Heroic Lich King and Difficulty in Raiding in WoW

The race continues among the hardcore guilds, with Paragon having cleared every boss but the Lich King in Heroic 25 ICC.  Given the time it took to do that, and the limited attempts, I have a sneaking suspicion Paragon is running a bit low for actually beating Arthas this go round.  I’m also not sure how many of these are alts; I’d guess some attempts are run by alt crews to learn the fights, then the real raid comes in.  But then…who knows?

Meanwhile, in 10-man strict land, From Chaos downed the LK in 10-man normal mode today.  This isn’t quite as much of an accomplishment as it might have been, as the encounter took a few hits, many of them making it slightly easier in normal mode.  More importantly, though, the attempt cap was lifted this past tuesday, so From Chaos may have benefitted from the 25-man version being beaten last week.  Either way, it’s an accomplishment, as attested by their lone standing at the top of the strict charts there.

Finally, 25-man Heroic Arthas doesn’t look like he’s going to drop easily.  Blizzard buffed 10-man heroic Arthas from 22.5 million health to 29.5 million health and, 25-man Heroic Arthas has…103 million health.  To put that into perspective, far as I can tell you only have to take him to 10% to “win” (you need to dps the rest down, but it doesn’t appear there’s any real fight mechanic…just spam damage, he’s stunned).   That means you “only” need to deal 92.7 million damage before his enrage timer hits at 15 minutes, which comes to 103k raid DPS.  With 7 healers and the generous assumption that your tanks combined equal 1 dedicated damage dealer, you have 18 pure damagers, who need to do an average of 5722 dps.  Well, that’s not much; in ICC Heroic gear, which they’ll have slowly been farming in reaching him, they should have no sweat easily blowing past that.

While true, Arthas isn’t particularly nice about standing still and letting your DPS spam at him.  He spends at least 2 minutes strictly un-targetable (casting Remorseless Winter), taking the average necessary dps to 6.6k per damage dealer.  He also spawns adds that need to be burned down.  Assuming you use Necrotic Plague to take out shambling horrors, you’ll be able to focus on the LK for the entirety of phase 1.  Phase 2, however, involves the raid burning down the val’kyr he summons or risking the loss of a vital raid member (as they’ve been dropped a cliff, they can’t even be rebirthed).  This has to happen while avoiding the potentially raid-killing defiles he casts.  Assuming you make it to Phase 3, the raid will need to burn down Vile Spirits on top of focusing down Arthas before he Enrages (or a bit after he enrages).

All this while healers are pumping out enough health to keep everyone alive through LK’s considerable AoE damage and the two tanks are properly picking up adds, handling his debuffs, swapping him through Soul Reaver, etc.  Also, dispellers need to be completely on the ball at all times.

Assuming that the Lich King’s damage output is unchanged and none of his adds have increased health, we can estimate what sort of gear/skill improvement would be needed to beat him on heroic.  Vodka’s US first kill videos are here and here; we can see it took them 11 minutes to down him, including remorseless winter time.  So they’d get 4 more minutes of time on him, or 27% additional time.  Since the Lich King has 61 million health in 25-man normal, they needed to wade through 54.9 million health to down him.  In order to take down Arthas in Heroic, they’ll need to do 69% more damage…more than double what the extra time can account for.

I believe I can guarantee that the adds will have more health and damage to the raid will be higher.  So assuming Vodka was playing at the absolute top of their game for their kill (a relatively safe assumption), there’s absolutely nothing for it: Arthas will require better gear.  I think it’s worth noting that it’s only just now, several days after heroic mode became available, that the prior bosses have been downed.  I guarantee they went and took out the unlimited attempt bosses first, hoping to pick up some ilvl 277 gear to make the wing bosses easier.

I suspect that gearing will be a tad slower for heroic mode: none of this gear can be purchased.  Heroic 25-gear is a return to Ulduar and non-badge gear: you’re just going to have to get the right drops off the bosses, and you can’t buy equivalent side-grades for any slots anywhere.

Further, normal LK seems to be something of a wall for many guilds.  If we look at World of Logs, we have an admittedly distorted, self-selecting window into the success rates of guilds against Arthas.  We see that 1.4% of logged attempts were successful on 10-man LK and .9% of attempts on 25-man were successful.  That’s out of about 13k and 15k recorded attempts respectively.  By way of comparison, Sindragosa has a 5.5% and 7.1% success rate, respectively.  He doesn’t seem to be going quietly into the night for many of the people meeting him.

Which brings me to the question: is Icecrown Citadel an easy raid instance?  Have raids gotten easier with the introduction of Wrath of the Lich King?  It’s certainly an issue that’s been argued over for…well, the entirety of the expansion.  I know someone who reaches back to BC to talk about the “catering to the casuals” and the slow decline of difficulty in WoW high-end pve content.  Blizzard devs have addressed this topic a few times, addressing a number of concerns.  It should be born in mind that their overriding concern has been getting this stuff in front of more people.  Accessible Naxx was one step in that direction.  Weekly raids that are pug-able is another.  And the LFG tool has been a pretty big step in getting even the most casual player into 5-man dungeons and the WoW group experience.

Another argument Blizzard has put forward is that fights actually more interesting and involved, and insofar as complexity of mechanic can admit of more difficulty, encounters are at least as difficult as Vanilla fights and more interesting to play.  Zarhym makes points along those lines in a recent thread, and it’s probably worth quoting:

The fact that at least 25 members of our player base have killed Sindragosa on Heroic difficulty doesn’t say much about whether or not the encounter is more complicated than, say, Twin Emperors. In fact, most of the raiders I’ve talked to who are at, or have killed the Lich King in 25-player mode say that Icecrown Citadel has provided some of the most fun and interesting mechanics this game has yet to offer. Most of the fights today are technically more complex than the fights of old, but the difference is in the tuning and the polish.

He then brings up another point that anyone who was watching at the time can probably remember quite well:

I can think of at least a couple of bosses that took a long time for any guild to kill where bugs were a major factor. As soon as the bugs were fixed, the bosses were defeated.

I believe C’Thun was one such boss, leading to something of an uproar at the world first kill, as the patch went live earlier in one region than the other, causing cries of foul.  Finally, he mentions another old standby of Vanilla and early BC raids: the resistance check.

There were also several bosses, take Princess Huhuran for example, which weren’t necessarily incredibly challenging from a mechanic perspective. The challenge came in the form of requiring that players farm resistance gear from that same instance for weeks and weeks until they stood a chance. Resistance fights have since more or less disappeared.

However, one thing Zarhym hasn’t mentioned is the enormous penalty involved in a wipe.  First, Vanilla raids involved 40 people.  If you wiped, you needed to get these people all sorted out and back into play.  If you were fairly deep in an instance, this might involve 10-15 minutes from wipe to next pull.  I watched a friend once, long ago, practicing Kel’thuzad in the original Naxxramas with his guild.  He had flasked for the attempts, and he would literally log out after a wipe while people sorted things out in Vent to conserve the duration.  And he wasn’t logged out for short durations.

On top of this you need to bear in mind that trash respawned on relatively short timers (usually 45 minutes, if I remember right).  That would give you 45 minutes from first associated trash pull to down the boss before you had to go through them all over again…and trash could be pretty rough in its own right.  The time between boss pulls only aggravated this issue.It’s for this reason, in fact, that winger dungeons and raids were and are so popular: you could more efficiently choose bosses to take on to better maximize gear gains.  In long, linear dungeons, like AQ20 or AQ40, you had to just march through the bosses, and if one boss presented an impassable barrier to you (probably more due to not having the necessary time than anything else), you were screwed.

Beyond even that, high-end raiding was a grind-fest.  At one point in time, all elixirs stacked with themselves (there were no Battle or Guardian elixirs).  You could also use a Flask, along with the elixirs.  That meant, obviously, that the dedicated raiders would blow their time grinding up all the mats for flasks, elixirs, scroll drops, making sure they had the right classes for the proper buffs, scrounging up difficult-to-get enchants, farming the legendaries, etc.  This took time.  The bonuses granted, particularly by elixirs, were so large that it was like adding at least another tier of gear to the raid…and Blizzard compensating by increasing the gear check bosses represented.  That meant alchemy grinding moved from being a great way for top-tier guilds to make content easier to making that grinding a damned necessity.

What this added up to was an enormous time investment if you wanted to raid seriously.  Bosses might only give you a couple attempts before you needed to clear trash to get back to them, leading to hours spent just learning a single boss fight…but most of those hours were spent on preparation and trash clearing.  The top-end guilds weren’t just good; they also were dedicated enough to play World of Warcraft in every free moment they had.

Understand that a lot of the time sinks mentioned above were removed often at the raiders’ requests.  Grinding was boring and ridiculous, and I think everyone breathed a heavy sigh of relief when the alchemy nerfs went in.  Since launch, trash has been a source of contention between raiders and developers from day 1, with developers contending that raiders might say they just want a series of bosses, but really, the trash adds a level of pressure to the entire ordeal.  Raiders argued that it was simply a waste of time and energy, particularly when trash was so ridiculously punishing that you might wipe more to a boss’s guards than the boss, and they dropped absolutely no reward themselves.

I guess I can point to Wrath as evidence in favor of the devs’ arguments.

So once you’ve distilled raids down to boss attempts, with minimal (or in the case of Trials of the Crusader, no) time between bosses, how do we define difficulty?  If fights really were more difficult pre-Wrath, in what way were they more difficult?  I’m forced to contend that the difficulty relied entirely on how unforgiving of mistakes fights were, and how ultimately punishing failure was.  Failure now is lost time running back, and even that is curtailed by winged dungeons or teleporters in Ulduar.  How long can Blizzard force people to spend learning a fight?  How do they make learning challenging without making fights wipe raids if someone messes up once?  Because the fact of the matter is that the internet facilitates the dissemination of learning; we’ve all become damned good at finding out what everyone else knows about a fight.  When a boss comes out, the entire community masters it together.  So it can’t simply be a learning issue.

So it becomes a matter of execution, and a question of failure to execute properly and the punishment for that failure.

One other, tangentially related point: the skill of casual players.  When I first purchased Starcraft, over a decade ago, it was basically my first RTS.  I’d seen Command and Conquer, and I’d played a bit of Warcraft 2, but that was the first time I’d purchased an RTS of my very own.  I consumed it, bashing my head against the game, eking my way past every mission, clawing victory from the gaping maw of defeat…until the last Terran Campaign mission.

All of the stratagems I had used to that point failed me utterly in that mission, and repeated attempts only netted me repeated failures.  I gave up in extreme frustration.  I came back again, maybe a year later, and tried again, only to be frustrated with defeat.  Finally, I turned to the internet; not for specific strategies for that particular mission, but to teach me how to play the game, because I obviously didn’t know.  What I rather rapidly learned was that I had barely scratched the surface of Starcraft.  In order to truly play that game, you have to see it in an entirely new way.  Starcraft as played by the masters is not the game you learned when you bought it first.

What I learned from this is that we are precisely as good as the situation demands of us.  I probably could have, through simply throwing myself at the mission ad nauseam, beaten the Terran Campaign and moved on.  Or I could have looked up a strategy for the mission, followed it, and moved on.  I would from then on remained absolutely and blissfully unaware of the actual game of Starcraft.  I would never have known I sucked at that game, because, frankly, I didn’t: I was good enough to solve the puzzles I cared about.

If you are never confronted with a puzzle whose solution is desirous and which makes certain demands of your skills, you have no incentive to learn.  Simply getting better just to be better isn’t enough; that’s too nebulous a goal to be realistic.  In what way are you better?  Measured against what?  You’ve already solved the puzzles presented…how much better do you need?  Unless there’s something beyond that, I know I’d prefer to work on things where there actually is a puzzle to be solved.  Then not only can I get better, but I actually accomplish other goals as well.

This is what confronts the casual player.  Why ever should a Death Knight exceed 900 DPS if they don’t have to in order to accomplish their goals?  If 900 DPS is sufficient to clear through daily quests and allow them to have fun doing it, the puzzle is solved for them.  Finding a way to do this faster simply doesn’t need to occur naturally to these people.  This is true, I suggest, because fun is the ultimate goal in WoW and if fun is had at 900 DPS, they have no incentive to increase that number.  It’s completely irrelevant to them.

Once you drag someone into a dungeon or a raid though, suddenly expectations change.  Now there is, in fact, a reason to improve.  An occasional jaunt simply isn’t sufficient encouragement to improve those numbers; why bother if it’s something you’ll only do every once in a while?  But consistent dungeon running will push people to improve enough to get to the end of the dungeon.  They want to see it – it’s a puzzle that provides a goal and a reason to improve.  They can also grab some awesome loot in the process, further sweetening the deal.

There seems to be a natural tendency among people who’ve ventured into group content to want to do larger group content.  I’m fairly certain the LFG tool is helping to fuel that further now.  Once you move into 10-man territory, or 25-man territory, there will be further incentive to improve.  It has been a rare person indeed who hasn’t learned and improved after sufficient time spent in raids.  I’ve seen people who would otherwise be completely casual subjected to a couple nights a week of 10-man running for a month or so come out with substantially better skills, both in touching their abilities and in situational awareness.  They can’t help but become better at it through sheer exposure.

I think, then, that, when assessing the skill of the population of players, one should take into account the incentive they have to be skilled, rather than assume they are already as good as they can get.