Monday, April 06, 2009
Green Armadillo from the Player vs. Developer blog is Ashamed of Victory, killing Sartharion on heroic with a horrible pickup raid group, and wondering if raiding is too easy if such a bad group can do heroic raids. Spinks even thinks that easy raids are killing hardcore guilds. But like so often, the whole discussion of "are raids too easy" is full of sweeping generalizations, which often totally fail to get to the core of the problem, and then propose solutions which only make matters worse.
As I said often enough, but guess I'll have to repeat every time, Naxxramas-10 has a near-perfect difficulty level for an entry-level raid dungeon. Making raids generally harder would destroy that, and get us back into the bad old days of TBC. Where the actual problem is, is the SPREAD of raid difficulties: There are not enough raid dungeons, and there isn't enough difficulty difference between normal 10-man and heroic 25-man raids, so there isn't enough CHOICE between easy and hard places to go.
So-called "heroic" 25-man raids are a particular problem. There is a persistent myth that guilds have a measurable degree of power, as in "my guild can kill Sartharion with 3 drakes up". People think that by looking at raid "progress", they can measure the power level of a guild. But that would only be true if everybody in a guild performed at exactly the same level, and doesn't take into account how raiding actually works for the various roles.
Tanks have a very individual and binary role, they either hold the undivided attention of their target and survive, or they don't. Healers work in a similar way when there are healing assignments: Either your target lives or dies; but then raid healing is often a more shared responsability. But damage dealing classes work on a collective sliding scale: The boss has a certain number of health points, so the SUM of all damage has to beat that number before the fight ends due to some enrage timer or other end condition. It doesn't matter who deals what damage, as long as the sum is high enough. Gevlon coined the phrase "good enough" for the idea that everything is fine as long as we just divide the sum of all damage needed by the number of dps players in the raid, and everyone is doing the required amount of damage, not less, but not more either. Of course real life doesn't work that way. One man's "good enough" is another man's "moron and slacker" (another Gevlon term).
But the problem with "heroic" raids is that necessarily due to there being a larger number of players, the spread in performance between the best and the worst is larger than in a 10-man raid. And as dps performance is measured by the game on a cumulative level, and not an individual level, it becomes possible for the people who perform better than "good enough" to carry those who perform worse than "good enough". Green Armadillo's PuG was lead by a freshly minted Death Knight dps in green and blue gear, who perfectly abused the system. While no being "good enough", he used his raid leader position to invite enough people to compensate for his deficencies, and the ninja'd the loot he needed, being the master looter. Obnoxious to the extreme, but not stupid; he cleverly used all the deficiencies of WoW's raid system and lack of social coherence to his advantage.
For guilds the matter gets even more complicated, because suddenly social aspects come into play. The spread between best and worst is not necessarily unwanted, because it enables the better player to gear the worst players up faster. Me squibbling with my guild on how to do Malygos boils down to a simple argument of whether we should close the gap between our collective performance and the performance needed to actually kill Malygos on heroic by A) practicing the fight by wiping with everybody or B) gearing up the people who currently operate below "good enough" level.
Some commenters on this blog fell into the easy trap of thinking that if my guild can't beat encounter X, then we all aren't "good enough" for it. But in reality we simply have the same problem as many medium-sized guilds: We gear up running one or two 10-man raids, I don't know any guild that manages to run 3 of them in parallel. And there is no fixed rotating roster, like most guilds we assign raid spots by selecting a reasonable class mix among those who show up, and some people just show up more often than others. So in the end we have a rather large spread between the best and the worst performers. Which, as long as we go to places where the sum of all that performance is "good enough", is no problem at all. As long as a guild succeeds, it doesn't really matter whether everybody performed at the same level, or whether some people were willing to perform better and "carry" others who performed worse. But if the spread becomes too large, which is more likely to happen in a "heroic" 25-man raid than in a normal 10-man, and the overall performance isn't sufficient, then it becomes more problematic. The cost of wiping, in terms of wasted time and gold, is carried evenly by everyone involved. But World of Warcraft is a game of diminishing returns, so it is much harder for somebody already performing well to further increase his damage output than it would be for somebody under the "good enough" level to get up to that level. Diminishing the spread between best and worst, in my opinion, not only relieves social strains from the guild, but also is more efficient than wiping together.
In summary I think the solution for 25-man raids is not as easy as simply making them harder by adding X% to the health and damage of the bosses. Instead we would need a different design for raid bosses in which there is more attention to the individual performance of everybody, and less collective measurement of overall performance with the possibility of wide spreads. Thaddius from Naxxramas comes to mind, or in one extreme case Razorgore in Blackwing Lair. Many guilds had problems with Razorgore, although collectively they performed well enough, but by having to split up the fight into 4 quarters, there was less possibility of spread, everybody had to perform "good enough" to succeed. That won't please everybody, but ultimately such design feels more "heroic" than using 25-man raids for easy gearing up.