11 March 2010

Lame title is lame

Posted a few days ago on the WoW forums, have yet to see conclusive results! The dragon picture doesn't fit well in the blog style used when this entry was made, and I need to stop worrying about whether to leave the date and time at when the webpage was loaded or to update it before submitting. Generic blog is generic! Sorry, I got bored at the end of writing this one ._.

    MMOs change.


The recently announced update to the stat changes that are coming with the Cataclysm highlight the adherence of Blizzard to the principle of making WoW as fun a game as possible, for as many people as possible. The basis of the changes is to continue to provide a wide diversity of choices in the game, with each option having meaningful results. This design is limited only by the reality of the interaction of different options within the game's mathematical and social framework -- the hopes and dreams of people provide a guiding force, which is not always apparent in the direction it will take the game.

Chief among the failed dreams of WoW is the valuation of past achievements and respect for those who combat the menaces facing this world, although the lack of respect extends to PvP contributions as well. This is posted in the tanking class role forum because the tanking role is essential in establishing order in the realization of PvE goals, but mechanics affecting the success of and respect for the tanking role have implications for all classes.

A recent thread on this forum posed the question, is threat fun? The most basic and correct answer is, threat is fun when it offers a situation where the player can benefit from making the correct decisions and the risk from guessing incorrectly can be mitigated by other players turning a dangerous situation into an advantage when played well, and when the complexity of the situation does not detract from an overall goal that turns anything before it into a grind. In raid situations, threat is not fun when the only conceivable outcome of the tank losing aggro is the raid wiping. In dungeons, threat is not fun when it only increases the time that must be spent farming badges for the day when the gear from those badges is needed for the real objective. The hidden quality of Warcraft, and the basis of its immense popularity is that individual goals have a group benefit that may not be immediately apparent, but these results are disconnected from each other in too much of WoW's current design. This must change.


In the Hybrid Tax sticky in the damage dealing class role forum is this contextual definition:

Hybrid = can respec to fulfill a different role (damage, tanking, or healing).
Hybrid != can fill multiple roles at the same time.
Hybrid != has awesome, amazing buffs or utility.
Hybrid != pure. Beyond that, there are no shades of gray among hybrids.

It is not possible for any player to competitively fill multiple roles at once in current WoW, nor has it ever been possible. In the original design for WoW this was simply because the offensive capabilities or mana efficiency of healing classes was reduced to ensure PvP balance. In later expansions, with all classes now viable for raid dps, it is because the mechanics of the game do not allow effectiveness in multiple roles. This has harmful effects: no one except a pure tank can absorb a single hit from a raid boss, so raiding is designed so that no one has to, becoming predictable. There are no choices to be made in gear or talents when there is no increased difficulty from doing more damage and no options that provide an alternative benefit to other players. To provide content to those who do feel these restrictions are not fun, alternative versions of endgame content are created that can be done without single-role specialization that is more effective in the current design of the game.

The announced intention for Cataclysm is to increase the survivability of non-tanks in a raid setting, but two major barriers to this have not been addressed: the effect of diverse stats on item design, and the quantization of raid threat mechanics into predictable performance leading to further increase in specialization.

Stat weights in item design

The current item design mechanics of WoW are that an item with the most stats is one with many different types of stats, all of approximately equal proportion in the item budget. An item with only strength will provide a reasonable amount of dps to a plate user. An item divided into strength and crit rating will provide more. An item divided into strength, crit, and hit will increase damage by even more. An item with strength, crit, hit, armor penetration, haste, attack power, agility, and expertise will provide even more, even if the item budget has been divided into all these different stats, even if a single point of strength would provide more benefit than a single point of any other stat. Even if a class only values agility at one-fourth of the value of strength, a small amount of any given stat on an item is virtually free. Conversely, a single stat when increased takes up a disproportionately large share of the budget as it becomes larger than other stats. With several stats for the class role on every item and only a small portion of the item budget allocated for stamina, this is why dps has inflated so much in WotLK relative to health pools, since not only does stamina have a small share of the budget but dps increases polynomially with multiplicative stat mechanics such as crit, haste, hit, and armor penetration.

The current method of adding stats in an item's budget was fine for original WoW, where the main stats were Strength, Agility, Intellect, Stamina, and Spirit, and each class role benefited from only one or two of those stats. But that design is no longer sufficient in providing interesting gear choices, and must change. Recognizing the importance of roles, each type of stat should be more expensive when paired with other stats from the same role, and the groupings of stats for different specialties should be much cheaper when paired together. Avoidance is a special case as cheaply providing it on all gear would not scale well in PvP without counters, but this fits well with the intention of making the same boss attack challenging and counterable for tanks and non-tanks. Stamina in large amounts would take up a reasonable amount of the item budget when it is only competing against one or two other stat groups, instead of against five discrete dps stat types on the item.

Revising threat mechanics

If it's reasonable considering the possiblity of increased computational load on the server, revising threat and aggro would not only allow the rest of WoW's design to make non-tanks less squishy giving them more choices in gearing and talents, but would also further the respect for raiding achievements in WoW overall by increasing the accessibility to challenging raid content for players with non-skill 'handicaps' like scheduling or social obligations and by making mob behavior more fun.

  • A proportion of the threat for any offensive ability is shared between all mobs of the same social aggro group that are targeting the same player that the mob is targeting.

    • Best way might be to take a certain amount (like 30%) of the threat and divide it between said mobs including the mob that is the target of the offensive ability, but to drop the proportion of this splash threat that would go to the original target. (Another way would be to reduce the threat of the original attack by the amount of splash threat.) Examples:

    • Tanking two mobs and using a 1000 threat ability on mob 1. 30% of threat splashes, resulting in (1000 + 0) = 1000 threat on mob 1, and (300 / 2) = 150 threat on mob 2.

    • Death knight is tanking five mobs, a warlock uses an aoe that does damage equal to 500 threat on every mob. Threat splashes from each mob, with a total of (500 + [(150 / 5) * 4]) = 620 threat on each mob.

    • Warlock in above example pulls aggro from aoe, and all of the mobs except the primary dps target head for the warlock. Death knight uses ability that does 3000 threat on one of the mobs heading for warlock, causing (900 / 4) = 225 threat on the other three mobs making them attack death knight again.

  • When a mob, such as a raid boss loses aggro from the current target because they died or used a threat reduction move while below the top threat on list, instead of heading for the highest threat player on threat list the mob will use the 'effective' or 'apparent' threat which is decreased for players farther away. Implications:

    • Even if someone who is trying to tank is not at the top of the threat list, they can still gain aggro if they are next to the boss and the highest threat player is far away and running as fast as they can.

    • If the boss goes crazy and heads for a squishy, others nearby can react by either running away from that player (example: mage), attempting to heal the squishy and keep them alive but risking being the next target (example: holy paladin), or trying to gain aggro themselves if they have more survivability than the boss's current target and the tanks are far away or have had their aggro reduced by the boss (example: boomkin).

    • This mechanic can be viewed as a temporary restriction on the healing or dps capability of individual players that forces others in the raid to adapt and heal, and also as a temporary imposition of tanking capability on specific players in the raid with frequency proportional to threat generation when combined with unpredictable threat mechanics explained below.

  • Kiting a mob and being out of range will decrease threat, causing it to switch targets if threat is low enough. Aggro switch from kiting does not occur until threat is significantly below other players. If using the 10/30% rule in current mechanics, then boss will not switch until threat loss on the kiting player causes someone to exceed those thresholds, or effective threat could be used to establish a threshold that varies linearly with range.

  • Threat for all players on threat list is reduced proportionately more frequently on fights, maybe every time targets are switched. This increases the effect of threat-producing moves which are used in these situations, compared to if they had been used prior to the switch. Simultaneously, dps are forced to interrupt their attacks until a tank has established aggro, unless the dps is attempting to tank by pulling aggro off someone more vulnerable. Could be a reduction by half but the objective is just to keep visibility on recent events and retain volatility in later stages of an encounter, and frequency and reduction might depend on the design of a specific encounter that is intended to be more or less volatile.

  • Threat-producing events that are below the threshold for pulling aggro still have a chance to cause the mob to switch targets, deliberately reducing the amount of time any specific player is tanking the boss.

    • Provides smoother curve of risk vs reward for dps, instead of an infinitely sharp transition to 'raid wipes' by causing 100 more threat-per-second to push over the threat cap.

    • Allows balancing of class abilities so that high threat production and skilled play are required for tanks, instead of being given for free to avoid bad tanks automatically causing wipes when paired with bad dps.

    • Allows balancing of bosses with attacks that will not always one-shot players who are not tanks, by forcing tanks to combine tanking talents with hybrid tanking/dps gear to achieve optimum mitigation and tanking reliability ratio.

    • Simplifies boss drop itemization requirements by allowing more generic items and tier sets, with the optimum threat/mitigation balance to contain only a few pieces of specialized tanking gear when using tanking talents.

    • Each boss tunable with constants that determine how likely they are to unpredictably switch targets. Patchwerk might have no chance to switch targets unless someone somehow managed to exceed the threat cap. Battleguard Sartura might be an example of the type of boss designed to frequently switch targets.

    • Server load: instead of calculating on every attack, might total threat increase for each player during a certain time period, and periodically check these totals to see if boss switches targets. Special attacks by the boss, and possibly even autoattacks might be good times to check, as would drops in threat in the current target due to threat loss abilities or kiting as described earlier. Also gives some measure of predictability for unpredictable mechanic giving tension in fight. Simplest design for chance to aggro would be [(player's effective threat, based on distance) / (tank threat)] * (amount of threat player incurred since last periodic check) / (time since last periodic check) / (sum of [constant for size of significant physical attack, scaled to raid threat-per-second expected] + [(effective threat/tank threat * tps since last check, summed for all players in threat list including tank) / tank threat]) * (constant specific to boss, determining frequency of unpredictable aggro switch) * (expected frequency constant for the mechanic tuned for entire game). Chart:

      / \
      / pick from:
      ⌜---------------⌝ ⌜------------------------------------⌝
      |Constant in tps| | Everyone incl. tank | ====>
      ⌞_______________⌟ ⌞____________________________________⌟
      | [sum of]: (effective threat/tank threat) * tps
      | |
      V V
      (no switch) x rate constant for mechanic frequency, boss

    • This simple (my head hurts) design, with chance to draw aggro below the threat cap proportional to total threat multiplied by current threat-per-second, means that danger is proportional to the square of dps, while flat threat boosts like Misdirection and Tricks of the Trade linearly decrease the chance for a switch to occur. Other designs might give greater chance to draw aggro at low total threat levels such as after feign death, or have different low-fight-intensity behavior. This one will give a certain rate of aggro switch each time the check is made unless the division by time is removed to make it 'threat dealt since last check' instead of 'threat per second'.

    • When aggro switches to random player below the threat cap, all threat above the new target is flattened to the threat of the current target, as well as threat being halved. This is distinct from aggro switching due to the tank's aggro being reduced by the boss or by the boss's current target dying in that threat leveling does not occur in those situations.

    • The tank is not expected to be max threat in melee range. Offtanks should be trying to exceed the tank's threat and even reach the melee threat cap to switch aggro, allowing the previous tank to switch from mitigating damage to catching up in threat. If a ranged player draws random aggro or the boss uses a threat reduction move on the current tank, an offtank will probably pick up aggro before the previous tank does.

    • Threat above the tank for ranged players represents increased risk should the boss go wild, but will be floored and halved if any player unpredictably draws aggro below the threat cap. Healing threat-per-second from spam should exceed tank threat-per-second, so that healers have a reason to care about proximity aggro from high threat if the boss goes wild. A situational threat limit on one healer allows versatility from others in the raid.

  • Healing threat may benefit from a change if the server load can support it: when a target is healed, threat is distributed in proportion to the share of that target on mobs' threat tables. Intention is for aggro management in chaotic situations to be controllable on a smaller scale. Examples:

    • A paladin is dpsing a boss when an add appears. The paladin picks up the add and begins to tank. A priest heals the paladin for the equivalent of 4000 threat. Prior to the heal, the add has 10,000 threat on the paladin and no threat on any other players, while the boss has 160,000 threat on the paladin out of a total of 3,200,000 threat on the entire raid. The heal causes (4000 * 100/105) = 3810 threat on the add, and (4000 * 5/105) = 190 threat on the boss.

    • A shaman is healing a dungeon when the tank loses control of one of the five mobs and it attacks the shaman, who heals herself. Prior to the heal, the shaman's threat on the mob is just slightly above that of the tank, on the current party focus is 30% of the tank's, while on the remaining three mobs the shaman's threat is slightly below that of the tank. Due to social threat splashing described above, the three dps in the party also have threat close to that of the tank on the four unfocused mobs so much of the shaman's healing threat is correctly distributed to the focused target, but with the mob no longer targeting the tank splash threat does not reach it and without intervention the proportion of heal threat going to the mob targeting the shaman will gradually increase until it really, really hates the shaman! (lol) But heals on the tank will lead to less threat going to the mob attacking the shaman due to the lower proportion of threat that mob has towards the tank.

Overall, the lack of viability for a hybrid playstyle in current WoW is due to both the linear dropoff in effectiveness between the extremes of specializations for different roles and raid encounter mechanics that would not support any such hybrids even if they did exist. The ideal raid would neither be composed entirely of players that were geared and specced as hybrids nor one composed entirely of specialists in one role each. It would have a mix of both, with the specialists filling one role for the entire duration of an encounter with greater effectiveness than a hybrid trying to do the same, while the hybrids filled gaps as needed either due to requirements for or impositions on class roles on the raid as a whole or on individual players. Itemization for all classes would be easier too, since standard gear would support all roles for players with the proper spec, with only a small drop in effectiveness compared to wearing gear meant only for that spec which would also mean more risk in a raid.

World First

Moving beyond tanking, there are more things that can be done to increase the fun of and respect for incursions against enemies residing in the instances of WoW. Many of these possibilities support each other in the attitudes they would encourage towards accomplishments, but some of them may sound quite radical compared to what has gone before.

Extend the concept of item scaling beyond Heirloom items. Gear at the level cap causes considerable problems, including but not limited to, the mitigation of plate continuing to increase its advantage over cloth, fixation on weapons as a source of imbalance in both PvE and PvP, and the almost total reliance on gear for casters as the source of scaling above level 60 with almost no increase in the base power of spells. A portion of the scaling for both weapons and armor should be shifted to level, with the effective result that the dps (or spellpower equivalent) of weapons and base armor value of items would increase only slightly with each tier at the level cap, and would only reach the 'expected' values for their iLevel in the next expansion when items of that power have become standard. Moveover, there is no reason not to extend this partial scaling to all items regardless of what their original value was, to extend the usefulness of items that are no longer the best available. If, for example, two-thirds of item scaling from items at the cap was from character level, then...

  • Thunderfury would have 77 dps at lvl 70, 118 dps at lvl 80, and approximately 189.5 dps at lvl 85.

  • An ilvl 200 epic 1H weapon would be unchanged at lvl 80 with 143.5 dps, but the dps (not stats) of weapons past it would be reduced, with the base damage of Bloodvenom Blade (Heroic) from Deathbringer Saurfang on 25 Heroic being reduced from 250.6 dps, or 175% of the ilvl 200 weapon, down to 179 dps or 125% of the iLvl 200 weapon. Bloodvenom Blade would not do the 'full' dps of an item with its ilvl and item budget until lvl 85, where presumably a blue weapon will be equivalent and have similar dps and stats. At lvl 85, the dps of the ilvl 200 epic 1H would have increased to 215 dps, still 35 dps less than Bloodvenom Blade (Heroic), but with greatly inferior stats to either that weapon or the lvl 85 blue which would match it.

  • Armor would be scaling similarly, so a Sunwell tanking item would have had reduced armor at the level it was obtained, and only reached its current armor value at lvl 80, while any less powerful gear that had not been replaced in Northrend would have still been adequate with the armor value that was only slightly below current Northrend blues, although the Outland gear would have had significantly less bonus stats.

  • In contrast to weapon dps, the bonus spellpower on weapons would not increase from leveling, because spells would improve while leveling from the automatic increase in the base power of spells with the removal of ranks in Cataclysm.

Note that the primary benefit of removing spell ranks is to remove the WotLK paradox of a spell staying the same in power but increasing significantly in mana cost from level while avoiding the TBC situation of spells being downranked to avoid mana costs while still reaching nearly the same effectiveness from scaling. This could be avoided, if Blizzard so desired, by scaling the mana cost slightly upward with levels (to keep spell effectiveness relatively stable while leveling post-60) while increasing the base amount of an ability proportionately with the mana cost, increasing the coefficient proportionately with the mana cost, but decreasing the coefficient proportionately with the 'expected' cost of the spell should the spell rank have been obtained at that level. This has the overall effect of giving all spell ranks a coefficient proportional to their mana cost.

The other situation is scaling player levels, and item levels, down to match the level of an instance. This has great advantages over creating Heroic versions of a dungeon, such as the Deadmines, at the level cap:

  • A high-level player can do instances in a challenging way with their friends who have just started the game, or even friends who are in mid levels.

  • The experience of attempting to overcome the boss with the storylines present at the appropriate level is preserved, with the boss always being as hard as the story asserts them to be for anyone seeking to challenge their reign.

  • Preserves the challenge of older raid content, making any achievements in them understandable even if there are no longer enough players at the appropriate level to do the raid in the way it was intended without overgearing or overleveling it.

The development necessary to remove spell ranks makes it much easier to scale player levels down to match an instance. Each instance would have have the highest level allowed in the dungeon determined, probably not higher than the highest level boss in the dungeon or even one or two levels lower: Deadmines might limit all who entered to lvl 19. However, all abilities and even talents would still be available, just with reduced capabilities to match the player's level. Gear would also be limited to the highest quality available in the instance, with different item colours scaled to match the the selected item level so going to the Deadmines in lvl 60 epics would not be easier than going to the Deadmines in lvl 80 blues. The stats on an item would be scaled as a vector.

Just as significantly, this would also apply to instances at the level cap. Any equipped items would be reduced to a level appropriate for the dungeon, which for many of the lvl 85 dungeons might be the first tier of Cataclysm epics. For a raid instance, it would likely be the tier of that raid or the next tier. Any items of higher power would provide reduced benefit inside that raid.

However, items should not be the main progression mechanic in Cataclysm for the more casual guilds that have traditionally found them necessary. ICC attempt limiting and current attitudes towards the ICC raid buffs has shown conclusively that a new method is needed for raid progression beyond just gear. The fundamental question is how to allow players to assist and fight alongside their 60-year old grandmother who may be very skilled but does not have fast reflexes to overcome the challenges of the game, while also providing the same story to a guild composed entirely of seasoned players at a difficulty that will take more than just a single week to complete?

If an encounter becomes easier over the course of a game to players with the same gear and abilities, whether because the encounter was nerfed to reduce the hitpoints of the boss or by giving a zonewide buff, it becomes impossible for players newly attempting the encounter to understand the achievements of those who defeated the boss previously, and this creates conflict in the playerbase in the sense of value. The individual goals, of defeating an encounter as soon as possible, come to interfere with the social objective of extending the boundary of possibility for the entire playerbase by showing that a certain difficulty can be overcome. This conflict can be removed, without increasing the imbalance and distance between tiers of raid gear, by balancing raid progression on individual character advancement specific to the raid instance.

This comes in the form of powerful buffs that draw on the essence of a particular raid instance or that focus on the weaknesses possessed by enemies in that instance, but these advantages can only come at the cost of a lingering weakness in said instance which will only go away completely after a period of weeks or months. Such is the price of power, but with each passing day the resistance to the malaise permeating the raid location will become stronger, without the loss of the power that caused it. In other words, it is a raid progression mechanic on an individual level where the benefits are not realized immediately. Given the large number of customization options in WoW, it may not be wise to make it too complicated, but here are some useful patterns:

  • "Increases damage of <ability> against <creature group specific to instance>."

  • "Reduces damage of <ability type specific to instance>." (see: Shadowflame!)

  • "Blackwing Lair Sickness - damage, healing, health reduced by 5%." (Up to 30% reduction if every single possible buff is obtained in a single week with no time to decay.)

To be fair, the level of the corruption would not decay while inside the instance, with the value calculated by adding all time spent in the instance for the week but it would have a minimal effect for all but the most hardcore guilds when an instance first becomes available. There are many possibilities about how to obtain such dual buffs/debuffs and how to present it, but it is clear that any benefits should be themed specifically on the raid instance in their description as well as functionality. Overall, as regards to raid progression tiers, it would offer much more of a sense of progression and achievement if raids did not feel the need to impose on everyone to become fully geared in epic equipment before even attempting raid content, as it devalues the sense of achievement in the playerbase. A mechanic like this, as well as quest-based mechanics such as a collection quest that would take weeks for the entire raid to do but only a day for a new guild member later on, offers the necessary slowing of progression without the need for easily obtainable epic upgrades at an expansion's launch.

The main discontent with the raid paradigm of original WoW was the imbalance that raid gear caused in PvP as a result of the dependence on gear as the almost sole progression mechanism, also causing many other problems with guild hopping and leeching between tiers. WotLK has offered its own totally different capacity to remove the ability for gear to provide a sense of achievement. If WoW focuses on new mechanics for raid progression with less reliance on gear to provide this limiting factor, the potential for blue-geared players to earn the respect of raiders and even contribute to raids themselves is much improved, and there becomes absolutely nothing wrong with an endgame where high-end epics are not easily grindable with 15 minutes per day of faceroll aoe. In this case, the only obstacle to the community valuation of raid achievements is the issue which has become a hallmark of WotLK casual discontent: endgame raid size and the disassociation of story from achievement.

Obstacles to the 40 person raid

It is possible to return WoW to the style of requiring 40 people to raid an endgame instance. It does not mean tuning encounters to be easy. It does mean fixing hybrids and threat to allow an understanding of one part of an ongoing fight without simultaneously having to observe every other part of the fight, and giving players and guilds the necessary tools to defeat content even without being able to form the full raid from a single guild.

The majority of guilds in WoW are small, social guilds. People are in these guilds because they like to associate, and adventure with, people they know and are comfortable with. There is no reason these guilds could not participate in endgame content if the players are skilled and are able to complete challenging content on a smaller scale. Scheduling and instance locking are one complication that prevented this from happening frequently in the past, but so were the motives that commonly inspire players to raid at all. Many casual players feel that raiding does not provide sufficient context to be able to feel that the individual motive, of wanting to raid, supports the group objectives of valuing goals that all sufficiently skilled and determined players are able to achieve. This is both due to the lack of game-supported rationale for visiting a raid instance in the form of quests, and also the impositions of scheduling and restrictive specialization requirements in the current functional framework, with the declared objective of many raiders being simply to collect items. There is nothing wrong with this, but with the community attitudes that inevitably emerge, currently gearscore obsession and in the past PvP advantages, many simply do not view gear as a worthy goal and the game provides no other goal for raiding that justifies raiding's requirements and constraints in scheduling, talent spec and gear.

In the context of WoW, benefit to the group and others is provided by the promise of a questgiver that a certain action is worth rewarding, and the description by the questgiver of how the objectives of the quest relate to our own goals and values. The only significant questline related to an instance in WoW was the forming of the scepter to open the gate of Ahn'Qiraj, and most of that took place outside instances. Blizzard in Cataclysm is already seeking to reduce the aforementioned, non-skill restraints on raiding of talent spec and gear by allowing more player choices to be viable for a raid, but there is an overwhelming need for interesting, deep, and relevant quests for the raid instances of Cataclysm as well as better tools to coordinate scheduling so that 40 person raid instances can, once again, be a reality.

Part of the difficulty of coordinating multiple guilds to attempt content is not only knowing what the completion status is of a partially-cleared, shared raid instance, but also on the status of other people bound to that instance and their willingness to continue it at a certain point in time. The tools provided to the community must be just as effective as the highly successful Looking for Dungeon tool in allowing guilds to communicate both short-term and long-term raiding needs and desires to other guilds and players. Just as important for any raid instance, however, is avoiding drama in the distribution of rewards, and while social guilds very rarely have any drama it is important to recognize that the social structures within a raid may need their own loot distribution authority to ensure timely and efficient operation without detracting from the story of the instance. For any 'PUG'ed members of the raid, being able to efficiently implement the 'Gold DKP' loot method within a subset of the raid might help.

Reviving the heart of the World of Warcraft

The congruence of individual goals with group values applies just as much, if not more to PvP, and as with the case of PvE these objectives are often hidden and are only revealed once they have been lost. When was the last time a random roleplayer on your server charged into battle yelling, For the Horde?

Arenas do not have an objective relevant to the interests of the group, except for the minority of the playerbase who do not view the roleplaying aspects of their character as an essential part of their experience in this world. The challenge of seeing how high a rating one can reach in a controlled competitive environment is something that many in WoW only did for one season, since further participation beyond that tends to prove little more than that you are interested in obtaining better gear. The world of Azeroth needs a return to the sense that by defeating a hostile and aggressive member of the opposite faction, you have helped your faction.

A rank system may be the answer. But while the past one was based on measuring PvP activity levels and rewarding the greatest champions with superb equipment, this one must be based almost entirely on creating a standard to measure individual victories against the opposite faction, with rewards only a small bonus for time and effort spent. This is, completely coincidentally, somewhat similar in form to another PvP game developed by a different company... but with key differences!

First, battlegrounds and what can be obtained from them. It is possible to measure skill in a mixed population of skilled and unskilled players by consistency of victory when the population contains a relatively high proportion of skilled players. For BGs in a team-based, premade bracket, a premade of skilled PvPers will have no trouble defeating a PUG premade formed in trade channel, so by ensuring high participation by skilled teams in that bracket, any item rewards that are obtained from it by the average player will be well-earned. But farming kills is not the same thing as completing battlefield objectives, and it may be advantageous in encouraging a competitive spirit both within and outside of BGs if these were separated, to allow players to earn rewards by accomplishing the clearly defined goals of a BG, while contributing to their faction's advancement on an entirely different level with PvP kills. Key points that would encourage wholesome and morally unencumbered slaughter inside and outside of BGs:

  • gear matching should work on the total number of players on each side, so if a party of five players with the best gear in the game join via the Looking for Battleground tool and are placed in a BG, their faction's team would enjoy a handicap of two or three players as those players already exist in the form of overpowered gear. To prevent unbalancing of the system in a way that isn't an issue with LFD, just prevent anyone from transfering gear from storage to their bags after receiving an invite to a BG.

  • The existing faction-specific BG rewards purchased with honor might benefit from being updated to a quest system dependent on achieving a certain number of victories from that BG and looting the corpses of the opposite faction, with any epic rewards requiring the premade team bracket for quest credit.

  • PvP competitiveness and community is built by interaction with other players and teams, not by interaction with a rating system. The existence of a premade bracket and restriction of forming a full successful team prior to entry into the queue validates the group contribution to the PvP conflict. The success of the bracket is formed around this, with any high-quality item rewards only ensuring the validity of the bracket and taking advantage of measurable difficulty in PvP.

  • May help to merge the mostly dead 'twink' bracket of no-XP with the team premade bracket.

    PvP rank

  • A smaller number of ranks to ease recognition. No one cared about ranks in the old honor system except Ranks 12, 13, and 14, and not only because they had teh epicz.

  • Rank is based on quality of play, not playtime. Honor loss on death and honor gain when getting kills, but opponents more than a few ranks lower do not reward honor nor do they lose honor on death, similar to the XP cutoff against lower levels.

  • Significantly flat progression for honor awarded on death for higher ranks, with main determination of skill based on rank being that the lower rank has no risk and low chance of reward, instead of a more extreme situation of the lower rank having normal risk but extremely high reward on a successful kill (encourages zerging).

  • Rank is understood to be a function of both gear and skill, and victory over a high-ranked opponent is seen as a more significant contribution to one's faction whether that high rank resulted from gear, or from skill.

  • A small number of rewards might be available of a PvP-oriented nature, but not intended to become necessary for PvP and not designed to be replaced with new rewards as PvE tiers are released. Rewards in a progression-based system should not be required to progress to the point of be able to obtain those rewards, so better if rank rewards are for less visible slots like accessories.

  • Any PvP gear identity, of visible nature such as weapon or shoulders/chest, would be obtained either from premade or rated BGs, or from arenas if they still exist.

  • Honor gains from PvP kills should also be seen as benefiting the faction in a concrete and visible way. One way would be to increase the rewards from BGs and world PvP objectives, in proportion to the honor gained compared to the other faction in a 'tipping' system similar to WG. Due to design of ranks, eventually both factions will get the same honor per week, even if it means all Horde have to be of highest possible rank and are awarded no honor for killing any low-ranked Alliance they encounter.

  • PvP rank as an indication of skill must take into account group size and dynamics in PvP. More logical to modify the honor loss on death, instead of modifying the honor gain which is more visible and goal-oriented.

    • Dying when heavily outnumbered would cause reduced honor loss.

    • Dying by being killed by someone who was heavily outnumbered by your allies might cause increased honor loss, considering other factors such as their rank and your rank.

    • When killed by a mixed group, dishonor is only awarded for the opponents who were close enough in rank to obtain honor, not for the opponents who were much higher rank (implying much better gear or PvP experience).

    • Best way to calculate odds in PvP is with 'attention', similar in mechanics to distributing threat. The reciprocal amount of attention determines the supposed difficulty in inflicting damage, so if a group of five players are engaged in combat with an equal group of five other players of the opposite faction, there is no modification to group.

Other PvP issues

With the introduction of flying mounts to Azeroth, this mechanic will significantly affect how the world is viewed. Several things could decrease the impact of this escape and afk mechanism on PvP servers, as well as increasing the enjoyability and realism of flying mounts overall:

  • Inertial physics for flying mounts. Should not be possible to hover, even with the gyrocoptor. Somewhat similar to the Brewfest quest ram, keep going forward, and climbing should be significantly slower than level, gliding flight. Flying should be a cinematic experience.

  • Might benefit to add a flight timer, with forced gliding and landing or even forced dismounting at the expiration of the timer.

  • Nets vs flying mounts and druids, that immobilize, force landing, and prevent movement and escape.

  • 'anti-aircraft' stations, possibly portable, where by utilizing a scope similar to an Ornate Spyglass one can view flying targets from a longer distance and immobilize or disable them.

  • A strict limit on the vertical detection range of gathering abilities to allow lower-level characters a competitive chance at nodes.

  • Disabling 'click to move' for flying mounts, as a quick Google search reveals that this mechanic is used by gather bots to avoid Warden detection.

  • Ground mounts should be disabled when diving beneath the surface of a body of water.

  • Redo spell resist system so the marginal advantage of higher amounts of resist is not so unbalanced.

videos used

World of warcraft alterac valley
So we pwned this funeral by Serenity Now
Teremus by Punchcat
BWL - Drakes in the Shell by Millenium
Yashe's Domination II
25k Ambush lvl 70 by Nition
Ahn'Qiraj The Movie by Jack
Fanaticism by Jamaz
KotS Ganker in the Westfall
2.4.0 PTR bug by Mewlyn
Holy Power Priest PvP by Panda
Vurtne Frost Mage PvP 1
Extreme Solo PvE by Demia
40 shamans vs 40 paladins preview by Ressorius
Feel the Pain vol4 by Niar
The Chosen Ones by Clan Rum
Paladin - World of Warcraft beta
Revenge of the windfury by Creed
Ahn'Qiraj - Eranikus and Maws by betrayal
The Last Ovski by Gegon
Shockadin goes to arena by Avalus and Oscuro
Polzielol : The True Story by Polzie
One Man vs Onyxia by Nabor
5 rogues vs Ironforge by Ludenben
Wormie hunter PvP
Candystriper 2 Mage PvP
Burn 4 the master of pain
Polzie PvP Part 2
How to get into Mount Hyjal by Shétàn
Spelldmg paladin by Dernos
New Ability by Gegon
Path of Blood by Grim
Shivan Frostmage PvP 2
Kazzak in Lakeshire
Tales of the Past III by Martin Falch
하이잘 성기사 홀리인페르노 (reckoning bomb PvP)
Hole of LoL
Szene Event Pinke Gnome
Maydie grand marshal PvP
WotLK Beta paladin preview by Kamil
Fist of the Heavens by Jamaz
Laughing Skull Horde Ironforge Raid
Lvl 70 warlock dps TBC beta
The Craft of War : Blind by Percula
TBC Mage preview by Gidgey
Melee Hunter by Kardonnis
The Memories of Sun
AFK by Sagacity
Solo BRD Emp by Gaeowyn
Unstoppable by Aelli
Onyxia Solo by Oniji
Awesome WoW Tricks by Chweet
My first wow PvP video by SA
Lorelol by Sykskar
DK Solo Karazhan by Mionelol
속력 frost mage PvP
Zero keeper Yogg-saron by Stars
Reflex is best by Grievous
A Night in Ahn'Qiraj by betrayal
Total Annihilation by Grim
Divinity II by Avengence
Polzie PvP part 1
scharlachrot by Patrizia
Alexis, Priest PvP by Francis
Soloed Chess Event in Karazhan, 70 Rogue
Paladincraft Ultimate by Grievous
Alliance Most Wanted by Death and Taxes & Death Dealers
Baskar the Hero
Run by Niallith
Fight against the Twilight Corrupter by betrayal
Zalgradis Paladin PvP 3
Circles by Shweex
Gwilldom vs AV Queue
Jimmy the World of Warcraft story
Clash of the Ovski by Gegon
Faxmonkey vs Onyxia
Divinity mage PvP by Avengence
Feel the Pain vol 5 by Niar
Vurtne 60-66 Mage PvP
Ovski Strikes Back by Gegon
Learn 2 Play by Myndflame
Further Proof Shamans are Overpowered
Vicious Mage Clazzi
Naxx Heroic Military Winge solo by Pai
Hunter solo Azuregos by Bigfluf
Untouchable 2 : 5-man Gruul by Gaeowyn
Millenium vs Ragnaros
Twin Emperors cinematic by Novakaine
Mantrid the Ninja
Tales of the Past by Martin Falch
Tales of the Past II by Martin Falch