Why Is Class Representation Considered?

Tanking	Why Is Class Representation Considered?
07/17/09 10:41:30 AM Patch 3.1.3 Lovenectar	I'm not sure I understand the logic behind using class representation when deciding how to balance classes for PVE. I've seen GC, on several occasions, allude that since a lot of people continue to play warriors, they're ok. It's as if Warriors are intentionally being left as is in hopes that people will reroll to help balance out class representation. Why is balance amongst class representation pursued? Does it really even matter? There are also comments being made that guilds are still using Warriors in Ulduar, so they must be ok. This doesn't make sense either. Yes, they CAN tank the content, but not as easily as other tanks and without providing the equal benefits other tanking classes provide. The fact that Warriors are still used, despite their shortcomings, makes me believe that there's probably factors other than class performance that are contributing to Prot Warrior representation. They've been around and been capable of tanking effectively for longer than any other class. Therefore, there will be more just because the amount of time they've been available. This extended amount of playtime would also lead to a stronger bond to their toon. A warrior of four years won't be as willing to reroll because another class is more effective at tanking and guilds won't can their long time tank just because he won't reroll... most of the time. These factors, and I'm sure there are more of them, make me wonder why representation is even considered. Why is a class deemed "ok" because people still play them? If you really want to bring the tanks in line with each other, wouldn't it be more effective to base any changes or tweaks on the performance of the classes? This attitude that, even though Prot Warriors have the lowest DPS, TPS and take more effort to AoE tank effectively, are ok because people still play them is a little perplexing to me. It's like dedication is being punished with massive amounts of inconvenience. Does class representation take precedence over class performance?
07/17/09 01:48:40 PM Patch 3.1.3 Ghostcrawler <Lead Systems Designer>	Q u o t e: This. If less people are playing rogues than say, hunters, Blizzard wants to know why. Is it more fun? Is it too powerful? Can they make hunter gameplay more fun? This is true. Q u o t e: Basically this is what I think of for why they consider it. GC has said that they are afraid of buffing warriors too much because they previously were THE tank class and that perception is still carried by a respectably high number of people. They want to make sure the other tanks are considered equally effective tanks, so they are being very careful how they buff warriors because if they over buff them they will have just destroyed everything they were trying to change about the majority perception within the community. They also take into consideration whether they are being effective in the current content and since they are they haven't had any need to do anything in an emergency mini-patch way. And so is this. Let's consider a totally hypothetical example (and I am being serious about that): Say we did some extensive data extraction from Ulduar and found that only 5% of guilds use Feral tanks when learning hard modes. Assume for the sake of my very contrived example that we could somehow select for those guilds with a potential to beat the encounters, but that the encounters weren't on farm yet. Assume that the sample size was somehow large enough that the statistics are not at fault in any of this data collection. (I'm trying not to let you Kobayashi Maru your way out of being able to resolve the scenario.) Now, let's also assume that we convinced ourselves beyond a shadow of a doubt that using a competent and appropriately geared Feral tank made most of the hard mode encounters significantly easier. Assume that the community also felt the same way -- that it wasn't a dark secret. The fair thing to do from a balance point-of-view would be to nerf Feral tanks. This will likely cause the percentage of them to drop from say 5% to 2% or virtually nil. A game designer should look at that and say: Yikes! You can argue that maybe the bear is just a horribly frustrating spec to play and so nobody does it despite its advantages. I don't really buy that though. Players tend to say that about all of the classes, and I don't see a lot of evidence that Ferals are somehow unique in this regard. Furthermore, many of our players will do things that are soul-crushingly frustrating if they think it might confer to them a small advantage, which is often why we nerf such things -- to save players from themselves so to speak. It's just hard to resolve how, in this particular example, why more guilds don't go stampeding towards druid tanks if they are overpowered. It's a tough question -- what to do with the overpowered but underplayed spec, assuming it doesn't have any crippling gameplay flaws? What do you do with the spec that is wildly popular but underpowered? Do you make them somehow less fun (even if it's relative) so players try out the other specs? I think saying "just make all the specs as fun!" is a cop out. We try to do that all the time, but I don't think that will ever result in as many shamans as warriors. This is why I say we don't balance around representation. We don't tweak numbers until we have 25% of each tank in Ulduar.(Or should the number skew higher towards DKs since they have more than one spec? Or should the numbers skew lower for paladins and druids since fewer races can be them?) But we do have to consider representation when we're making changes.
07/18/09 01:14:25 AM Patch 3.1.3 Ghostcrawler <Lead Systems Designer>	Q u o t e: Consider when DKs were new to the scene. There was a common perception that DKs were squishy tanks and needed significant buffs just to be able to make it through a heroic. Much less a hard mode raid. There were some skilled DKs, but by and large, most people agreed that DKs just weren't up to the other tanks from a survivability standpoint. The buffs rolled in and lo and behold, DKs were cruising around tanking things with no problems. Then, it became somewhat apparent that DKs were a bit too good. Now the nerfs roll in and it's just a bitter pill to swallow. The community feels betrayed because they were given those new shiny toys, but then had them yanked out of their hands. I think this is accurate, especially the psychology. If we wanted players to feel good, the right way to design is to ship their classes underpowered and then buff them quickly. That might not lead to the best game but it might lead to the happiest community (at least until they saw through our cheap trick in about twelve seconds). Q u o t e: Feral druids are in a similar boat. They're quite powerful, but there's still a perception that they're squishy. Why? I have no idea. People really do have some silly bias against ferals. Granted, these aren't top-end raiding guilds, but there's enough out there that the myth carries enough weight that when people are at the character selection screen they think tank and roll warrior or paladin. When they see druid they think, "healer or moonkin." Feral druid tanks are somewhat rare, especially among some of the most hardcore, progression-focused guilds. Now, as I say below, hardcore guilds may be the most likely to stick with their current MT (who to be fair, is likely a warrior because these guilds have been around awhile). There are definitely Feral tanks out there -- there are a lot of WoW raiding guilds after all. But if they are as good in 3.1 and 3.2 as some players predict, then why aren't there more of them? Why isn't every guild recruiting one? It's an interesting phenomenon and I'm not sure I could adequately explain it. Q u o t e: Ok, what bothers me about this point is that you basically encourage people to reroll a class they don't like as much by keeping it over powered because it is under represented. Also, a large enough population will generally move towards a class that fits there play style best regardless of strength. You will always have a favorite is what I'm saying. If you built a formula to explain class popularity (which would be impossible, but stick with me) both power and fun would be variables in that equation. Play style is important, but so is strength. This is particularly true in PvP where players tend to have less class loyalty (though many of them still have a favorite). Q u o t e: GC you your self have worked on RTS games in the past, so allow me to pose a hypothetical question to you. Say you created StarCraft and found that 70% of the people that played the game enjoyed playing the zerg over the other 2 races, would you balance the multiplayer to favor the other 2 races and make the zerg underpowered because they are more popular? Or would you attempt to keep the game balanced especially if it was seen as a serious competitive game. Iff 70% of games were won by Zerg (which is different from what you said, which was just played with Zerg) then I would be 99.9% certain that Zerg were overpowered. That's what happens in RTS games. There is very little sense of investment in your current or favorite choice (though there is some). I don't think MMO games work the same way. In part I think that's because the MMO version of "games won" is so slippery. Is it succesful times a boss has been tanked by a certain class? Probably not. Q u o t e: You can easily make stats show whatever you want. My comment on check the "55-55" thing was because unlike the other 9 classes, DKs start at 55 which means more people are inclined to make them out of curiosity. This is part of the point I was trying to make. I don't think you can look at DK numbers and say "Clearly DKs are overpowered or clearly they are the most fun class in the game." It's possible either or both are true, but more than likely all you're seeing is that they're new, start at high level and have an awesome starting experience.
07/18/09 01:14:44 AM Patch 3.1.3 Ghostcrawler <Lead Systems Designer>	Q u o t e: 5) Blizzard has to be incredibly careful with warriors or they risk making them clearly dominant again. Backsliding all the progress they have made in tanking representation. 6) Blizzard also has no good excuse to not make warriors (or any other tank) on par with the rest. The Warrior population is not going away and should not be a tool for measuring balance. I buy your point 5, but not your point 6. If warriors were terrible or not fun to play, they would drop fast. Players will stick with them out of loyalty to a point, but only to a point. I don't think you can argue the opposite though - that lots of warriors means they are overpowered. However, if they were overpowered, that would lead to lots of warriors too. Logic is fun. Q u o t e: Leveling up a druid is probably one of the worst experiences in the game, according to most people I talk to anyway. A paladin only being slightly better. A lot of work has been done to ret to make it into a extremely good damage dealing tree. Which makes leveling one both fun and fast. Wow, I disagree. I found leveling a druid to be much, much easier than leveling a warrior once you get cat. My warrior leveled on a stack of health potions and a high repair bill. Paladins are trickier, because they have some really nice benefits and some really slow aspects of leveling. I will leave you with the tidbit that paladins are the least likely class to be abandoned at low level. What does that mean? I'm not sure I have any idea. Q u o t e: That said, druids have been nerfed EVERY single patch in this expansion. This is not QQ. Its just a matter of fact. It's a curious fact though. It argues that we never nerf them enough because we keep having to do it again and again. Does that mean we have a double standard and are too nice to druids? Be careful trying to use facts like this to prove anything. Number of nerfs or number of patches nerfed are not very informative values. Q u o t e: I think the answer is very simple to be honest. If the spec is overpowered, even if absolutely no one plays it, it should be nerfed. Same if a given spec is underpowered. One's performance in a raid should not be dictated by how many other people play the spec. What I'm trying to say is that over power and under power are entirely independent from popularity. I don't think they truly are independent though, not by a long shot. I can understand that viewpoint from a pure game design standpoint, but I also think if we gave druids a 25% dps buff and it stuck that you would see players flock to them in no time. This is more true of PvP than PvE, but I still think it's true of PvE. We saw rogue numbers decline in 3.0 when they were underpowered and they have since returned. Now maybe my hyopethetical example above never actually happens, but I sort of suspect it does. Q u o t e: If we're talking about cutting-edge guilds, and the community generally agrees Feral tanks are better, the first reason can't be it- most of them would prefer the better tank class. The second reason is plausible- an existent tank could reroll to the better tank class, but he's giving up his epic loot, his epic flyer training, his four "Gigantique" bags and his Traveler's Tundra Funmoth. And for what? So he can, after a lot of work, be part of the overpowered class du jour? When you have no idea how long that overpowered-ness will last, it's a risky investment. It is an interesting phenomenon that some of the most cutting-edge guilds are the least likely to change. Now, they certainly have the resources and mindset to change if that's required. If we made a boss that could only be tanked by a mage with a half Arcane half Frost build, they would mysteriously produce one. But they tend to be conservative. They have their roster and they know what works for them. If their traditional MT can beat a boss, they will probably do it that way, even if another tank would give them an advantage. They would only use the advantage if they couldn't beat it the way they wanted to (and this does happen). For less than cutting-edge guilds, they might see more of a benefit in switching tanks. And yet... these guilds are also the least likely to be able to attract amazing players with good gear of other classes at a moment's notice, and they are likely to see a much bigger improvement just by tightening up their game than they are by chaning their roster. This is why I often say tank balance doesn't have to be perfectly equal. It just has to be close enough.
07/18/09 01:14:49 AM Patch 3.1.3 Ghostcrawler <Lead Systems Designer>	Q u o t e: I'd appreciate more discussions like this - that center around actual in-game play. There's often too many assumptions when we discuss tank balance - and I include in that people's home grown spreadsheets that spawn to 50 pages. Almost nobody analyzes in game play to decide which tank to use. The encounters are all that matter. If you are the best tank on Patchgalon, it doesn't matter if you can't beat the first boss in the Coliseum. That's not to say Patch parses are useless. Just remember that they are descriptive not definitive.

Blue colored topics are started by Blizzard Posters.
Topics are sorted descending based on their last update (last blue post) time.
You can find links to the orignal posts in the top-right corner of each post.
Update frequency is ~3 minutes.
Stats can be found here