Gamasutra: The Art & Business of Making Gamesspacer
Rethinking the Trinity of MMO Design
View All     RSS
October 22, 2014
arrowPress Releases
October 22, 2014
PR Newswire
View All

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Rethinking the Trinity of MMO Design

December 17, 2009 Article Start Page 1 of 4 Next

[Experienced MMO designer Brian "Psychochild" Green pulls at the MMO trinity -- Tank/Healer/DPS -- to examine whether or not this pillar of combat design can be pulled apart, modified, or even changed fundamentally.]

Designers, like many people, tend to fall into habits and patterns. Our games fit into specific genres, so they tend to be similar to other games in that same genre. Best practice is to borrow systems that work well in other games. However, sometimes a system gets enshrined into the conventional wisdom; once that happens, we rarely turn a critical eye to these systems and see if they are really meeting the design requirements. MMORPGs experience this type of "inherited design" for many gameplay systems.

The Trinity of Core Roles

One common design in MMORPGs is the "holy trinity" of class roles: Tank, Healer, and DPS (or damage dealers). As most games are about combat, these roles are about how damage is handled: Tanks can mitigate incoming damage from enemies, healers restore damage done from enemies, and DPS classes do damage to enemies.

Characters often require specialization: a superb healing class may not be able to do good damage without a significant change that hurts healing ability, or they will often not want to do damage because they need to save their resources for healing others. This trinity of classes forms the basis for most group-based encounters in a game.

There are other roles of characters possible in these types of games. For example, "crowd control" (CC) classes can temporarily take enemies out of the combat, and "buffer" classes can use abilities that enhance the abilities of other classes.

These additional roles are often combined in a class with a primary role: a healing class might get good buffs, a DPS character may be able to control extra enemies, etc. These other roles can also be taken care of by the core roles: a secondary tank can keep multiple enemies occupied if crowd control is not available, for example.

Finally, the DPS role sometimes has specialized categories: melee, ranged, single-target, area of effect (AoE), etc.

Our design goal for this article is to look at alternatives to the trinity design. First, we must understand the details of the trinity design, what design goals it accomplishes, and what type of design could replace it. For this discussion, we'll make a few assumptions:

  • Combat Focused. One of the main features of our theoretical game will be combat, and our class design will focus on this aspect of the game.
  • Group Focused. Although solo play can be an option in the game, there will be a strong focus on group encounters at the high end of the game: dungeons, raids, etc. in typical fantasy games. So, we will want different classes to still work together and complement each other.
  • Commercial Game. Our design must be for a viable commercial game. The design isn't going to be highly experimental or too off-the-wall to the point it scares away our potential player base. (We can still dream of making our ideal game full of our craziest ideas once we hit the big time!)

For this article, design is the only focus. Technical limitations tend to be specific to each game and therefore of limited use in a discussion. Other vital issues, such as community management of player expectations, are also beyond the scope of the article.

The History of Classes and the Trinity

It is useful to look at the history of character classes in RPGs to trace the origin of the trinity design. Classes were part of the first well-known RPG, Dungeons & Dragons (D&D). Each character had a class that determined his or her abilities: equipment that can be used, hit points, magical ability, etc.

Each class was based on a fantasy archetype, but without explicit roles; there was no rule that a Fighter could only absorb damage (be a Tank) and not be an awesome machine of death (be DPS). Each class had signature abilities, but statistics and options allowed characters to fill a variety of roles despite their class.

Over time, certain patterns of gameplay emerged: for example, the physically fragile low level wizard might prefer to be protected by the heavily armored fighter instead of rushing forward to engage enemies on the front line.

The MMORPG EverQuest (EQ) had the concept of classes and started the focus on the trinity of core classes. Inspired by a text MUD that was based on a D&D game setting, EQ had a class design that borrowed heavily from the older D&D system.

The core trinity of roles rose to prominence in high end raiding: a character tanked the boss, healers had to keep the tank alive, and other players hurt the boss without drawing too much attention to themselves. Each class had a single role it did well: Warriors were always Tanks, Clerics were always Healers, etc.

World of Warcraft (WoW) borrowed from EQ's class design and refined it. It added different talent trees that allowed a single class to specialize. WoW's expansions allowed some classes to fill multiple roles through talents; a Druid is able to become a Tank, a melee DPS, a ranged DPS, or a Healer based on equipment and talents chosen. A single character is usually expected to only fill a single role in one encounter.

Paper RPGs have been developing during the years as well. The fourth edition of D&D comes full circle, and the venerable paper RPG has been inspired by MMORPGs. In the latest edition, classes have gotten explicit roles as part of the class description, although it doesn't quite use the same roles as the trinity.

Other systems have gone away from classes and roles; for example, GURPS and the White Wolf's World of Darkness have had more flexible character development systems that rely more on lists of skills that any character can take rather than on strict classes and roles.

Article Start Page 1 of 4 Next

Related Jobs

Petroglyph Games
Petroglyph Games — Las Vegas, Nevada, United States

Illustrator / Concept Artist
DoubleDown Interactive
DoubleDown Interactive — Seattle, Washington, United States

Game Designer
Zindagi Games
Zindagi Games — Camarillo, California, United States

MOBILE Art Director
Treyarch / Activision
Treyarch / Activision — Santa Monica, California, United States

Senior UI Artist (temporary) Treyarch


Brian 'Psychochild' Green
profile image
Not sure why there's an EQ2 theme for the images with the article; I didn't pick them out. ;)

If you're interested in other discussions on game design, visit my professional blog:

Simon Carless
profile image
I think they were probably attractive panoramic MMO images, and you do mention EQ a bit, Brian :P (Also, Brian's blog is good and people should check it out.)

Mark Venturelli
profile image
When I talk with other designers, most of them overlook or are not familiar at all with Guild Wars' amazing Skill/Class hybrid role system. I think it's pretty unfair, but I can understand how this system wasn't used in many other games: it's ridiculously hard to balance and implement, and only super-designers like the guys in ArenaNet can execute it as elegantly as in Guild Wars. Anyway, that could change if more people tried new things in their RPGs and MMOs.

Joshua McDonald
profile image
While I'm very much on board with moving beyond the tank/healer/dps system, I have to disagree on your proposed solution.

My concern is that the result of attempting to implement what you suggested would simply be a dungeon crawler of yesteryear, such as Diablo II, Sacred, or Silverfall. These were reasonably fun games for their time, but there was no real teamwork. Playing with friends simply meant that you were chopping up bad guys in the same general area and occasionally cast a spell on each other.

The design challenge, in my opinion, is not just interesting character design but interesting team design. I'm a big fan of games that implement what I call "1+1=3": Games where the whole is much greater than the sum of its parts. The current crop of MMO's succeeds in this but fails in that battles are too repetitive and there are minimal benefits to making a character that doesn't work exactly like all of the others of your class/spec. Essentially, everybody tries to build the same team and uses it for every encounter.

I have a lot more I could go on with, but I'll wait to see what kind of response I get first.

Bernardo Gomes
profile image
Brian you made a good point here, but I think that most of the designers fall in the sacred trinity thinking on time and cost to develop.

As you sugested, The trinity doesn't work well on sci-fi based MMO's but it's there at Star Wars Galaxyes and will be on Old Republic.

Here's my 2 cents:

What I think that would work will be a multiclass aproach like a Warrior/Priest or Mage/Warrior and so on.

Philip Daay
profile image
The article is fairly definitive on the three roles. Great job. I do think you might have overlooked one crucial aspect of MMO's in your discussion. For whatever reason, MMO's have always focused their gameplay around the archetype of a group of players defeating one very powerful boss-type monster, capable of killing non-Tanks in a few hits. As you infer, this comes from our pencil/paper AD&D days...the brave party fights the evil dragon or sorcerer. But, this very paradigm literally forces the design evolution of MMO's to give Tanks more mitigation and taunts, Healers powerful cures and gives DPS their spike damage.

But, who says that MMO content must focus on a group killing one powerful boss or elite?

Imagine if we inverse the archetype. A Fantasy MMO is published where one player or small groups of players must now overcome huge hordes of less powerful monsters. This assumption will force design changes away from the three roles. Namely 1) Tanks become irrelevant since the whole party can be attacked. 2) It moves the game design axis away from Tanking/Healing to the DPS classes, placing a premium on Area-On-Effect. 3) Game designers will feel pressure to give all classes mobility, survivability and/or heal.

I'm not saying such a game would be superior to WoW/Aion/whatever, I'm just pointing out that there are underlying archetypes in the very nature of what we consider a standard raid fight in an MMO. Those archetypes are dictating our evolution to some degree, and you can't move away from the Trinity unless you change the archetype. It's such a basic design decision, but it affects everything.

David Wipperfurth
profile image
I was about to write a condensed version of what Joshua McDonald posted until I read his post. Nice work Josh, well stated.

The only one other thing I was going to say was that the nicely written slow build-up of this article, whom's goal was originality, ended in a suggestion so vague it could be used to describe over a half-dozen games already in existence.

You had said at one point that the downside of a radical idea would be that it would take a lot of time to develop. Still, how much better would this article have been if the result was a radical suggestion. I'd like to think that these kind of blog articles would be an excellent place for Brain Storming, where creative ideas could be nurtured and distilled without the loss of many man-hours or ego. Yet all I've found so far are these sort-of 'safe' articles. --Which definitely have their merit, this was an excellent article, but still very 'safe'.

Jorge Bilello
profile image
I always thought that the system used on Dungeon Siege where players don't have any ability but they grew on them as they make use of them was very good and "real".

Joe Cooper
profile image
Changing the context can toss things up a bit.

One of my favorite games is Civilization 3.

The tiles are positions you can hold and capture, and there is no healer class. From that you get these four:

Attack - Optimized to take a tile

Fast - Likewise, but will withdraw if losing, and can make multiple moves if terrain permits.

Defense - Optimized to hold a tile

Artillery - Strike neighboring tiles and augment defense.

Workers - Can build a fortification.

One interesting thing however is how the emphasis on differing roles _shifts_ through the game. At a certain point, you don't have an "attack" unit and only have a "defense" unit - the infantry - and your attack and fast units are no longer able to take infantry's positions, so you have to bulk them up with lots and lots of howitzers (your artillery unit). Eventually you get tanks which are both fast units & attack units, and bombers which are a sort of artillery... It shuffles the deck a bit.

Owain abArawn
profile image
I'd like to see MMOs get away from the fantasy motifs, and move more towards realistic historical settings, but where players would still interact in meaningful ways. So, a game based on the setting in the movie 300 would be interesting, with players using the Greek phalanx infantry tactics or perhaps the more lightly armored Persians approach. Another option might depict Roman Legions against barbarian hordes, or the medieval warfare with Knights and so forth, but no casters. Added into this where historically accurate, perhaps archers, calvalry, siege engines and on and on. There is plenty of fertile ground to be covered without always having to fall back on the same tired tank, healer, nuker formula. Surely, given the span of human history to draw from, developers can come up with something new.

Luis Guimaraes
profile image
The trinity is there because that's what MMO players want.

Bart Stewart
profile image
Brian, I think I've posted some similar thoughts on your web site in the past, and it looks like we're still on very much the same wavelength where the "holy trinity" is concerned.

I've written a couple of detailed essays related to this subject, so I'll just point to them: "The Archetypal Origins of MMORPG Group-Combat Roles" at
-of-mmorpg-group.html discusses where the four archetypes (Fighter, Mage, Healer, Rogue) come from (and, as you do, calls for the restoration of the Rogue archetype!), and "An Alternative to Aggro" at
native_to_Aggro.php discusses how the concept of aggro is at the heart of why the holy trinity keeps showing up in MMORPGs, and proposes a new way of thinking about how PvE enemies can target players that allows other roles to be developed that are a better fit for a game than tank/DPS/support.

Beyond that, I think this article is excellent. A lot of people complain that they're tired of the tank/DPS/support model, but because they haven't thought about why it exists and what goals it serves, they are rarely able to propose practical alternatives. This article does a very good job of clearly describing the basic reasons why these three roles seem to click so well with MMORPG players (and developers).

In particular (from my perspective) this article makes the crucial point that the tank/DPS/support roles exist to serve the goals of "combat" in "groups." What this means is that if you're willing and able to question either of those goals, you no longer have to be stuck with the holy trinity.

Not doing grouped gameplay would be a little odd for a MMORPG, so grouping probably needs to be retained as a design goal. But suppose someone created a MMORPG that wasn't almost entirely about shooting other avatars in the face and taking their stuff -- what if a MMORPG was designed to provide combat and non-combat content in roughly equal measure?

Imagine a MMORPG that, while including plenty of interesting combat, also offered challenging crafting puzzles and emotionally engaging stories as core gameplay elements. In such a game, where force-based destruction is not the overwhelmingly central motivator for role design, why would you need to clone the tank/DPS/support model? What other grouped roles become possible and practical when the focus of a MMORPG shifts from being about zero-sum destruction to non-zero-sum cooperation and creativity as gameplay?

Yes, that's a bit out there. :) Consider it a design challenge....

Brian 'Psychochild' Green
profile image
Mark Venturelli wrote:

"When I talk with other designers, most of them overlook or are not familiar at all with Guild Wars' amazing Skill/Class hybrid role system."

I find most people aren't familiar with Meridian 59's system, either, which I mentioned in the article. It uses a nice hybrid system that I think works well. (I maintain M59, but I didn't design the original system.)

Joshua McDonald wrote:

"I have to disagree on your proposed solution."

David Wipperfurth wrote:

"Still, how much better would this article have been if the result was a radical suggestion."

The problem is that radical suggestions have been made for years about online games, even back nearly two decades when I started reading on Usenet. People have been complaining about the "holy trinity" design for as long as that term has existed; I think it was coined because someone wanted to criticize the design.

The problem with proposing a radical design is that it is a lot more work. I also thought it was amusing that we could go to the progenitor of the current systems and look for ways to change the current design fashion. I wanted to show that even a mild change can create some interesting results. Detailing a more radical solution would also put this already long article to truly heroic lengths, and would limit the audience even further. My goal here is to spur discussion, like Bart Stewart's insightful comment above.

Finally, I am a professional game developer/designer. My radical ideas will probably go into a game I develop. :) Not much profit in spilling the beans here if I want to use my concepts in a future game. (Anyone willing to fund game development can contact me for details. ;)

Luis Guimarăes wrote:

"The trinity is there because that's what MMO players want."

Before EQ1 you could say, "Open PvP is there because that's what MMO players want."

Before WoW you could say, "Required grouping is there because that's what MMO players want."

Before EVE you could say, "Fantasy games are there because that's what MMO players want."

Not to say these are the very first games to break the paradigm, but just because something is fashionable doesn't mean it's the best solution and always will be. Yes, some players like a specific aspect, but that doesn't mean that alternatives shouldn't be discussed.

Thanks for the comments, all!

Derrick Whittet
profile image
It's not there because that's what players want. Players, in the most case, don't know what they want. It's just what they know, and in many cases they've never seen anything else.

I have to agree with Brian's proposal in this. At a fundamental level, I think the biggest problem these combat systems run in to is healing. Once you bring in healing (in whatever form) you immediately drop into a Holy Trinity setting. Even if there is no "tank" role; you still end up needing designated players to take damage so the healers know who to heal.

A game without (or with only minor) healing - focusing instead of positioning and avoidance - opens a great deal more tactical options in regards to protecting damaged party members, while allowing for a much better solo/nonstandard group play experience (all characters are equally capable of being self sufficient when necessary).

In traditional RPG's such as D&D, we accomplished this very well, with healing being extremely limited: Typically a very small number of casts of minor healing ability. You didn't race incoming damage vs. healing output, it was much more an emergency measure.

Finally, it's much more (I know, I know, these are bad words to use here) "realistic" or "believable". A starship, for example, could quite likely be repaired(healed); but it would have to withdraw from the thick of the fighting, likely with friendly help covering it's retreat; and the repairs capable within the timeframe of a single combat would be minor - more likely to be restoring lost functionality than restoring to perfect shape.

So, extended to a fantasy setting: Remove/lessen healing. Have damage reduce combat ability, but positioning be critical (collision detection etc). Focus on avoidance (be it through extreme mitigation via heavy armours or simply avoidance) while allowing players to actively rescue friends.

This doesn't just work well for PvE, it's a MUCH more involving model for PvP settings. The Holy Trinity fails badly there - and is a key example of why PvE games struggle with PvP.

Mike Engle
profile image
Aggro mechanics are the true reason for the trinity, as Bart points out.

Aggro is successful because it takes a simple question ("Who should mobs attack?") and turns it into a formalized game system. It's a game unto itself; a pattern to be identified and interacted with in interesting ways.

The removal of Aggro is possible, but it reintroduces the "Who should mobs attack?" question and removes a marginally fun minigame. Neither is crippling, but they certainly warrant solid planning of how your game will cope with these differences.

"Eliminating Specialized Roles" is a dangerous phrase, since specialized roles are a significant part of role-playing game popularity. It becomes a less dangerous phrase if we're not talking about the elimination of specialized roles, but instead the elimination of being locked into a SINGLE role.

Creating new roles (to replace removed ones) is a simple matter of creating player needs, then creating class(es) with tools to address those needs:

1. Mobs call for help, aggroing additional mobs in a very large radius (enough so that whenever it happens, you have to run or risk death.)

2. A Silencer class is added, who can spend half the fight keeping mobs suppressed of sound.

In the above scenario, every good group will need someone to fill the role of a silencer. The trick, of course, is to develop roles which are a little less arbitrary than this example (as silencers don't exist in real combat, but heavily armored knights *did* charge headlong into combat as shock troop "tanks".)

Oh and lastly: Puzzle Pirates. Ship combat. MMO. No trinity to be seen. Specialized roles. Yeah, it breaks the "standard combat" rule established on page 1, but trying to solve the trinity in the context of standard combat (ie without adding new mechanics like Mobs Shouting For Help) is going to meet with lackluster success anyway.

Joshua McDonald
profile image
Mike makes some good points. I'd like to comment on his first three paragraphs in particular.

It isn't necessarily the existence of aggro rules that create the limited Tank/healer/dps situation but the particular aggro rules and character abilities. A few modifiers to the aggro rules wouldn't entirely eliminate the trinity, but it would certainly make it less concrete.

Imagine if you simply changed two rules:

1. Make healing cause a lot more threat, but only on enemies attacking the person being healed.

2. Take away some of the tank's aggro controlling abilities.

Suddenly, there's a need for DPSers who can take a few hits. Maybe even DPSers with a taunt ability to pull attention off the tank temporarily so that he can be safely healed. A tank starts to need to consider his own damage dealing because he can't just use his built-in abilities to always control aggro. Likewise, a backup healer would be extremely valuable in aggro management, and tanks with healing capability (Paladins or druids in WoW) would find that extremely valuable, both for directly keeping party members alive and for managing aggro.

Essentially, you just made combat enormously deeper and required characters to do more than strictly hold to their role. This can be both a good thing or bad thing, depending on who you're targeting your game to. People who prefer simplicity and grinding won't like the system, while people who are looking for a difficult fight where they need to make intelligent decisions on the fly will love it.

As Mike pointed out, aggro rules will always be important, and further, players are likely to always manipulate them whether that was the design plan or not (even completely random aggro gives openings for manipulation). Better to plan for it from the ground up.

Mike Engle
profile image
While aggro rules will always be important to developers, importance to players is going to be relative to how much manipulation players have over them. Make it a minigame (WOW) and players will write Threat Guides. Make it uncontrolled and it might not enter their minds at all (like in the countless genres which lack aggro mechanics.)

No survivor player "tanks" in Left4Dead. Zombies just come at you. It's chaos. Uncontrolled. Your best bet is to send the guy with high health up front and hope for the best.

An MMORPG could be made with similar AI. In fact, Guild Wars is darn close.

Bart Stewart
profile image
There's another point worth considering, which is the notion of retaining the very distinct "trinity" roles but allowing players to switch back and forth between roles easily (when not in combat).

This approach retains the desirable simplicity of tank/nuke/heal roles without permanently locking the player's character into a single one of the roles. I'm very curious to see whether Sony's "The Agency" -- which will allow players to change roles by simply putting on a different set of clothing -- winds up being popular because of this combination of structure and flexibility.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

Sean Parton
profile image
Fantastic article. Irrespective if you disagree with the sentiments of moving away from the Trinity, it's a great dissection of the beast.

That said, I'm not certain how easy it will be to pull away from it. When push comes to shove, even most suggestions here are basically calling a rose by another name.

On Healing: I believe that this is the biggest crux of the Trinity. When you have a dedicated healer, everything else just falls into place (which Derrick Whittet pointed out fairly well, so I'll not to repeat too much). I really like what 4th edition D&D did, though, which while you still have some classes that can be pure healers, each class has it's own minor ability to heal itself and usually a few ways to defend itself as well. When combined with the focus on dealing with encounters that have multiple enemies (instead of always boss and minions or just solo's), it means that most everyone can stand alone, while not necessarily needing a healer.

Still, healers really help, and even 4th edition D&D still has roles. But partying with roles missing is possible; it just requires DM intervention and some changes to encounters to accommodate for that.

Enrique Dryere
profile image
League of Legends, DotA's spiritual successor, somewhat achieves this. There are still tanks and dps, but survivability comes in a variety of forms. I know it's not an MMO, but its combat is quite similar, showing that tank/healer/dps is not necessary for PVP. In fact, it's not necessary for many types of encounters.

It only becomes necessary when you want to create stand-and-raid encounters where someone sponges the damage, someone negates it, and someone deals it. As tired as this system is, an MMO without it may still feel lacking.

The solution I came up with when I first grappled with this subject a few years back was to eliminate the inevitability of damage. That is to say, in the current MMO model, the boss WILL deal damage to someone, which must be mitigated or healed. If the fights went more like those of an action game, in which the player is expected to avoid the enemy's attacks in one way or another, you remove the necessity for tanking, thereby removing the need for healers as well.

The big drawback here is that setting up unique encounters of this type would be a lot more work than the current standard. But perhaps this is what MMO designers should be focusing on instead of voice acting and other garnishes.

Sion Williams
profile image
Enrique, I like your suggestion but then your describing a run and gun approach. TI can think of one limiting factor here; the area of combat would have to be small to avoid players tactically kyting the boss around. High dps classes like mages have already been seen to do this in WoW with a further 2+ members in the group it would be completely plausible that a boss never gets a hit on someone.

Eric Mangrum
profile image
Great article that i think hits on the head what is the problems of the MMO genre right now. The problem is almost every game that has come out that I've played that even slightly shifts away from this model seems to die a quick painful death. I can think of two examples, first Conan, which was a miserable commercial failure but mostly do to game play issues and lack of content. It had the "trinity" in place but game play was so diverse that it was almost a non-issue. It just lost a lot of subscribers because there just wasn't much more to do, it was boring. Second is Vanguard. I really enjoy this game still, but it had huge technical issues that it still hasn't really over come. It is great because it offers the "trinity" archetypes but it also offers several hybrids. My two favorites being Blood Mage a healing/dps and Disciple a tanking/healer. A group of these with a Shaman, which can either be a tanking/healer or DPS/healer or DPS/tank makes for an awesome group. In fact that is the group setup me and some friends run with. It requires us to manage agro without taunts, watch each other because not one of use is a strong enough healer to spam heal ourselves in a real battle, and still we have to kill the bad guy.

The problem with both of those examples is they had huge technical issues at launch so they lost the WoW kids almost instantly and have simply been left4dead by the developing company. Nothing in the MMO world will ever compete with WoW so developers need to learn to cater to a smaller group and provide technical and polish support to make the games last a long time.

Jonas B
profile image
The tank-role has to go. Aggro-management as in 'AI-Monsters will automatically attack the tank' are stupid and remove the need for tactical movement and placement. And they don't work in PvP, so a player has to unlearn PvE strategies in order to be competent at PvP.

I agree with the comment above that Guild Wars has one of the best, most refined and most interesting class/skill systems.

While GW has its 'flavour of the week'-builds, this does a lot to keep PvP interesting. The FOTW can be countered with different anti-FOTW builds, which are not as good in general but very good against FOTW. As soon as too many people copy the FOTW, everyone else runs the anti-builds, and FOTW is a losing strategy.

Also note that GW does just fine without any artifical aggro-management mechanics (except for blocking/tactical use of terrain).

Rod Keller
profile image
Great article, all the comments are fantastic and offer a lot of different insight.

A common idea it seems from most people is to remove a piece of the trinity from the game as an option. Healing for instance, can either be a widespread application for each character or removed almost entirely, especially as a class option and made to be a out-of-combat option. Either are quite valid. You could make armor and health relative among all classes so no one tanks better than another. You COULD go so far as to balance each class so that none deals more damage than another. Each of these is a way to break the trinity.

I played Asheron's Call for a year and everyone had a healing ability. It had a chance to fail depending on what type of healing kit you used (ones that healed more had a bigger risk) but everyone could do it. You also had life magic which pretty much had the same effect, only, it took mana and never failed. A healer was nice to have, but surely wasn't necessary. The game itself had great design principles but also a lot of balance issues. Had they been smoothed out, I would consider it a solution to your quandary.

Removing healing from combat and having each player use their pool of health diversely to deal with incoming damage is another interesting design principle. Something D&D puts into effect to some degree.

I agree with the above posters that the trinity is meant mostly to fight 1 of the biggest monster that can be farmed for XP or another resource. Given a situation like Left 4 Dead, when the hoard comes barreling down on you, a completely different group dynamic may need to be established. AOE damage and effects are more important and a single tank becomes ineffective.

A bold idea may be to make combat a non-focused objective; instead part of what needs to be done. What if you need to get and hold capture points on a map while attacked by large groups of creatures. Sure, the trinity may be the most effective at holding a point, but are they efficient in fast travel or stealthy captures? Maybe the fact that they have no option of splitting up because of how weak they are alone will cause a problem when needing to do multiple objectives in quick fashion?

What Joshua McDonald said was interesting about the proposed solution being like Diablo II. However, I don't think that's a bad thing. Diablo II might seem like a poor team game, but that's because most of us play on players 8 and solo the game. If it had difficulty that actually required team play, or team play elevated what and where the players could do and go to, it'd be a lot more fun and you could play your own way no matter what.

Honestly I think there is a lot of outs for this simplistic design, if designers so choose to do so.

Nick Schubach
profile image
Why does everyone downplay the importance of crowd control? A good crowd control class (aka: EQ Enchanter) is neither a tank, nor a DPS, nor a healer. It's a support class that makes all their lives easier. It can't do any one of their jobs, but it can help them do all their jobs. It's the ultimate support class that falls into none of the three main classes you mention.

Holding an enemy is not the same as taking damage (aka: tanking) for the team.

Nick Schubach
profile image
Actually the more I think about it, a healer only helps a tank do it's job, so technically, there's only two jobs in an MMO. Taking damage and dealing damage. Everything else is a support role.

Mihai Cosma
profile image
So overall, bring back Diablo 2 in MMO form.

Also, another thing sparked to mind when you mentioned the removal of health bars. Having to make a certain chain of events, which can be set via skills or attacks, to determine wether the monster is killed or not. Imagine this like getting a certain percentage of succeses in Guitar Hero, so at over 30% basic monsters die, and bosses need 90% precision in what skills or attacks affect them, and the lenght of the chain. Imagine this less of a chain and more of a quota. How many different effects and attacks can you attack a monster with. Put this to be randomly generated each time, and you have a system where any class can and will be useful.

Derrick Whittet
profile image
In regards to aggro management as it relates to the Holy Trinity:

Who Should The Monster Attack is a good question. The game mechanics should indeed address this. However, I strongly believe as we move forward - and technology improves - this design needs to move from the highly contrived minigame that exists now to a more believable system. The mob would most likely initially target a nearby player - different mobs may have different initial target preferences, but that has little impact. Once the battle is well underway however things should be determined more logically.

Lets assume, then, that collision detection is in place. Friendly Fire can happen; or at the very least you can't shoot through your friends. There is no, or minimal healing: It's not a DPS vs. Healing race. Facing matters - as there is no healing(or little), avoidance or total mitigation(blocking, dodging, etc) is critical, and turning your back on your opponent severely penalizes said avoidance/mitigation.

Aggro is surprisingly easy to calculate. The mob needs to calculate the damage he's taking from various sources (who he hates more); and compare that to how much damage he will take turning his back on players he's currently in combat with (when avoidance and mitigation drop enormously, incoming damage skyrockets) to chase after another target.

There are two key points to note with this: Firstly, it translates perfectly to PvP. No more "Two Games In One". Second, it allows players to fulfill roles they are accustomed to without being forced into a specific class chock full of silly arbitrary threat abilities. It allows you to ease players into a more involved gameplay system, with them none the wiser that the underlying game is completely different - until they have time to really get into it.

But, isn't this then just "calling a rose by another name"? No. Because any player can be the "tank"; and it can change mid-combat when necessary. Other characters may or may not be as good at it; but because the role relies on actual player skill and action rather than arbitrary abilities, it's much more natural feeling. It makes PvE feel much more like PvP (and thus, vice versa). It leads very well into group vs group PvE encounters too (don't you feel that the group vs. single mob encounter design gets old?). The roles still exist to some extent, but in a natural and immersive way.

The aggro minigame is not a bad concept, but the current implementation is just too arbitrary. Moving in this direction guides players into a more in depth combat but easily grasped combat system without shaking up their world too much in one go, moving away from the tired and worn out Trinity.

Owain abArawn
profile image
"holy trinity only applies to pve."

Werner hits the nail on the head! Nothing anyone else has said better illustrates the artificiality and superficiality of MMOs in general, AI in particular, caused, as I have expounded upon elsewhere, by the lack of realistic collision detection, which it teh root of all evil in MMOs.

If you develop a game where players have significantly different tactics when it comes to PvP and PvE (assuming you even permit PvP), that should be your biggest red flag that your game is going to have balance problems.

The lack of realistic collision detection results in artificial AI behavior, requiring 'agro' mechanisms to compensate. If you introduce effective collision detection, then even if the mob (or the player) really wants to get to the squishy archer (considering a non fantasy setting - no casters), he first has to get past the shield warrior. If there is only one warrior, that might not be that hard, but if there are more than one forming a shield wall (phalanx tactics!), that may be very hard or perhaps impossible, thus rewarding strong player interaction. Once you break through the shield wall, or flank the enemy formation and attack the lightly armored archers or the undefended backs of the warriors, the defenses should realistically collapse.

In a non-fantasy setting, you don't need healers, at least not during combat. The idea of the slowly decreasing health bar where your combat effectiveness remains constant while your 'health' slowly erodes is highly unrealistic anyway. In real life, most blows are parried, evaded, or blocked, but once you get even a single serious hit, your combat effectiveness drops precipitously. Instead of heath slowly eroding, it is stamina you have to conserve. Fight long enough, and you no longer have the strength to continue. If you survive the battle, your wound can be bound back in camp, where you can rest and recover your strength to prepare you for battle the next day.

So here is an example illustrating how the Trinity can be discarded, and yet you can still have compelling game play that should apply equally well to both PvP and PvE. All it takes is a little imagination and a willingness to innovate.

Mark Harris
profile image
FF XIV is supposedly using a system similar to "The Agency" that Bart mentioned, whereby you can change roles on a single character at almost any time (I assume outside of combat) by changing equipment (in this case weapons). The different "professions" level up individually depending on how much xp you gain while in that role. I'm also curious to see how these systems are designed and whether or not they catch on.

Also, a few of you have mentioned direction and facing and shield walls and whatnot. I've never played Age of Conan but I remember reading that the combat system did take some of that into account. Was there not some directional combat or combat options to direct your attacks away from the "shielded" part of your opponent? Can someone who has experience talk about that system and how it succeeded/failed?

Owain abArawn
profile image
I played Age of Conan briefly, and the melee combat system used was ok for PvE, but it was too slow to be effective in PvP, in my opinion.

It worked by showing a curly bracket over your opponants in the direction thier defenses were oriented. If they defended equally in all directions, They'd have a curly bracket left, right, and top center, so they would get some defensive mitigation no matter what direction they were struck from. In PvE, the mob would sometimes double his defense from straight on, and would display two or perhaps even three curly brackets top center, and would leave one or both sides undefended. As an attacker, it would be to your advantage then to do a direct damage attack against the undefended side, or a combo attack that ends with an attack to that side, resulting in increased damage to the mob.

If I remember correctly, in PvP, you didn't see the brackets on another person, so you usually just mixed up your attacks rather than trying to second guess things. The pace of fights in PvP was usually too rapid to make switching your own defenses practical, so most people ususally chose for a balanced defence rather than using shield hopping, but if you noticed that a person ALWAYS used an overhead stroke on attack, i suppose you could use that to your advantage, but I never saw it in PvP myself.

So the approach was interesting and different for melee PvE, but irrelevant for PvP. It had no effect against casters for either PvP or PvE. Unfortunately, AoC had too many other bugs. One dungeon would frequently cause my system to crash and reboot when I went into it. That and other bugs pretty much killed AoC for me.

Porter Woodward
profile image
Something that really stuck out in the comparison to old-style D&D (or AD&D as is the case) - is that those games often involved something more than combat as the only activity. When you boil the entire MMO experience down to _just_ combat, whether it's PvP or PvE - the structure is pretty limited.

Mention of the Thief or Rogue style class as a DPS class in old-style D&D really misses the point. They could also disarm traps, open locks, pick pockets, etc. The ability to disarm traps is a huge damage mitigation capability. When a character's role outside of combat is considered - there can be a lot of utility and a breadth of play options. When combat is the end-all, be-all - the dimension that matters most about a character becomes what they can do in combat.

Of course, actually designing a game that requires strategy more complex than "rock, scissors, paper" (rsp) is much more difficult. And while the 3 roles: tank, healer, dps (hate those names - talk about ruining the suspension of disbelief) end up being somewhat more complex than rsp - the single dimensionality of combat without other factors really mattering limits options.

Most MMO games don't seem to focus much on mobility based skills. Whether this means being able to bypass barriers by flying over, teleporting, etc, or just actually making maneuver a key combat capability - not many seem to do it well. Position (high ground) also seems under-utilized. Skills which allow combat avoidance (disguise, invisibility, or group invisibility) to bypass grunts also seem less available.

And, well actual "roleplay" options in most MMOs are extremely limited. You can't really negotiate with NPCs to achieve victory - victory is achieved through combat and defeat of the enemy almost exclusively. Including story elements and puzzles is extremely difficult in MMOs these days. The social nature of the games, combined with networking technology really undermines the value of adding such elements - as it's likely anyone playing can read all about the solution to some puzzle or the story's big reveal somewhere on the web.

lee loo
profile image
tank, healer, dps is actually not the original holy trinity, in eqlive it was warrior, cleric, enchanter, after eq developers started scaling back crowd control more and more because adds became a joke and most people had difficulty playing good cc which left us with todays trinity.

Thomas Langston
profile image
Collision detection is not required for shield wall style tactics. Snares and push effects, especially when combined and when used as very short range auras or stacking effects, can easily create the same effect of "You shall not pass!"

Tom Hardy
profile image
An important note for this discussion is that EQ actually had the elusive 4th pillar of groups.

The Enchanter and the Bard served completely different roles in the groups outside of the conventional trinity schema: crowd control. This particular gameplay was a favorite of many. Unfortunately that particular group component has been trimmed down or removed from most games since then.

Like the presence of a tank or a healer, designing the game to assume the existence of a crowd control class in every group led to difficulties in balancing the game. Content design overall is more complex with the addition of this archetype.

One game in particular embraced the archetype: City of Heroes with their Controller archetypal class.

When I look at a game like WoW today, I still see a watered down EQ, in that encounter designs are firmly centered on the trinity. Granted many classes in WoW have some specific and/or singular crowd control abilities, but not to the extent of an enchanter in EQ. A typical 10 man raid group in WoW possesses less overall crowd control facility than a single enchanter from EQ.

EQ Deserves a kudo for including that aspect and I do wish that more designers had been up to the challenge of continuing the idea in later games.

Zebulon Rogers
profile image
A lot of good ideas here. As Philip Daay said,

"But, who says that MMO content must focus on a group killing one powerful boss or elite?

Imagine if we inverse the archetype. A Fantasy MMO is published where one player or small groups of players must now overcome huge hordes of less powerful monsters. This assumption will force design changes away from the three roles. Namely 1) Tanks become irrelevant since the whole party can be attacked. 2) It moves the game design axis away from Tanking/Healing to the DPS classes, placing a premium on Area-On-Effect. 3) Game designers will feel pressure to give all classes mobility, survivability and/or heal. "

AoE such as crowd control would do nicely here. A scenario I imagine would be a group of three or more characters surrounded by advancing troops. The caster's job is to slow down these troops, making their advance a manageable problem, whereas DPS whittles them down in their impaired state before the effect ends or more troops push in to assault. Tanks could still play a role here if the DPS were ranged. Tanks can act as wall-like barrier against the enemy, and slowing them down via aggro would allow ranged attackers to pick them off from behind said barrier.

Really, I think the players should be given an MMO where they can put together different groups and experiment to see what combos work best for themselves. This way it can also help, in some small way, eliminate the pesky "useless" class. Let's say we have a rogue in a group against the aforementioned hordes, and the group has less than a minute to prepare. The rogue could sneak about and establish a perimeter using slow down traps such as ice or encumber. They could continue doing this throughout the battle as well as doubling as DPS throughout the battle with backstab fighting techniques.

The imagination is always the limit (well, and budget), and there are innumerable scads of different angles you could take to make it fun. There is room for both tradition and progress in this style of gameplay.

Jules Mazarin
profile image
I found Nick Schubach's comment about the Enchanter role in Everquest interesting. I played a Dark Elf Enchanter as my main for about 5 or 6 years on the original EQ, and saw it go from a superbly challenging and appreciated role to crack dispenser (selling the KEI intellect buff by the bank) during that time. As Nick said, the EQ enchanter didn't fit any of the "trinity" roles, but enchanters had powerful abilities that enhanced any group's chances of survival and success. The enchanter was Crowd Control personified. During a period when "adds" often meant that the group was going to wipe, a good chanter could save the day. Using mesmerization ("mezz"), he could keep mobs quietly lined up, waiting their turn to be killed by the group's damage dealers. Instant or very fast casting stuns could mitigate damage, or give the chanter time to land the mezz. The ability to charm mobs meant that the chanter could (temporarily) recruit an extra fighter for the group. Though charming was not nearly as valuable as the mezz, it was a heck of a lot of fun. And what other class could turn themselves into a piece of furniture to avoid aggro?

I'm thinking that the solution might lie less in moving away from MMORPG classes as such, as maybe in better and more classes, employed in more imaginative ways. The trouble with this proposal is that there were reasons why the enchanter class got shafted in EQ ("nerfed" is not even close to the word for it), and those reasons

continue to operate today, constricting MMORPG design to accommodate just three roles.

I think that the first and foremost reason is that players do not like the game to be "too hard". Chanters were hard to play well, and hard to play with. The group had to understand how the chanter worked, and apply appropriate tactics. If everyone in the group didn't understand that waking a peacefully sleeping mob before it was done killing the first one was a bad idea, then the chanter's work went for nought.

Another reason (as Brian points out) is that too many gamers believe in the proposition that "that which can't be counted doesn't count". You can write scripts and complicated formulas that analyze DPS; you can't count damage that is NOT done because a chanter mezzed three mobs during a fight. I HATE this focus on numbers; to me, it's an absolutely fun-killing thing. It makes play into work.

A third reason is that too many people enjoy direct violence. To them, whacking, slashing, crushing, or frying mobs is what's fun. They want to see Big Numbers in the damage stats when their character does its thing. I like violence as much as the next player, but I like it to be applied with some finesse (at least some of the time).

Though I don't know what the designers were thinking, I believe that they were responding to what they perceived the players wanted: put the major focus on DPS, and give most classes a reasonably good tanking capability. That meant group encounters were not threatened by adds, because these could be off-tanked by the pally, the rogue, or (even) the bard. So what were chanters good for? They could slow--but not as well as the Shaman. They could do fifth-rate damage, and apply buffs that could be bought by group members before going to fight.

If the EQ developers were correct in their perceptions of what the players "really" wanted, then I suppose there isn't much hope that there's going ever be another game that I enjoy nearly as well as the first couple of years of EQ. The game I'm playing now, Aion, has some pretty nice features--but the variety of class roles seems very limited to me. In fact, let's see, you've got healers, tanks, and...uh...DPS.

Chris Berkholtz
profile image
I think Dota/heroes of newearth is a fantastic way to limit the dependence on the trinity. If it could technically be adapted to an mmo setting, which I believe is within reason, would be a significant change. The games have the classically 3 primary roles however they are no more important then traditional secondary roles. You don't need a tank, you have ai allies, you can kill your opponents before the ai dies or heal the ai. you don't need a healer, its reasonable for a fighter to step out of the fight to heal themselves. Automating the tank in this way destroys the trinity. tanks and healers are still there but a healer doesn't need to be primary because the ai tanks are more or less disposable. In dota you have rogues, healers, summoners, mages, various cc, tanks and mixes of all of them. They are all effective in there own way, and combinations of characters are more or less powerful but always effective. I suppose an alternative for a similar system would repeated and rapid revivals, though not technically challenging it would be difficult to design.

So what would it look like? perhaps you had wagons of troops being transfered to the front lines, the player hopes in the wagon and gets taken out. The battle might take place in a perpetual castle assault, or hordes of demons spilling from a cave. The players assist in fighting the opposing forces which are ai and players or simply a more powerful ai. A player gains experience is shared. In order to stop players from simply standing around and watching, you should have fairly low damage coming from the npcs, however not so low as to disable a healer. Another way to enforce participation would be a less predictable aggro system. Perhaps the ai allies would be more likely to come to your aid if you are healing them, or dealing allot of dmg. A system like this would also allow for some amazing world pvp with tangible benefits.

On the down side I could see this doubling server and bandwidth requirements.

Sean Parton
profile image
Porter Woodward said "Mention of the Thief or Rogue style class as a DPS class in old-style D&D really misses the point. They could also disarm traps, open locks, pick pockets, etc. The ability to disarm traps is a huge damage mitigation capability. When a character's role outside of combat is considered - there can be a lot of utility and a breadth of play options. When combat is the end-all, be-all - the dimension that matters most about a character becomes what they can do in combat."

Well, my response is a bit of a sidetrack from the Holy Trinity discussion, but...

The main reason why things like disarming traps, open locks, etc is downplayed in a lot of games today is that it's not fun, either for the rogue, for the party the rogue is with, or both. It's basically jumping through hoops for a benefit (or mitigating a hazard). For example, in Dragon's Age, you'd be a fool to not have a rogue in the party, due to the profuse amount of locked chests, and the fair bit of traps that can occasionally be found. But you know what? Having to switch to a different character to open a chest, press a button on a trap, or poke an NPC (with a 10 second cooldown!) to get an item isn't fun. It's a chore. This becomes even more of an issue in a group situation where you have to wait on one guy (and even worse, hope he's not retarded about it and knows what to do).

Back on topic though, giving a class a utility that isn't used in combat (or at least most combats that are expected in the game) isn't addressing the issue of designing an alternative to the trinity. And as many people here have pointed out, irrespective of what design may intend, people will probably still try to make parties based on the trinity anyways, even if it's not actually the most optimal method. It reminds me of what I've heard about Red Faction: Guerrilla; many people complained that it was too hard, but it was because they were thinking as you would in many games in that genre, not taking into the account the feature of being able to destroy through walls/etc to make your own path.

With that in mind, perhaps the best way to go about mitigating this is to find a strong group of MMO gamers who want something new, and working with them to ensure there is an initial base who wants to go in that direction?..

profile image
I would have liked to see a better proposal to conclude this piece. The "use of tactical options in combat" is a pretty obvious choice at adding enough variables to the mix to overtake the specialized class system. Nearly every game that isn't an MMO already does this. So the solution to fixing the MMO is to make it into another type of game? Obviously. I would have liked to see a condition in the piece about keeping the statistical and MMO/RPG game mechanics intact while modifying the class/skill system away from the trinity system.

A few suggestions:

Secondary race/tendency type stat that is basically a second class attached to your main class. The second class won't gains skills, but will have improved tendencies toward different types of enemies. For instance, "skilled at tearing apart soft skins"(+dps to soft skins) while their main class is tank. This would allow them to fill tank primarily, but when their party is fighting soft skins they'll see a dramatic increase in their DPS effectiveness. This would allow them to advertise "LFG tank with tearing tendency". Raising their chances of finding a group as either a tank or a DPS or they could even be both in the same group depending on where the group was fighting. This is just one example, but with a well designed tendency system you could see some very different yet effective customization options.

This switches the design pressure from 100% on the class system itself to 50/50 class and environment. The world design can even get creative by mixing and matching different types of enemies in various spots/dungeons to ensure a diverse group could succeed.

Chan Chun Phang
profile image
Just some random suggestions which would also break the trinity:

1) Use a scissors/paper/stone format. This would optimally work in group-based gameplay. Essentially this would force players in a group to swap roles depending on enemy composition, where some elements are better at defending against another element, and another element better at attacking against that same element.

2) Push the damage and/or strategic focus to the fore. For instance, implement a system where the objective of defeating the enemy is a lot higher than keeping alive. The most obvious method is to give everyone regenerating shields/hp.

@Chris Berkholtz

Actually, GuildWars already does that. You can have accompanying AI heroes and/or hirelings (which are basically preconfigured AI characters) to handle the tanking, DPS, or healing aspects. In a sense it boils down more to strategy and teamwork

Sergiy [Svargas] Yevtushenko
profile image
Nice article. Want to share my thoughts (partially inspired by AD&D4E):

Basic ideas:

1) to have "aspects" that can be "borrowed" to teammate if he is in trouble

2) to make "tank" & "healer" role as part of "aspect"

3) to make all characters damage dealers with 1 aspect

Sample implementation:

1) Every character have "life steal" property (can be found in Diablo 2) - ability to get % of damage done to hit points

2) Two "aspects" present - "healer" & "absorber"(tank)

3) Healer can transfer the health he steal to selected teammate

4) Tank can increase damage absorption of selected teammate, taking absorbed damage in his hitpoints

So team play in this system is more about "react on situation" - there is no dedicated tank character able to survive under boss's fire, but there is teammates who help to survive the person taking beating.

This system solve lots of problems present in "holy trinity" design:

1) No need to hurt solo performance - characters basically have same DPS & similar survivability

2) As result of 1, one on one PVP balance is also much easier to achieve

3) Healing & tanking is not boring duty anymore; character does utility job as part of DPSing

Mark Harris
profile image
Thanks for the info, Owain. I had wondered how the AoC combat system worked in practice.

@ Sean

"With that in mind, perhaps the best way to go about mitigating this is to find a strong group of MMO gamers who want something new, and working with them to ensure there is an initial base who wants to go in that direction?.."

Sounds similar to how EVE started, and CCP has built a very successful business around that dedicated community.

Kumar Daryanani
profile image
The holy trinity exists because of AI behaviours. Most if not all Diku-mud offsprings rely on 'aggro' (EQ) or 'Threat' (WoW), where AI enemies attack whoever is at the top of their aggro list. Thus, the tanking role is about being at the top of that list as well as being able to withstand the ensuing punishment. All other classes must then perform their role without getting above the tank on the aggro table.

Rather than changing the classes themselves, then, it is the behaviour of the mobs that needs to change to move away from the holy trinity. The challenge is to create less-dumb mobs that are still predictable, and then giving classes the tools to deal with those behaviours.

A good example of this: in WoW there are is at least 1 instance of a boss that doesn't have an threat list, it simply attacks players randomly, but it's damage output is low enough that healers who are on the ball can, with quick reflexes or teamwork, keep all raid members healed while they whittle down the boss's health. The reason this boss works is because it isn't the norm, so this erratic behaviour is part of the encounter.

Replacing the threat table is huge challenge, because it works so well at creating the tension of combat while establishing a predictable path to success: control the mob's damage output and win.

As Brian mentioned, the Holy Trinity is an equation: if boss damage output - party damage mitigation < party damage output - boss health, the party wins. The tank and healer form the first part of the party's equation, while the DPS encompass the second part - although WoW has done a pretty good job of spreading some of the damage mitigation aspects to the DPSers as well, in the form of AoE damage, damage zones, and so forth.

If you want to eliminate the tank role, boss damage needs to be spread out in such a way that it still creates a challenge, but can be overcome. So instead of having one character with huge amounts of health, boss damage should be manageable by anyone, under the right circumstances. You can give characters damage mitigation on different scales (essentially the difference between Heavy, medium, light, and no armour), damage avoidance (dodge, parry, block, stuns, etc), and so forth. Another idea would be expanding defensive abilities, and having all classes have some amount of them, to be able to 'pop a cooldown' to survive if necessary. Abilities that cause monsters to attack specific characters (taunts and misdirects) would also have a place.

I'm rambling, so I'll let this cool off a little and see what other comments crop up.

Todd Edwards
profile image
There are some great ideas here, if they were all put into one game I think it could set the new bar for MMOs. I think some of the ideas would work better with traditional fantasy games compared to other genres though. Most gamers seem tired of fantasy for now for their mmo's so it is really important to start looking to radical changes in how we design the next gen of mmo's.

Getting rid of health is a good step, just using energy but make it something that slowly regenerates on its own so you still need good dps to beat the mob. If someone is close to losing all their energy then they would need to step out and wait or use an ability that makes their energy recharge faster but they can't be attacking for so many seconds to use it. Let everyone have a blocking ability so if the mob focuses on them and they are low on health they can become their own tank without having to rely on someone else to notice them. Also if the other party members aren't doing enough damage to take the focus off that player then their death is because of the party members and not because the boss hit too hard, which keeps a need for players to play responsibly.

You could have a pet with abilities that do a certain function like give a debuff or buff so that you could still focus on doing dps but also be a support role at the same time. Instead of having tank abilities like taunt you could have more of a focus on controlling the movement of the mobs so they can't get to the other players, it is more interesting to clip a mob or use an ability to knock it away so it can't get to the other party members then just spamming taunt every time it's cool down is up. To avoid having a mob that never hits anyone the mob could have a damage aura so the tank would eventually have to fall back and get their energy back so they don't die.

This would leave the dps vulnerable so not to kite the boss all day. On top of this you could have different abilities for the boss such as a pull move where the player who is farthest away would get yanked to the mob's feet if they were actively attacking the boss. Or if the boss did not attack someone for 5 seconds it would stop chasing and put up some kind of shield that could only be broken by some kind of melee attack.

I think that if someone took the skills from weapons from Phantasy Star Online, collision detection from Warhammer online, and npc helpers to let you be more soloable from Guild Wars and using spells from multiple classes to do more damage from Dragon Age Origins ( casting oil on an area then lighting it on fire) you would have a successful mmo that players would play longer before getting bored and leaving compared to some of the current mmo games out there. I know there are other games that do these things as well but these are just the ones that came to mind first. Great ideas from everybody.

Dan Reed
profile image
I believe the holy trinity exists for one reason - it works. I am all for innovation, but there is a reason why WoW is so successful. They took this holy trinity and developed it for the masses while at the same time making it possible for each role to successfully solo to the top level. Now there is an added ability to switch specs on the fly to redefine their role before entering combat in group content. This makes it easier than ever before for the player to experience that content the way that they want.

Wow has also introduced other situational tossups to the traditional plate tank - take for instance the mage tanking in the High King Maulgar fight. Between the mages own damage shields, and spellstealing the shield from the monster they tank - these most squishy of dps are able to mitigate most if not all of the damage. I believe it would be nice to see this kind of dynamic included more often in raid encounters.

Anytime you have player vs monster encounters - it's going to be kill or be killed, win or lose. Speaking purely about combat encounters, there are only so many ways this conflict can be resolved. You can either kill the monster fast or you can kill it slow. When it comes to dealing with the monsters damage, you can either avoid its attacks altogether, mitigate it's attacks while fighting it, heal attack damage while fighting, or a combination of all 3. This gives 6 combinations of dealing with the monsters damage output, and 2 methods of killing the monster(fast or slow). I believe that there are only 12 possible combinations to kill it - before it kills you.

I believe that Innovation will come from non-combat areas more so than combat when it comes to group dynamics in the MMO's of the future.

Andrew powers
profile image
If this was just a matter of allowing classes to fill different roles, then WoW has already done it, and the problem is solved.

There are different forms of tanking which exist in WoW, but are not fully utilized. Drain Tanking, for example. Affliction Warlocks have a few spells that damage the enemy and simultaneously heal the Warlock. There is a balance between how fast it damages, and how much it heals while the enemy also damages. I think the Sith Inquisitor will employ some type of healing like that in Star Wars: The Old Republic. It allows for a player to do two roles at once.

As Dan mentioned above, there are other types of tanking, like the mage taking spell damage. Mages have an interesting tool, a shield that allows them to take damage to their mana in place of health. They are not wasting stats on a lot of stam, but can still take a hit. Situational tanks, some for spells, some for melee, creates some variety. I thought DKs were going to fill the role of spell tanks, but it turns out they take on melee damage as well as a warrior and don't bring anything new.

As stated previously by several posters, crowd control is really missed. Even in TBC, WoW had use for CC. Now, the bosses hit so hard, the tanks have so much stam and avoidance that "trash mobs" are just useless filler. It used to be before a fight the a mage sheeped one mob, a rogue sapped another, and a hunter was tasked to chain trap another. The tank would then have to choose which was the next target, and the rest of the group had to know who to fight and who to avoid. When someone messed up comedy and angered ensued. Now it is just the tank rushing in and everyone else AoEing the group down. The tank/spank/heal dynamic isn't so bad where there are still other activities that need to be done. When it is just boiled down to these bare bones, as WotLK is, then it becomes monotonous. Especially when other games follow suit.

Roles can't be taken away, or it all becomes a bunch of people fighting side by side, not fighting together, like in Champions Online. I would like to see a game experiment with this, put out four or five classes, give every class a different style of healing (fast but weak, strong but slow, group overhealing, single target healing...) and a few forms of crowd control, that vary greatly by class. Put in situations in the dungeon that could be controlled, going one way sets off alarms and more mobs come, killing a particular boss makes players hated/friendly to others in the dungeon. This could make for creative group play, and group cohesion, or total chaos no one wants to be a part of. It's a risk, but it could pay off. Even if beta just gave players a peek at leveled out characters in these experimental dungeons, it would be enough to see if this is what people want, and proceed from there.

Mihai Cosma
profile image
Everyone is still thinking in terms of the system. The system works that's why it's used, and in some games it's been perfected. I don't think it needs more tweaking to make it less then what it is, i think we just need new systems, like what i've thrown in my previous post.

Nollind Whachell
profile image
To me the holy trinity works because it's about diversity which is extremely important within systems and communities, even more so because it promotes interaction. No one unit or class can work on it's own. It's what makes RTS games so enjoyable, complex, and deep. You need a group of different units, all working effectively as a collective whole, to overcome your opponent.

So no I don't see a problem with the holy trinity concept. What I do see is a problem with classes being pidgeon holed into a particular trinity category. For example, people immediately think a warrior should be a tank because of their ability to absorb massive amounts of damage. Yet really all a tank is about is being able to hold the aggro of creature and avoid getting killed. Why couldn't a rogue do this just as well with a high degree of avoidance (i.e. ninja-like)? I noticed the same thing with my Healer in Allods Online when he switched from Endurance gear to Agility gear. He went from being able to "absorb" massive amounts of dmg to being able to "avoid" massive amounts of dmg.

As for healing though, particularly during a fight, ya that's a tough one. It's pretty much the domain of healer classes to do that. After a fight though, ya any class should be able to heal themselves with a variety of items (i.e. consumables, bandages, etc).

Michael_Gama Sutra_Gersten
profile image
So here's a question: Why have Agro at all? What does it really mean?

If a "tank" is someone who manipulates agro so that the mobs all gang up on them, then what it really means is that they are manipulating the NPC AI.


Does this then restrict other players in PvP? Are they now restricted, somehow, to only attacking one person?

Does that make sense?

Let me toss out a different concept: someone with a zone-of-control. You can't go past them without either going around their ZOC, overrunning them (massive numbers surviving attacks of opportunity), or taking them down.

What happens if you get a row of 5 of these guys? Why, you've got a real front line -- behind them, archers and mages are casting _distance damage_ (different from DPS).

Suddenly, the "raid" (large group) game -- distance damage with well defined front lines (WW1 games, anyone?) -- is different from small group or solo play. PvE and PvP look the same -- you can't ignore the front line, you have to deal with it. The concept of the "break" piece/move from Rock/paper/scissors comes into play (ala catapults, reavers, flying longrange-ground-attack zerg creatures, etc), as does the idea of "My archers are out of range of your archers, but I can still hit your front line".

You still have people filling the role of agro management and damage taking, but no longer in a "special cased for the NPC's" manner -- now it affects PvE and PvP the same.

It gives an immediate difference to small v. big battles.

Does it allow someone to solo? Sure, if your ZoC character does enough close damage to fight. Then you could have solo close combat progression, or teamed close + range; your range character can solo by doing "Move-shoot-scoot" tactics, etc.

[quote]1. Make healing cause a lot more threat, but only on enemies attacking the person being healed[/quote]


[b]Why should healing cause ANY threat/agro at all?[/b]

If you are fighting an intelligent creature that knows what a healing spell is, then yes.

But if they don't know what a healing spell is? If they are "animal intelligence"? Then no agro at all. They aren't doing anything that an ordinary animal would recognize as a threat.

Luis Guimaraes
profile image
I'm surprised with the ammount of comments on this subject.

When I was beggining to game design aspiration in my teen days, I used to wonder about what MMOs could be... nowadays I'm just skeptic if any MMO will get my attention for a long, long time.

Read @Porter Woodward@ and @Jules Mazarin@ posts for further explanation.

I keep my opinion that MMO player will simply reject any change or improvement in their simplistic games. Making any kind of system better than "trinity" is easy, making the players like it is the hard part, they're just move to WoW if you try it, and WoW gets enough money to try anything else.

John Petersen
profile image
I dunno, maybe leave the whole trinity thing alone and add more solo content instead. A soloer might be gimped in group PVP, but is ferocious in one on one.

I think MMO's as a whole need a make over.