Project Lore

Love 'em or hate 'em, rogues have got to be sneaky. It's a good part of what defines the class, allows them, as leather-wearers, to deal melee dps without instant death, and sets them apart from other DPSers. So it's no huge surprise that Vanish, one of the key skills that rounds out that specialty, has long been debated because it's, as some players would say, "broken."

Just how is it or isn't it broken? Different players of different classes will give you varying answers. But the debate revolves around players/mobs/pets still being able to attack a rogue who has vanished, and a number of AoE/DoT effects that will immediately pull a vanished rogue back out of stealth. Personally, my main is a rogue, so I tend toward the side that vanish should allow a little bit of lenience to allow some time to get away, heal or regenerate some energy. But I certainly don't want it to be overpowered, either, because then it will inevitably get a nerf.

Enter Patch 3.3 on the PTR, and game developers announced hefty plans to unveil a Vanish fix. The hoopla began about a month ago, with blue poster Ghostcrawler announcing Blizzard's intentions on the WoW forums:

We have a change for Vanish in place for 3.3. You will get to try it out soon (tm). As promised, if it proves a significant buff to rogues, we may have to compensate elsewhere. Just because it hasn't worked as intended doesn't mean it will be balanced when it does.

Good news, right? A fix to vanish that may mean developers will have to reexamine its strength once it works as intended. Seems like a happy middle-ground to all sides. Not all agreed, and massive grumbles ensued from all sides. Ghostcrawler attempts his "dispel QQ" skill with another post early this month:

"Roguerage!

We need to make Vanish work as advertised. As I've said, if this results in a massive net buff for rogues, we will compensate in other ways. However, the intent isn't that it's another version of Cloak. We just want to compensate for server lag, missile travel time and the like. It's entirely possible 1 sec is too generous to accomplish the job and a smaller number will work. That's the kind of thing we will be evaluating through the PTR process.

Unless my understanding of the spell is completely wrong, Flare actually removes stealth, so it should work in any case. I will verify this part tomorrow"

That post went up the same day that the PTR became active. In the first version of the PTR patch notes, the fix was this: "Vanish: For the first second after this ability is used, neither Vanish nor Stealth can be broken by taking damage or being the victim of a hostile spell or ability." A week later, the change was updated and the updated notes reflected that the fix now would only be applied for a half-second after Vanish or Stealth. But even that small of a time frame is problematic, as Ghostcrawler detailed in another post last week:

We're not happy with the current Vanish fix on the PTR.

Even at 0.5 seconds, the rogue sometimes gets hit by a missile already in flight. Yet, even at 0.5 sec it's too easy to screw up using AOE to break the stealth, which is a totally legit counter in our minds. Therefore, making the immunity longer or shorter causes other problems. As I said earlier, the attempt was *not* to make Vanish into Cloak of Shadows Jr. We're not trying to offer a big PvP buff to rogues. We just want the ability to work as advertised. The intent of the 1 sec (then 0.5 sec) immunity was to provide a little padding so that the ability didn't fail in situations where it is supposed to succeed. I'm not sure it's doing that though, and it's causing other problems.

The way we want it to work: If you Vanish after a single targeted spell is cast, you're safe, even if the missile is still in the air. If an AE spell is cast on you, even if you just went into stealth, it should break you out.

We're working on a hopefully better fix now where we record the time that a missile leaves the caster and compare that to when the Vanish is activated. There still might be edge cases though that this doesn't fix, if for example the cast and the Vanish happen very close together and someone's client isn't updated quickly enough. TLDR: It's complicated.

For perspective, in patch 2.4 prior to LK, Vanish was more reliable in the situation I'm talking about above: used after a spell was cast would make you immune to the spell. It had some other bugs at the time, but all of those that we know about have been fixed. We have made many changes to the way spells work in the 3.X cycle, and one of them made the situation bad again. We need to get the conflict between missile travel time and Vanish resolved the way it used to work. The bad news is that since this is a code fix, it might not make it into 3.3. The good news is that because it is likely a server fix, it doesn't need to wait for a major patch.

I'm also going to stamp this with "no promises" since unreasonable players may start complaining if Vanish isn't fixed in exactly one week from now. :(

P.S. I used the word "spells" a lot above because to a developer, everything you do in WoW, even a melee attack, is a "spell." It isn't necessarily confined to what players think of as spells, such as Fireball.

And as promised, in the most recent PTR build, Vanish changes have been reverted and may not come at all in 3.3. Yet another example of why the PTR patch notes are never final. So just when will a vanish fix come? In standard Blizzard fashion, no date is provided but Ghostcrawler hints that the fix could happen anytime. What is certain: game mechanics are complicated, and Blizzard seems to be taking into account how a fix would potentially affect all classes, not just rogues. So what's a rogue to do in the meantime? I suggest iTZKooPA's essential vanish macro - that at least prevents you from doing anything that would pull you out of stealth.

Another certainty: there seems to be plenty of strong opinions regarding vanish. Some people say any fixes would cause it to be overpowered. Others think it's virtually useless as it is currently. Where do you stand on the issue?