I checked 19,258 player-fights: just under half your raid drinks nothing

In brief. Across 800 boss kills from 416 guilds — 19,258 individual player-fights — 47.7% of raiders did not drink a single potion. Not one. DPS are twice as likely to potion as healers or tanks. It collapses on farm content and climbs on progression. And the gap between hardcore guilds and semi-hardcore ones is the biggest split in the whole dataset: on Serpentshrine bosses it is 79.8% against 57.3%, which survives every control I could think to apply.

There is a particular kind of raid leader argument that goes "everyone needs to be using consumables" and is met with "I am, mostly". Nobody has the numbers, so it goes round again next week.

I have the numbers now. They are worse than I expected, and more interesting than "people are lazy".

What did I actually measure?

One request per boss kill, covering all 25 raiders at once. Warcraft Logs reports a potionUse count per player per fight, so a single call gets the whole raid.

The sample:

That per-guild cap matters. Without it a handful of guilds that log obsessively would carry the entire result, and I would be publishing a fact about them rather than about the game.

One thing to hold onto throughout: this is a floor, not a ceiling. A potion drunk in the two seconds before the pull lands outside the fight window and does not get counted. Pre-potting is real, especially for DPS opening on a boss, so the true rate is somewhat higher than every number below. It is not higher by twenty points — but it is higher.

How many raiders actually drink a potion?

52.3% use at least one. So 47.7% go the entire fight on nothing.

The average raider drinks 0.62 potions per boss kill. The shape here is the interesting bit: almost everyone who potions at all drinks exactly one. Fewer than one in ten drinks two. The population is not split into "optimisers" and "slackers" — it is split into people who drink their one potion and people who drink none.

Does it depend on the boss?

Enormously. This is the single biggest thing driving the numbers, and it is why any headline figure needs the context underneath it.

Per boss, sorted by how much people care:

Two of these deserve a moment.

High King Maulgar at 15.6% is not negligence. The median Maulgar kill in this data lasts thirty-five seconds. You do not burn a two-minute potion cooldown and a few gold on a fight that ends before the buff would.

The Chess Event is 0.0%, and that is the correct answer. Across 317 player-fights, not one person drank a potion during the Karazhan chess game, because you are not a person during the Karazhan chess game — you are controlling a chess piece. When a method returns exactly zero on the one encounter where zero is right, I feel a lot better about the other rows.

Which roles are drinking?

DPS, mostly. Within Serpentshrine, 67.9% of DPS drank something against 38.1% of healers and 33.3% of tanks.

Those are same-tier numbers, so nobody is being penalised for farming easier content:

The class table is the same story wearing different clothes. Across the whole sample: Mage 67.4%, Warrior 64.7%, Rogue 64.7%, Warlock 63.8%, Hunter 62.5% — then Shaman 45.7%, Paladin 44.6%, Priest 39.2%, Druid 36.3%. The bottom four are the classes most likely to be healing, and it is the role doing the work in that ranking, not the class.

Are healers just being lazy?

Partly, but there is a real mechanical reason too, and it deserves saying before anyone gets shouted at. Potions share one cooldown, and a healer's decision is genuinely different from a rogue's.

• A DPS potion is free upside. Haste or Destruction, drink it in the burn window, there is no downside beyond gold.
• A healer's potion is insurance. If mana held, the Super Mana Potion stays in the bag, and on a 107-second Hydross kill mana usually does hold. Not drinking it is sometimes correct play, not sloth.
• Tanks are in the same position with a different flavour: the potion competes with a healthstone and with just being healed.

But look at the shape across tiers. Healers who drank something: 12.6% in Gruul's Lair, 26.5% in Tempest Keep, 38.1% in Serpentshrine. Healers absolutely do potion more when the fight demands it, which means the low overall number is mostly "the fight did not demand it" rather than "healers do not carry potions".

Even so, three in five healers going through a Serpentshrine boss without a potion is more than mechanics explain. Some of that is people who genuinely do not have one in the bag.

Do hardcore guilds actually potion more?

Yes — and it is the biggest gap in the data. On Serpentshrine bosses, hardcore guilds are at 79.8% and semi-hardcore at 57.3%. Twenty-two points, on the same bosses.

I checked this against the obvious objection, which is that hardcore guilds raid harder content and harder content gets more potions. Splitting by tier kills that objection:

Read the Gruul row: on farm content, the gap vanishes completely and hardcore guilds are marginally the lowest. Nobody potions Gruul, including the people who potion everything else.

So this is not "hardcore players are more diligent human beings". It is a decision being made deliberately and collectively — this fight matters, drink up; this one does not, save your gold. The difference between the tiers is not effort. It is that somebody decided, and told everyone.

That is, incidentally, an extremely cheap thing to copy.

What can this data not tell me?

Which potion anyone drank, for a start. Take the following as the limits, because a data post that only lists its strengths is an advert.

• Not which potion. potionUse is a count. Identifying Haste versus Destruction versus Super Mana needs the cast event stream, which is a request per player per fight, and that key is shared with the live site.
• Not pre-pull drinks. As above: every number here is a floor.
• Not healthstones. The field reads zero across all 19,258 player-fights, which I do not believe for a second in a game with warlocks in it. That is a tracking gap, so I have left it out rather than announce that nobody uses healthstones.
• Heavily EU. 18,191 of the player-fights are EU against 1,067 US, because of which reports happen to be cached. The US figure of 46.4% is real but rests on a much thinner sample, and I would not build a strategy on the gap.
• Only guilds that log. Guilds that upload their raids are plausibly more prepared than guilds that do not, so 47.7% drinking nothing is probably the optimistic version of the true population number.
• No casual bucket. Casual guilds do not produce enough multi-kill reports to survive the sampling, so they are absent rather than zero.

What should I do with this?

Decide which fights are potion fights, say so out loud, and stop asking for "consumables" in general. The hardcore/semi gap is not a work-ethic gap. It is a communication gap.

1. Name the pulls. "Potions on Vashj and Kael, nothing on Gruul" is a rule people can follow. "Bring consumables" is a mood.
2. Stop asking for potions on trivial fights. Maulgar dies in thirty-five seconds. Asking for a potion there spends goodwill you need for the fight that matters.
3. Check who actually did it. Not to punish anybody — to find out whether your rule reached the raid. This is the one thing an audit of a night tells you instantly, along with the enchants, the gems and the flasks.
4. Expect healers to be lower, and be fine with that. A held mana potion on a short fight is correct play. A held mana potion on a six-minute progression attempt is not.

The short version

• 47.7% of raiders drank no potion at all on a boss kill. Of those who did, almost all drank exactly one.
• Content drives everything. Serpentshrine 60%, Gruul's Lair 22%, Karazhan Chess Event 0% — the last of which is correct.
• DPS 67.9%, healers 38.1%, tanks 33.3% within the same tier.
• Hardcore guilds hit 79.8% on Serpentshrine against semi-hardcore's 57.3% — but on farm content the gap disappears entirely.
• That gap is a decision that got communicated, not a personality difference, which makes it the cheapest thing on this list to copy.
• Every number here is a floor: pre-pull potions are invisible, and only logging guilds are in the sample.