How to Actually Evaluate a Healer

Healing is the hardest role to evaluate from a log, and the healing meter actively makes it worse. A healer who tops the healing chart is often just the healer who threw the most inefficient healing into a group that did not need it. If you have ever been told "your healing was low" after a clean kill where nobody died, you have run into this problem. Here is a framework that actually works.

Why raw healing lies

Effective healing only counts health actually restored. The moment a target is full, every additional point you pour into them is overhealing — wasted output that still inflates the meter if you are not careful which number you read. Two healers can do the exact same useful work; the one who spammed expensive heals into full health bars "wins" the meter while having played worse, because they wasted mana and global cooldowns that a real emergency would have needed.

The first correction is simple: never evaluate a healer on raw healing. Expand the healer on the Healing tab and look at the overhealing percentage per spell. High overhealing on reactive spells is a sign of panic healing or poor timing; some overheal on shields and heal-over-time effects is normal and fine.

Start from damage taken, not healing done

The single best reframing: a healer's job is to keep people alive against incoming damage, so start from the Damage Taken tab, not Healing Done. Was the damage the group took avoidable or forced? If the raid is taking enormous avoidable damage, no healing chart is "good" — the team is bleeding survivability and the healers are just treading water. If damage taken is low and controlled and nobody died, the healers did their job well even if the meter looks modest. Healing is downstream of damage taken; read them in that order.

The metrics that actually matter

Deaths. Did people die, and were those deaths healable? Cross-reference with the death log. A one-shot is not a healing failure. A slow ramp death with the healer's cooldowns unused is.

Cooldown usage. Healing cooldowns are most of a healer's value. Check the timeline: were major raid cooldowns used on the big predictable damage events, and were enough casts gotten out of them across the fight? A healer who held a major cooldown the entire fight and never pressed it is the equivalent of a damage dealer who never used their burst.

Mana and efficiency. Did the healer end the fight with a reasonable mana curve, or did they run dry early and contribute nothing for the last ninety seconds? Efficiency is invisible on the meter and decisive in long fights.

Damage contribution. On many fights healers are expected to also deal damage when healing is light. A healer who kept everyone topped while contributing meaningful damage in the calm windows did more than one who only healed.

Healer parses and why they are noisy

Healer percentiles exist, but they are inherently noisier than damage parses because healing demand is shared and variable. If your co-healers absorb all the damage events, there is nothing left for you to heal and your parse craters through no fault of your own. This is why healer parses are best read as a loose trend over many pulls, never as a verdict on a single fight, and why comparing two healers in the same group by parse alone is meaningless. See What Counts as a Good Parse for why population matters here even more than for damage.

A practical evaluation checklist

To assess a healer from a log: confirm nobody died to healable damage, check overhealing is not wildly high on reactive spells, verify major cooldowns were used on the big damage windows, look at the mana curve for the long fights, and only then glance at the parse as a tiebreaker. Do that and you will judge healers fairly, which is more than the healing meter has ever done.