Kerulak gazes up at Ragnaros during 40-Man Raid work in Molten Core. |
Fine Tuning
Raiding in Vanilla was unlike anything I had participated in throughout my gaming career. Just getting over the initial hump of requiring 40 players to be online at once, and coordinating their efforts in real-time struck me as somewhat of a miracle. Only once fireballs were flying across the giant dungeons, felling great behemoths left and right, did we begin to realize that we would need to proactively find ways to improve, even if it meant a tiny little little bit each week. These marginal improvements that we'd apply from week-to-week would infamously come to be known as baby steps, which was our only real weapon against the competition: other raiding guilds vying for bragging rights and new raiding recruits -- some of whom could easily come from our own roster. It was my job as Guild Leader to provide a raiding environment that was both fun and successful. Other guilds were doing the same thing, so I had to find a way to do it better.In order to stand out amongst a sea of raiding guilds, Descendants of Draenor needed to collectively assess class role, ensuring that we would approach our raids with the highest level of professionalism possible. Other raiding guilds were failing by being extremely wasteful; they would send players into raids without knowing who should play which role, instead just playing “however they wanted”. By looking at some of the world's best raiding guilds (at the time) we discovered a fact early on. Each class in WoW had a specific role in raids, and in order to be effective, deviation from those predetermined roles wasn't an option. In many cases the choice was obvious: A Rogue, for example, is a class that leaps out from the shadows and quickly ends his foe's life with razor-sharp daggers or swords. In a raid, rogues play only one role: Melee DPS (Damage Per Second). Rogues fall into a "pure" category class, alongside Mages, Warlocks, and Hunters. They do one thing: kill bosses in the face until dead.
The remainder of classes in WoW, however, can play multiple roles. This is based on how they have their talent points applied (what we refer to as their Talent Specialization or "Spec"). A Shaman, for example, could be specced to hurl lightning bolts from a distance, making them a Ranged DPS class. Or, they could be specced to utilize two one-handed weapons up close, swinging them at their target with the fury of the winds, standing side-by-side with a Rogue. This style plants them firmly in the Melee DPS department. Alternately, they could spec a third way, remaining at the back of the raid, blanketing their friendly party members with a protective layer of spells which replenishes health and grants shields. This style of play falls into the Healing category. We refer to classes that can play multiple roles based on their talent specialization as "hybrids". During Vanilla, the hybrids of WoW were the Shamans (if Horde), Paladins (if Alliance), Druids, Warriors, and Priests.
Just Because You Can...
While multiple roles were available to the Shaman in a raiding situation, there was truly only one spec that was viable due to the way Blizzard had balanced each of the nine classes. The cards of fate revealed to me that Shamans were to play the role of a healer, nothing more. The raiders in Descendants of Draenor quickly assessed what their role was, and embraced it, and if they didn't, they would be replaced by players that did. To lead by example, I was the first Shaman to lay down my weapons, and take on the responsibilities of a healer. I began to spread this belief system across the raiders of Descendants of Draenor, so that they, too, would embrace their class's fated role. Many did without hesitation. Some, however, remained in denial. They had originally chosen their class because of the emotion it instilled in them when they saw the class tearing monsters apart upon the lands of Azeroth. Choosing a Druid, for example, may have spoken to certain players because of their ability to shapeshift, taking on the form of a Bear that was virtually un-killable, or perhaps to a Cat, entering stealth (like a Rogue) and sneaking up on a foe, tearing them limb-from-limb.
Being told that in order to perform at your maximum capacity, you were to stay in the back of the raid and cast healing spells on your party, came as an awful blow to some. Once provided with this revelation, many threw up their arms in disgust and walked away from raiding. Others bit their tongue and played their fated role as was expected of them...but they did not enjoy it nor embrace it. This produced a large subset of players that were often performing at 50% or 75% of their maximum efficiency, and was a sad result of Blizzard's design we had no control over. Flexibility in class role would not come until much later.
There were ten bosses in Molten Core. With our raiders now playing the role cards they had been dealt, if we were diligent, we could get the first three bosses done in two hours. I thought our speed was reasonable, but at the amount of bosses in Molten Core, the reality of making an entire clear in one evening seemed unfathomable. We only had a few hours each week to make our attempts, and because we employed such a wide variety of players across North America, managing multiple time-zones and keeping players focused and awake left us a small window of time to execute content. There had to be a way to increase our speed.
As Kerulak finalizes raid invites, Haribo announces healing assignments. |
The Moment of Clarity
One night, I stumbled across a video uploaded by a European WoW guild called The Axemen, who performed a full clear of Molten Core in 1 hour and 30 minutes. After watching the video with mouth agape, I had a moment of clarity. Immediately, I sent it to my Warrior raid-leader, Ater, telling him we needed to crank our clear times up. He pulled the video apart piece-by-piece and began discussing it with the other Class Officers. The Axemen employed a strategy known as "Chain-Pulling": the raid is in a constant state of pulling trash and killing it...as trash is about to die, tanks are already moving into position to pull the next pack. We began to train our raiders to expect this kind of chain-pulling so that they were not surprised when there was little time to break, sit, drink...breathe. We pulled monsters non-stop.In order to facilitate this chain-pulling, our healers (having a finite amount of mana) had to come up with new ways of maximizing their endurance. Previously, they could sit down and drink after each pack of trash or boss was killed. Our new chain-pulling methodology removed this as a viable option. As we pulled trash faster and faster, our healers slowly expunged their mana pools until wrung dry like a sponge. This reduction in raid healing cascaded into player deaths, which only served to slow us down further. My Priest officer, Haribo (named after the candy company) came up with the concept of "Healing Buddies" to remedy this situation.
Haribo's idea was simple, yet elegant. Every healer would be paired with another healer. As we began to clear trash on our way to each boss in the instance, one "buddy" from each pair would perform heals on the tanks and raid, while the other just followed along, doing nothing. Then, as each of the first "buddies" began to run low on mana, Haribo would spam out a “SWITCH!” macro in a channel he created (unsurprisingly named "dodhealers"). This would alert the healers to swap to the other buddy in each pair. Now, the new buddy would assume the role of healing the raid, while the first buddy followed along, slowly regaining mana. This saved us from having to constantly stop at various times between trash pulls while the healers drank and regained their precious mana. With Chain Pulling and Healing Buddies in full effect, we were able to increase our boss kill count from three to five in the first two hours, killing Lucifron, Magmadar, Gehennas, Garr and Baron Geddon. If it were a particularly exceptional night, we could even fold Shazzrah, a sixth boss, into that two-hour block.
Raids were moving noticeably faster. As with anything you do in life, the faster you go, the greater the chance of making a mistake. Some players simply did not adapt as quickly as we had hoped; either they had computer performance issues that held them back, or they simply lacked the reflexes to respond at a moment's notice to an emergency situation. Players would randomly find themselves face down in the dirt, killed because of a simple mistake. No amount of "better play from other raiders" could help offset this. We needed a contingency plan. Additional research revealed many hardcore raiding guilds were utilizing "out-of-combat battle resurrection". This surreptitious strategy called for certain players to keep a healthy distance behind the core raiders doing the work on the boss. By remaining far away from the action, they would kept themselves out-of-combat, which granted them access to spells normally disabled during the raid. One such spell was resurrection: the ability to bring a dead player back to life. Resurrect spells were disabled in-combat to add a degree of ownership to the raid; simply rezzing the dead over and over throughout the course of the fight meant instantly trivializing encounters, allowing players to sloppily die in fire with no repercussions. Theoretically, it was a valid guard-rail to empower players to be accountable as a team. In practice, unsolvable latency-related issues leading to lag-driven death were not considered for. Out-of-combat battle rezzes were a reasonable-if-slightly-unethetical way to combat infrastructure mechanics for which Blizzard provided no legitimate solution to.
We now not only had a structure in place to facilitate an efficient raiding system, we had proof that it worked: an entire 40-Man raid instance was cleared. We were filled with excitement and pride, and as we strode through Orgrimmar, displaying our weapons and armor from the instance, we began to turn more heads and garner additional interest in our guild. This came at a pivotal time, as we would soon discover (as other raiding guilds had), the difficulty of Molten Core and the next instance, Blackwing Lair, were as different as night and day.
We were about to get a wake-up call. One that would require that addition of a single Rogue that would forever change the course of history in Descendants of Draenor.
Raids were moving noticeably faster. As with anything you do in life, the faster you go, the greater the chance of making a mistake. Some players simply did not adapt as quickly as we had hoped; either they had computer performance issues that held them back, or they simply lacked the reflexes to respond at a moment's notice to an emergency situation. Players would randomly find themselves face down in the dirt, killed because of a simple mistake. No amount of "better play from other raiders" could help offset this. We needed a contingency plan. Additional research revealed many hardcore raiding guilds were utilizing "out-of-combat battle resurrection". This surreptitious strategy called for certain players to keep a healthy distance behind the core raiders doing the work on the boss. By remaining far away from the action, they would kept themselves out-of-combat, which granted them access to spells normally disabled during the raid. One such spell was resurrection: the ability to bring a dead player back to life. Resurrect spells were disabled in-combat to add a degree of ownership to the raid; simply rezzing the dead over and over throughout the course of the fight meant instantly trivializing encounters, allowing players to sloppily die in fire with no repercussions. Theoretically, it was a valid guard-rail to empower players to be accountable as a team. In practice, unsolvable latency-related issues leading to lag-driven death were not considered for. Out-of-combat battle rezzes were a reasonable-if-slightly-unethetical way to combat infrastructure mechanics for which Blizzard provided no legitimate solution to.
Descendants of Draenor clear their first raid in its entirety: Molten Core |
Bearing Fruit
Employing a delicate mix of improvements, either self-created or shamelessly stolen from other successful guilds, resulted in a dramatic increase in our efficiency. We were able to defeat Onyxia (our first official Internet Dragon!) on November 21st, 2005, and were soon farming her regularly for better weapons and armor. Meanwhile, we continued to clear content in Molten Core, faster and faster each week, pushing deeper into the instance, collecting up the materials necessary to craft fire resistance gear; equipment vitally necessary to withstand the fiery wrath of Ragnaros. Then, on the fateful evening of February 10th, 2006, Descendants of Draenor returned to Molten Core, stood before the mighty Elemental Lord of Fire, and delivered a final, fatal blow, sending him back into the depths of Blackrock Mountain. I almost went deaf with the victorious screams and cries of forty players calling out in digital unison, piped into my headset via our Ventrilo server.We now not only had a structure in place to facilitate an efficient raiding system, we had proof that it worked: an entire 40-Man raid instance was cleared. We were filled with excitement and pride, and as we strode through Orgrimmar, displaying our weapons and armor from the instance, we began to turn more heads and garner additional interest in our guild. This came at a pivotal time, as we would soon discover (as other raiding guilds had), the difficulty of Molten Core and the next instance, Blackwing Lair, were as different as night and day.
We were about to get a wake-up call. One that would require that addition of a single Rogue that would forever change the course of history in Descendants of Draenor.