On Sartharion +3, cutting-edge guilds who normally tank with a warrior would (perhaps somewhat begrudgingly) use a DK to tank instead because the benefit was so massive. The same thing happened on Vezax where good guilds replaced their warrior tanks with DKs. Death knights were overpowered at the time.

We are not seeing the kind of massive swapping one class of tank for another in ToC or the preliminary Icecrown testing. It’s just not happening. If paladins or druids are too good or warriors or death knights aren’t good enough, we’re just not seeing them getting used for those fights to anywhere near the extent that DKs replaced other tanks on the encounters mentioned above.

As just one example (and I wouldn’t put too much weight behind this), the number of paladin tanks used to test Icecrown on the PTR has been almost imperceptibly low. It’s possible there were actually none this weekend, or else the numbers were so small that we missed them. Sure these are not the hard mode encounters generally being tested, and you can come up with your own explanations for differences between how players might approach PTR and “real” raiding. But these are the kind of lopsided datasets we have to deal with when we’re looking at who is using what tank and how that relates to success.

I’m not talking about balancing around representation, as players often mistakenly claim. I am asking the question that I have asked before: if one or two of the tank classes have such a superior and unequivocal advantage in making a fight easier, why aren’t more guilds using them, especially given the evidence that they have swapped tanks on previous encounters? I understand individuals of you may state it has happened to your guild. That’s fine, but understand you are in a small minority. Overall, it’s just not happening.

One conclusion you can draw (though certainly not the only one) is that whatever differences in survivability there appear to be “on paper,” just don’t make a significant difference in the actual fights, or at least not enough of a difference that guilds feel compelled to switch or are held back by not switching.

As an aside: there are some specific fights where the specific mechanics lead guilds to tend to use one tank over another, say a druid to tank Thorim hard or a shield-using tank on adds for Anub hard. Most players aren’t as concerned about these one-off encounters because it feels like such an encounter-specific problem and doesn’t feel like anyone’s job as guild main tank is in jeopardy. (Individuals of you may feel differently of course.)

I know there are lot of threads on this topic right now. I avoided many of the ones that are doing a good job discussing numbers because I didn’t want to derail them.

Q: When you mention Sarth+3 or Vezax hard mode, you make it sound like it really mattered that DKs were better tanks for those encounters, but then when discussing Thorim hard mode or Heroic Anub, you make it sound like it doesn’t matter that certain tanks are better for those encounters.

A: No, I totally agree that the distinction may be subtle or even subjective. The way I would describe it is that druids are good on Thorim because of Unbalancing Strike, and not just because “druids are better tanks.” If every, or even many, bosses had Thorim’s attacks, then druids might be perceived as too good overall. (And they may be, but not because of their lack of reliance on defense.)

To use a contrived example, if there was a boss that could be beaten by anyone but more easily defeated by Spell Reflect, you might see a lot of grumbling about how good warriors were tanking that particular fight, but you wouldn’t I expect see a lot of folks arguing that warriors were too good across the board. That case has less to do with say EH or cooldowns and more to do with a specific mechanic of that fight. But as I said, it’s subjective. If the druid was so good on Thorim that you felt gimped without a druid and felt like you had to recruit one, then it would cross the line and doubtless that line would be at different positions for different folks.

Q : So basically you’ve realized that you posting tends to make threads worse more than they make them better and have decided to post in the threads that are terrible QQ in the first place?

A: If I wanted to comment or criticize someone’s numbers, I would post in a thread about numbers. If I wanted to discuss philosophy, I might look for a different thread. Overall I try to avoid both the worst and best ones.

Honestly though you would do us all a favor not to put too much emphasis on whether a thread is rewarded or punished by a blue post. When I have something I want to say, I try and say it. Requiring me to analyze the absolute best place to say it just takes away time I could be using to address another issue elsewhere.

Q: So, unless Kungen, Xav and the other warriors that have been tanking for their guilds for 5 years quit….

A: You named those names, not me. I feel the need to respond here so that my original post doesn’t get locked into “Blue only cares about the top 1%,” because of your interpretation. I feel if I had not added the caveat that we look at reasonably good guilds that someone like you could have easily posted “Oh great. Because some idiot chose to tank Icecrown with a Voidwalker instead of a tanking class, they dilute the data from groups that actually know what they’re doing.”

Q: First, you are limiting it to within a guild when you should actually be looking at it as different guilds using different tanks.

A: While we are fine with talking philosophy or design goals, we generally don’t talk about how we come up with our decisions or specific data. We want you to focus on the end result (the game) and not on trying to criticize our data-collection techniques. Both are valid things to comment on if this was a peer reviewed journal and you wanted to point out flaws in our setup. But these forums are targeted towards feedback on class balance and mechanics not on how Blizzard operates.

Q: GC likes to ask ‘If it is unbalanced, why don’t guilds swap’. The answer is because of the massive barrier to finding, gearing, and training a new tank. Often times the task is impossible due to availability of certain classes. In case you haven’t noticed, there aren’t as many paladins and druids in the game. And there are a ton of experienced warrior tanks from Vanilla wow days, when warrior *was* the tank. Guilds don’t have a choice of tanks, they have to work with what is available. Good feral tanks don’t just spring up out of the ground fully formed.

A: I can accept that the challenge of finding and gearing new tanks can affect individual guilds. Overall I am suspicious that it accounts for the fact that very few guilds use paladin or druid MTs on progression attempts. It is arguing that guilds might desperately want druid or paladin tanks because everyone knows they make the fight easier, but they just can’t find any. Yet when DKs made the fights easier everyone managed to find those. I have a hunch that if we made a fight in which stacking Blacksmithing BM hunters with Nether Ray pets conveyed a significant advantage that we would see Nether Rays coming out of the woodwork. Players typically find a way when it matters.

We totally get that many long-term guilds still use warrior tanks because they always have. We did see many of those guilds switch to DKs for certain fights though. They by and large are not doing that now. Why not?

Q: The way you operate directly correlates with how things happen in the game. Questioning your methodology is directly talking about the end result. The end result is merely a product of your methodology. It is always valid to question methodology when it influences end result. To think otherwise is to believe one’s self to be beyond reproach.

A: We don’t run our meeting schedules, marketing plans, office configuration, compensation structure, shipping schedules or other information about our operations with our customers or the community. That isn’t to say they are above reproach. We just aren’t particularly interested in your feedback on them. We are very interested in your feedback on WoW classes. Stick to that topic if you want to be heard.

Q: This is an interesting point. One thing I’ve noticed is that warrior tanks are still in raids, but as off tanks now, not as main tanks. This is something that blizz might not notice as easily, they just count the number of tanks with certain achievements, they don’t know what job the tank did. Even some of the more famous tanks have put themselves in an OT role, for the sake of progression.

Right now paladin and druid get put in front of the bosses. Warriors clean up the adds in the back. And poor DK’s are dpsing.

A: I can understand how that might be your perspective, but if you could see the whole picture you might understand that you are dramatically overstating things. It may be true of your guilds or your friends or the guys you talk to on the forums. It is not true overall. Most bosses are tanked by warriors. Most hard mode bosses are tanked by warriors. Nearly all server first boss kills are done with warrior main tanks. Paladin tanks are so rare that it sometimes doesn’t even feel fair to compare the data.

However, we do know that sometimes those warriors will step aside because another class conveys enough of an advantage that they must feel it’s worth it. This was true of say Sartharion and Vezax. It is not true in ToC on anything remotely of that scale. We don’t think it will be true of Icecrown based on testing so far. If it becomes a problem — if we think that raid groups with warrior or DK tanks (or even druid or paladins) are struggling more than other groups — then we’ll do something about it.

Q: Blizzard does not have an internal raid group capable of Heroic Mode 25 man raiding. They would be at best world 1000 in the ranks to venture a guess. Ghostcrawler admitted that the internal group had trouble with Yogg-Saron *normal*, all 4 keepers used, during ulduar testing.

This is why PTR exists, or at least one big reason. Without the bleeding edge guilds that test there, no Quality Assurance would be done at all on heroic mode encounters.

Consider the number of hours a week many heroic mode raiders spend raiding, farming, researching, etc. It is a full time job, one that blizzard developers simply do not have time for, and blizzard isn’t going to hire players to test for them when they can get us to do it for free.

A: We test hard modes and we employ outstanding players who know the game very well.

We are also smart enough to recognize that independent testing can be very valuable. Some of the best information we get out of the PTR is seeing what solutions players try and come up with to handle various encounters. That kind of coverage works better the more players you have who partake. It’s also great to see how players with no previous exposure to the encounter react to it. Once a tester here has seen the fight, they are no longer inexperienced.

Q: Because Ulduar “ended” at almost the same time you nerfed DKs. Coliseum has had one EH testing fight, and one fight where mechanics favored warriors offtanking and Druids maintanking. The other fights can have any tank thrown in without much worry since they’re more about raid coordination.

Again, my money is on you not seeing a lot of tank swapping till the hardmodes of ICC are ungated/locked. However, at that point I really think you’ll see this shift you don’t see now happening just like DKs were shifted to in Ulduar. The EH, avoidance and cooldown discrepancies between Paladins(and to a degree Druids) and the other two tanks are too high to not make a difference on any fight that challenges tank survival.

A: If your point is that swapping isn’t happening now but might happen in the future, I think you have more of a leg to stand on, simply because we so far lack the empirical evidence to support or not support that prediction. Some of your compatriots here are arguing that swapping is happening now. It’s not, at least not on a large scale.

Q: GC i sure hope your vision for “bring the player not the class” is for every player to have a tank of each class and swap it in for fights that need it… cause that’s what it really is feeling like right now. If things are so bad, why aren’t people switching chars?” is never a proper excuse to discount the math

A: That’s not our vision. In fact, part of the reason we wanted to make all four classes generally good at everything was to avoid the burden of guilds having to gear up a DK for those relatively rare magic fights for which they were originally (before LK shipped) designed to tank.

I’ll counter your final quote by saying that theoretical models are never a proper excuse to discount empirical data. Empirical data in this case can only be gathered by raiding. Making estimates by using math does not somehow get you closer to the truth just because the mathematical operations themselves are pure. If doing math solved all the world’s complicated problems, we’d be surprised as a society a lot less often.

Q: Because there a difference between having an easier time, and being a fool not to switch. Using a paladin or druid will no doubt be a better choice in current content, but this isnt enough to counteract 5 years of dedicated service.

Now, early on Vezax and Sarth 3d, you would have had to have been a complete fool not to use a dk because of their cooldowns. It was because of a minute IBF. I know this, I know you know this, so why are we playing these games?

A: If we know that players will switch tanks at some point and yet they are not switching tanks now, then perhaps any perceived imbalances among tanks are not that big a deal in reality, or at least not as bad as some of these forum posts might have you believe.

If your contention is that many of these discussions are about slight yet annoying imbalances among tank classes, well okay, but my sense is that many of these frequent posters are describing a situation they think is much more dire.

Q: Guilds don’t switch tanks when it’s do-able with one tank, but 10% easier with another tank. 10% easier with another tank is still a problem, though, and one that you have, thus far, not shown any desire to fix.

A: It’s just challenging for you to prove that a 10% difference exists, or even that 10% somehow crosses the line. I’m not even sure what “10% easier” means. That the fight is 10% more stressful for your healers? That you progress 10% slower than other guilds? That you have a 10% more likely chance to wipe? That means that for every 10 times you fight a challenging boss that you’ll wipe once because you picked the wrong tank. That kind of prediction seems like it would swamped with the myriad of other variables that could affect your success. High latency probably makes a fight much more than 10% harder or easier.

Q: 1. Warriors are generally considered to be very “fun”
2. Paladins/Druids have a fairly “static” rotation which can lead to boredom

A: Plenty of players find warriors not fun or enjoy paladins or druids. I’m not sure you could quantify that at all. And even if you could, I just don’t buy that guilds wouldn’t find a way if it really made progress significantly easier for them. Maybe not every casual raiding guild could, but the hardcore ones would do it quickly. Someone would be willing to take the hit and play the less fun class in order to advance. If it made progress slightly easier for them, then yeah, maybe, but if we’re talking about “slightly easier” then aren’t there more pressing issues to handle?

Q: How in the world did you go from me questioning your data collection methodology to your day to day activities like meetings and work schedules? What I brought up was perfectly valid. I questioned the metric you are using to determine if something is imbalanced or not. I think you are looking at the wrong thing as your litmus test for determining if there is an imbalance.

I am on the topic of class balance because I am questioning what you are looking at to determine if there is a class imbalance. I think you are looking at the wrong data. That is a perfectly valid way to debate balance. Quite frankly I think you should stick to the topic. You are the one bringing things completely unrelated to the topic (you data collection’s focus) by saying you don’t run the day to day activities of running a business by the players.

Also, I don’t like your veiled threat to ban me. It is extremely unprofessional.

A: It wasn’t intended as a threat to ban you. I was just trying to point out that we’re not going to take very seriously feedback, criticism or suggestions on how we gather data. Sorry if that bothers you. I can understand why you think it’s a legitimate topic to discuss, but we disagree. That’s just not part of our development process upon which we want to have a public debate. We want you to focus on the classes, not data gathering. That’s not why we are here on the forums

Q: I’m betting on at least 50% of the bosses having a huge EH requirement, then we’ll see who is tanking

A: My prediction is warriors will main tank more bosses in Icecrown than all the other tanks (probably more than all others combined), even the hard modes, even Arthas. There might be specific fights where other tanks do slightly better.

Q: Honestly, the only thing I want to know is why the DK cool downs still have resource costs. For the life of me, I can’t figure it out.

A: Slightly off topic, but you contribute a lot to these boards so I’ll answer it. The original intent was that DKs press different buttons when tanking than when doing dps. Warriors definitely do that — few Arms or Fury warriors incorporate Revenge or Shield Slam and then Shockwave and Devastate are talented. We needed abilities for a DK to use that sacrificed dps for survivability. (The alternative is you don’t need a tank spec at all and can do as much dps while tanking as when not tanking, modulo gear.)

Given the realities of tapping for runes and the like, I don’t think the reality ended up where we wanted it. It would be closer to the design if DKs dropped say one Blood Strike for a defensive cooldown while dps DKs did not. That would be a tough change to implement at this stage with all sorts of negative ramifications. However just taking the costs off with the intent of putting them on again in the future would be unpopular with players. We’ll make unpopular calls when we think they are worth it. I’m not sure this is.

Q: Nothing about anub himself is Gimicky, at all. In fact, tanking him is as standard as they come. The only “gimick” to the encounter is with the adds, which need to be tanked by a block tank due to the “gimick”. On anub, you want the best tank—you use a druid or paladin, end of story.

Certainly no one has tanked Anub himself on a DK or Warrior. That would be too hard because of class imbalance.

A: I agree that Anub is a difficult fight to discuss because of its mechanics, but if you try hard enough you can find justification for why every boss (other than Patchwerk maybe) is a gimmick and shouldn’t count. However let me also throw out that a lot of guilds recognize that the hard part of the Anub fight isn’t Anub at all, but managing the adds. They choose to send the player who is either their best-skilled or best-geared or just their traditional MT to do the hard job. That player is most often a warrior.

I agree that druids are very common to tank Anub himself. However so far it is not the case that the druid just ends up as the new MT for the guild. Most often the druid just takes that one fight, just as druids often take hard mode Thorim. Yogg 0 is a popular paladin fight, probably because a million things spawn quickly to beat on the tank at once, requiring both strong threat generation and survival against multiple adds. However once again, this trend (if it’s even that) has not led to the paladin tanking the rest of Ulduar hard modes.

I might even concede that every fight if you break it down enough might slightly favor one class over another. As long as it isn’t always the same tank and as long as it doesn’t lead to guilds who “picked the wrong MT” being roadblocked, we don’t think it’s a problem. We have no evidence that picking a warrior MT puts you in that category, while we have a lot of evidence that correlates warriors tanking with success.

Q: Only because there are numerically far more warrior tanks available to do so.

A: The question was who was going to end up tanking Icecrown. The smart money is on warriors. You can bring up EH discrepancies and other issues if you’d like and we’ll consider them. But if you’re the kind of player who is going to state for a fact that a significant number of guilds are replacing their warrior MTs with paladin MTs, I am here to disabuse you of that notion.

Q: I hate when people [or Blizzard] say any of the following:

1) Guilds still employ more Warriors as MTs than any other class.
2) The boss has been killed more times with Warriors as the MT than any other class.
3) Guilds have more Warrior Tanks than any other class.
4) Only a small fraction of the Warrior Tanks have been benched or replaced by a different class.

Every one of these is can be explained due to population differences and balancing population is not balancing the classes. All of the above can be held true simply by having 100 warrior tanks for every 1 other class tank.

A: We only bring up the point because we know that raids or guilds *will* switch tanks at some point if things are bad enough. The fact that they aren’t switching now could be evidence that things are not as bad as some of you claim.

Q: I don’t have the population statistics, but it’s pretty obvious that there are simply more prot warriors than Prot Pallys, Bear Druids, Tank DKs… and probably more even if you combine them all. That does not mean that Warriors are OP.

A: I agree it does not mean that warriors are OP. But does it mean we should nerf the tanking class that nobody is really using? Or maybe it means that class balance is close enough (with regard to tanking) that players can generally pick which class they want and do okay.

Q: Just something slightly off-topic, have you ever thought it might just be a better idea to start your own threads once you find a thread you want to discuss and just post your thoughts there and let it go from there. That way you won’t find people running to a good thread to derail it. Just an idea.

A: I’ve thought about it, and done it on occasion, but that tends to minimize the back and forth between players and myself. When it’s a new thread, people sometimes seem to think that means it’s an open “Ask GC about anything” thread. We’ll continue to try different things though and see what works the best.

Sorry for going offtopic a second time.

——————————————————————————————————

We’re just going to have to agree to disagree that the only reason there aren’t more paladin or druid MTs is because they are hard to level or somehow less fun than warriors. I still have a strong suspicion that guilds would find a way. Our players have proven themselves capable of doing outlandish even painful things that confer only a very small advantage. Difficulty is rarely barrier to min / maxxing.

Therefore, if you believe paladins and druids are better than warriors, then I think it must be the case that one of the following is true:

1) The tanks are still close enough that there isn’t enough motivation to use a paladin or druid.
2) The more casual guilds are using paladins and druids. The hardcore ones (ironically those usually accused of being most obsessed with min / maxing) just power through the imbalance with their warriors through skill and determination.
3) ToC and perhaps early Icecrown are not a good showcase of the situation so the sidelining hasn’t happened yet. But it’s coming.

However, even if you believe number three to be true, you can’t use that to justify the kinds of statements I am reading on this forum about how all the warriors have already stepped aside for paladins or that warriors are a dying breed or that your guild replaced you for a paladin in quest blues or whatever. This last part is what I keep trying to hammer on. You can say it might happen in the future. It’s not happening now in anything resembling a large trend.

——————————————————————————————————–

Q: All they need to say is:

“We’re investigating the possible tank imbalances by running stress tests on all classes of all gear levels with multiple points of failure embedded intentionally to ensure all tanks are reasonably balanced on a metric of real survivability. We are running these tests and closely observing the results and will make adjustments as they are needed”

that’s ALL.

leaving us in the dark is sheer arrogance and indifference towards their consumer

A: We do that all the time. From the moment players first brought up their concerns we looked into it. The problem comes then if we say “We’re happy with the state of things,” many of you reply “No, no, you must have done something wrong, or you would be agreeing with us.”

Q: For most of this community, that “philosophical” vision is dependent on Armor, Health, Damage Reduction, Cool downs and to a lesser extent avoidance/threat. If a tank comes up short in almost all those fields, then he is going to feel bummed out about tanking.

Maybe our “philosophical” views are different but many of us came to these terms about what defines a tank based on years of playing your dungeons, GC…Trying to get us to see things differently probably isn’t going to happen.

A: Our philosophical vision is that you can pick one of four tanking classes and generally be successful on any encounter. It’s not a big deal to us that someone “wins” the stamina contest as long as everyone could still fight the boss within a relative degree of success from each other. Determining that relative degree is pretty hard without a large sample size of actual raid encounters. You can estimate the differences. You can attempt to model them, and many of you do. You can make anecdotal comments about how things feel when you’re raiding. All of those are important. Successful boss kills are the most important metric, but also probably the hardest thing for the average player to measure. There is no reasonable equivalent of the dps dummy for tanking (not that dps dummies don’t come with loads of caveats for determining dps).

Put another way, cats have a ton of attack power, but it feels less contentious because the dps specs can compare their dps. (Though even in that case, there is a lot of quasi-religious debate about whose dps is highest, even though it really depends a lot on raid synergy, player skill and the encounter in question.) Holy paladins have a ton of mana, but it’s not as simple as the most mana = the best healer. It’s hard for tanking specs to compare their “tankiness” so they tend to look at things like health and armor (which to be clear are both amazing stats for a tank). But the “tankiness” is far and away more important to us, and at the same time avoiding homogenization is also really important. We want death knights to have good cooldowns, druids to have high health, and paladins and warriors to rely on their shields (ideally we’d also have paladins and warriors more distinct in that regard).

Q: I’ve never seen evidence that Blizzard wants a given tank class to be special, for one to have a niche that they’ll have, especially since they scrapped the anti-magic niche for DKs.

They want tanks to have their own special ways of doing the same thing, but not necessarily doing one thing better than the rest intentionally.

A: Yeah, we have to be very careful about the whole “I’m supposed to be the best at X” thing. It is something we’ve grappled with on the dps and healing specs too. It’s fine for warlocks to have a lot of dots and for that to help distinguish them from mages. It gets tricky when warlocks then protest that they should have more dots than priests because “they’re supposed to be the dot class.” We aren’t necessarily trying to give unique superpowers to every spec. In fact, it’s more accurate to say we’ve been moving away from that.

What I was trying to say though is that we want the classes to play differently — the flavor goal you guys sometimes poke fun at. Druids have a lot of armor and health. Death knights have a lot of cooldowns. Warriors use their shields to mitigate damage. You shouldn’t take that to say “DKs need buffs because they’re supposed to have the best cooldowns.”

I know the distinction may seem subtle. I’ll also freely acknowledge that A) shields have not kept up as a good form of mitigation, and B) warriors and paladins overlap way too much as the “shield guy.” Warriors still get a lot out of their shields. Shield Slam is not something a DK or druid can do. (Now please don’t go around quoting “GC says to bring warrior tanks on hard mode fights for Shield Slam!”)

Q: this would be true if you could all accomplish your task with 0% fail rate. as soon as any fails are integrated you actually see the benefit of having these advantages.

A: But paladins have a fail rate too. Furthermore, nobody has demonstrated that the fail rate is lower for paladins. The best anyone can do is to say “On paper it appears that paladins might have a lower fail rate,” or else “In my experience, paladins have a lower fail rate.” You can’t discount either of those as sources of information, but neither can you discount empirical evidence that suggests in fact that warriors probably have the lowest fail rate when you look at how many warriors successfully MT and defeat the hard encounters. So then you are in kind of the awkward position of arguing that even though warriors are relatively more successful and paladins are frankly relatively rare, the warriors deserve buffs because on paper it appears like they should be worse tanks.

Now for those of you saying there isn’t a problem now but there might be at some point in Icecrown, I think that is a more defensible position (if only because the empirical evidence doesn’t exist yet to support or fail to support that thesis). Part of this job is definitely trying to look at trends and fix problems before they get out of control. We don’t think this one is going to get out of control, but we’ll see once more players are tackling hard modes (with appropriate gear). I know some of you are saying it will be too late then, but we’re also not going to pre-fix problems until we’re pretty convinced they are actually going to be problems.

Q: Didn’t they say the ICC bosses will hit faster and for less damage? And in the same breath they said the hardmodes will test tank survivability?

A: I know that seems like a two-faced answer and it bugs me that I didn’t explain it better.

The “Icewell Radiance” is designed to account for very high avoidance levels. It lets us lower the damage bosses do per hit because their hits will connect more often, keeping boss dps the same but lowering spike damage.

However, I also know what that message says to some people. It says that if I go into Icecrown and wipe on a boss all night that Blizzard lied to me about boss damage. We’re just sort of bracing ourselves for the inevitable “I couldn’t beat Marrowgar because he hits so hard. I thought bosses wouldn’t hit hard.”

Bosses hitting less hard does not mean bosses can now be facerolled. It just means they hit less hard. To be fair, most of the folks posting in this thread understand the difference. But it’s not uncommon for us to see our quotes misused in other forums where we can’t necessarily bop in to correct the misunderstanding.

Q: Yet those points are never refuted. Instead the response is ‘tanks aren’t rerolling, thus no problem’.

How about I toss the burden of proof to the other side. Write me a compelling theoretical model that describes, with facts from current tanking, and current encounters, why a paladin is equal, not superior, to a warrior. Not based on player behavior, based on facts. I’d love to read it. I’m a science type person and willing to entertain the notion that I’m wrong.

A: I’m not sure if you were addressing us specifically Regill, but the reason we don’t do that is the same reason I have often given:

We don’t want to be in a position where we turn over all of the power to design and balance the game to the community. We don’t want to be in a mode where we, the developers, have to get buy-off from you guys to make the changes we think are right for the game. We don’t mind trying to explain our philosophy on things, but once it goes from that to having to prove to the forum posters that we’re making the right decision I just think we’re going to end up in a very bad place.

I know that’s not “fair” in the sense that we ultimately have the power and the burden of evidence is 100% on you guys. But at the same time, understand that we’re not asking you to balance WoW for us. We are very interested in your feedback, however. That’s a subtle distinction that players can’t always get. They don’t follow why if we listen to them, they can’t necessarily force us to agree with them.

EDIT: I see later where you state you were talking to defenders of the status quo and not necessarily us. That’s cool — I wasn’t necessarily trying to single you out. I still think my reply above might help to explain our point of view a little better.

Q: So, just clarifying, Blizzard is using population as a measure of balance with the justification that if people aren’t rerolling, then it’s not a problem? So, this means that blizzard is specifically looking for people to dislike their class so much that they drop it entirely in lieu of whatever FOTM class their is. Why is that necessary? Why does it have to get to the point that people are doing making decisions on what class they play because of class balance rather than what tank they want to play?

A: Some of you guys are way too hung up on this rerolling thing. I’ll try and explain it one final time, and then if you’re going to continue to try and claim I am saying something else, there’s not much I can do. The point is not that we’re trying to force players to reroll. The point is that if warriors in both cutting-edge and more casual guilds are still beating bosses and guilds are still overwhelmingly relying on warriors to tank content (in the face of evidence that suggests they will swap tanks when the warriors struggle too much) then maybe the perceived differences among tanks really aren’t that great.

Q: What’s the needed amount of people complaining in order to be proven right since apparently that’s the only way to be justified. Is it 90 people? 100 people? 10,000 people? Are the arguments being presented justified only if their are 50% of the warrior population stating it?

A: The necessary amount is one. One player making a good argument is sufficient to sway us. But it works the other way too. We have to be allowed to disagree with you as well. When you slip into the mode of “They are wrong until they make this change I want,” then you’re not really granting us the ability to design the game. You’re specifying how we have to design it. I don’t think you’d like the end result if we designed in that manner.

Q: THE ONLY REASON PEOPLE RE-ROLLED DK WERE SIMPLY BECAUSE IT TAKES YOU NO TIME AT ALL TO GET THEM TO 80, AND WHEN DK WERE OP, ULDUAR HAD BARELY COME OUT,…. it’s not like it was annoying to gear a tank up then, and it’s not now really, but at the same time it is, there are “key” tanking pieces you can’t get easily, such as a good weapon for a DK.

A: Posting in caps doesn’t make your argument any stronger. People pages before this one acknowledged that a mere 55 level head start is not enough to make a progression-focused guild give up on finding someone with a vastly superior advantage to tank their raid.

Given that they don’t seem to be doing that, maybe just maybe the advantage isn’t that superior.

Q: This conversation keeps going in circles.

If the tanks are overpowered, why aren’t they being used? Because of the myriad of reasons you want to stick with your main toon, right? In that case, why did people not stick with their mains when the DK was overpowered?

Is it because there aren’t enough cases in TOGC where the Paladin / Druid is overpowered? Then how do we know that there are going to be enough cases in ICC? What if there still aren’t enough cases in ICC?

A: I agree. I’ll keep following this thread, but I think it’s pretty apparent to the folks who have been sticking with it since page one that we’re now getting a lot of the same arguments brought up on page 55 than have appeared several times now. It’s the nature of forum communication to some extent, so I’m not sure there are easy solutions to that problem.

Q: “people aren’t switching” is not a valid excuse to deny balance in a game. stop that noise

A: Warrior tanks can tank every boss in the game, and more often than not get the world first and server first kills on those bosses. More raids use warrior tanks than any other. Most cutting-edge guilds use warrior tanks. We can’t detect a higher “failure rate” on warrior tanks, and frankly I’m not even sure what that means given that most everyone here agrees there are far more experienced warrior tanks than other tanks out there. Some of you are arguing “when we switch to the undergeared paladin, we do better,” yet very few groups are actually switching. The best we’re stuck with is some kind of fuzzy “it’s harder on the healers,” which is pretty hard to quantify, especially since the healers aren’t running out of mana healing anyone.

This is why the developers keep going back to what we are seeing happening in the game, which is warriors tanking stuff

Q: no one ever claimed that paladin advantage now is DK Sarth level. to compare them at all means you don’t understand the issue, GC

A: Then I’d argue we have spent 57 pages debating something that might very well be a pretty trivial problem even in worst case scenario. Even in the DK Sarth days, there were plenty of other tanks out there. We just thought DKs were so good that it was jeopardizing other tanks having a job. That does not appear to be happening now. As I said above, if you’re worried that we have a problem because another tank does the job 5 to 15% (or whatever) better than you, you’re going to have to explain what that actually means and why it’s a problem

Q: I know you have a lot to digest on this mega-thread Ghost, but what happens if Sarth and Ulduar hardmodes are repeated in Icecrown hard modes, and Warriors/DKs can’t tank stuff effectively compared to the other two, to the point where their jobs are in jeopardy? Will you be able to fix it while Icecrown progression is still relevant? If you can’t say yes to that, doesn’t cutting down on that 15%+ EH difference(new gear scaling) make more sense than hoping it doesn’t become a problem?

A: Yes, if we see paladins or druids (or warriors or DKs for that matter) repeating what happened on Sarth or Vezax, then we will do something while Icecrown is still relevant.

When is the appropriate time to buff or nerf is something that is always going to be subjective. Buffing someone when we aren’t convinced they need it (assuming it affects relative power compared to another class) is as dangerous as not buffing someone when they do need it.

And again, it’s not as simple as guilds having a favorite tank for specific fights. Druids on Thorim or warriors on Anub adds didn’t cross the line for us, though I understand why it might for some people.

Q: What do we do about the fuzziness that will occur in ICC due to the difficulty of the content? ICC and especially the hardmodes should pretty much rock the crap out of us and expect really tight play, with some fight or two being M’uru quality. There’s been a lot of talk so far back and forth, and some of it quite cynical too, about how if your Healers just healed a bit better, that death would be a success and it isn’t the problem of the class.

A: I don’t have a straight forward answer for that, in that we don’t have some kind of algorithm that returns a positive when things are too tough. We will just have to keep evaluating how things are going and how we feel about progress and success over all.

I will say that, ironically perhaps, the guilds that get early or even first kills are often the least well geared because they went through the instance so quickly while everyone else had weeks to farm the easier content before they hit the hard stuff. The difference between the best raiding guild and the worst raiding guild is on the order of 1000s of percentages — much higher than any of the numbers players have been tossing around here. Skill can overcome some amazing deficits. I say that not as “Suck it up” but that if you think your healers can’t possibly keep you up that maybe you’re not approaching the encounter the right way.

Q: i’m just saying. to any complaints we bring to bear, “all tanks are completing content” is the answer we get

which is completely dismissive of the concerns and not relevant if the content is not balanced around the people who have the level of skill to easily beat content regardless of said discrepancies.

it’s like going to an obstacle course and then getting told that you shouldn’t have problems because jackie chan doesn’t have problems doing it

A: I would like to think that my posting in this thread like twenty times would suggest we’re not dismissing your concerns. My concern however is that anything short of “Okay, we’ll buff you,” is going to be interpreted as dismissing your concerns.

(Source)