key goes over time.

Numbers That Survive a Slow Run

If you want to compare your run against a faster one, you need denominators that do not grow when the group stands still.

Damage per enemy forces point

Total damage divided by the enemy forces the group actually killed. Two groups clearing the same dungeon land within about 1% of each other on this number no matter how long they took, because the same mobs have the same health. If the number differs, somebody killed different things: overpulled, skipped, or fought the same pack twice.

Boss-only DPS

Bosses have fixed health and near-100% uptime for everyone, so boss damage is the cleanest read on raw throughput. Trash DPS is dominated by how many mobs were pulled at once, which is a routing decision, not a damage one.

DPS over active time

Each player's damage divided by the seconds that player was actually doing something. This removes the walking, but it can flatter a player who stood still a lot, so read it alongside the two above rather than instead of them.

The time budget

The most useful number is not a rate at all. Take the difference between the two runs' keystone times and split it: how much came from bosses, how much from trash, how much from time spent not in combat, how much from killing mobs the key did not need. That split says what to change. "Your DPS was 12% lower" does not.

Get that split for your own key

Keystone Delta takes two Warcraft Logs Mythic+ runs of the same dungeon, or finds a comparable run for yours, and lines them up pull by pull: route, pull sizes, boss damage, standing-still time and per-player output, all normalised so a slower key does not fake the numbers.

Open Keystone Delta

Summary

Warcraft Logs' Mythic+ DPS is damage divided by the whole run, walking included, and it is the same denominator for every player. A slow run lowers everyone's number. Parse % pools every key level and is nearly meaningless at high keys; Key % is the one to read, and it inherits the same problem. To compare two runs fairly, use damage per enemy forces point, boss-only DPS, and a time budget that says where the minutes actually went.