The Trinity, AI, and EQ:Next – How I Lost Faith in the Trinity

xkerran-warrior-roars-in-ashfang-1375467886.jpg.pagespeed.ic.SBuC_6c8WK

It was August 2013 and like many other MMORPG players, I watched enraptured as Sony Online Entertainment’s David Georgeson presented us with our first glimpse at their upcoming title, Everquest: Next. I was surprised by their feature set, and resigned myself to a cautious optimism as I tried to temper my rising hype. As I’m sure most MMO veterans would agree, it’s best not to get too excited too early about these things. Then I heard it – EQ:Next will be ditching the trinity. I’ve always been a fan of the trinity, and my experience with Guild Wars 2‘s removal of it did nothing but solidify my belief that it was the cornerstone of good MMORPG combat; if anything could lower my interest in the game quickly, this was it. However, it was not long before my mind was changed.

After a bit of research, I stumbled across an old Gamasutra article from 2009 by Intrinsic Algorithm’s Dave Mark, one of the leading developers of Storybricks, the company whose AI has been licensed for the production of EQ:Next. Dave Mark and the Storybricks team are working closely with the developers at SOE, and for all intents and purposes, I believe they can safely be considered a part of the development team. My new position on the trinity has been heavily influenced by his article, and it’s definitely worth the read given the impact Dave Mark’s views and work will have on the final product of EQ:Next. I now believe the trinity to be an outdated mechanic where the question is no longer if it’s going to go, but only when. While I can’t be sure that EQ:Next will be the game to pull it off, I’d like to think that they have a good chance based on Dave Mark’s contributions to the team and the fact that the way they’ve described their combat appears to be more or less exactly what he prescribed in the article above.

How the trinity works

To properly explain my change of heart, we need to properly lay out what exactly the trinity is, and how it impacts the games that employ it. A lot of people mistake the trinity as being the only way you can have roles and tactics, a notion which stems from from a fundamental misunderstanding of exactly what the trinity does, and the mathematical role of specialized tanks and healers in a group setting. It’s worth noting that other roles such DPS and Control that are often present in the trinity, while important, are only peripheral in a trinity-based system and irrelevant to what makes the trinity what it is; despite the name, it’s really more of a duality.

Games that employ the trinity are essentially completely balanced around the effective health points (EHP) of the group as a whole. Those that have played ARPGs like Diablo are aware that EHP is calculated based on a variety of stats that can neatly be categorized as damage-reduction and health-regeneration. What the trinity does is use roles to have players specialize into the components of EHP so as to maximize the group’s EHP on the whole. The tank specialized in damage-reduction, the healer specializes in health-regeneration, and when done this way, the group can tackle challenges they never would have scratched the surface of before.

For a basic overview of this, imagine that a tank mitigates 40% more damage than the other players in the group. When the tank is getting hit, this effectively means that the tank has 40% more health, and that regenerative effects from the healer are 40% more effective, as the health restored will last 40% longer than they would on the other players in the group. The synergistic benefits of specialization increase as both the tank and the healer increase their proficiency at their respective roles; the more the tank increases his defenses, the more effective healing becomes, and through this, their combined efforts become much more powerful than the group could otherwise hope to achieve.

It’s important to point out that having tanks and healers as roles alone is not what makes the game trinity based; it is the mathematical relationship between those roles and the way encounters are balanced around that in not just a mathematical sense, but also a tactical one.

How the trinity is holding back encounter design

This is where the problem with the trinity arises. The non-tank members of the group cannot sustain the hits that the tank is taking, and if the damage amount of the hits was to be lowered so that they could, then the tanks and healers would be at risk of becoming superfluous or never challenged. This creates a situation where only one tactic is viable for every single fight. The tank has to be getting hit all the time. That’s not to say that challenging encounters can’t be designed with a trinity-based system; they can and they have been for years – it’s just that almost all of them have essentially been the same in that regard.

This limitation also creates an incredible barrier to the creation of truly dynamic encounters, as all damage dealt to the group has to be done so in a predictable way that is either manageable by the tank, or specifically scaled to the capabilities of non-tank players.

Brick WallPictured: the brick wall traditional MMO AI bangs its head against.

How Everquest: Next plans to tackle the problem

Some MMOs have attempted to address the problem by removing the trinity, but have done so without adding in a system to replace it. The general consensus has been that this has lead to games with zergy, uncoordinated encounters where the group roles that remain are watered down into near-insignificance; removing the trinity will not be an adequate fix – it has to be replaced. Dave Mark’s proposal and the one embraced by EQ:Next is that the solution lies in the type of AI employed by the game. I realize that an AI specialist proposing AI as the solution to a problem is hardly ground-breaking stuff, but he has a point.

With more intelligent AI, the basic and simplistic aggro system common in contemporary MMORPGs can be replaced, and it should be possible to design encounters that have different mechanics not relying on the EHP maximization of the trinity to function. There can still be tanks and healers, it’s only the mathematical relationship between the two and how they perform their role that needs to change.

Player roles in EQ:Next have been described as being MOBA-esque in design. In a MOBA, healers are really better described as support (note: I am not saying you will not be able to play a healer in EQ:Next), and tanks have a variety of ways to perform their role that are more interesting than just forcing everything to attack them. To blatantly completely fabricate a theory, I could see a situation where tanks perform their role through a variety of techniques, such as by physically blocking shots to intercept damage, by using guard mechanics to transfer a portion of allied damage to themselves, by preventing the enemies from reaching their target with the use of snares and knockbacks, by reducing the damage dealt to other players through the use of debuffs, or by using other methods we have yet to imagine. For actually implemented examples, look to games that have implemented PvP tanking using some of the mechanics I just listed. The tanks still serve to reduce the overall damage taken by the group, they just do it using a different mechanic than directing all of the damage to their face.

EQN Terran and Human Female in Cave

So what type of improvements is EQ:Next promising, exactly? The most relevant is that of utility-based AI’s effect on combat, a subject Fantomex has already covered on this very site. When engaged in combat, the AI will weigh the utility of all of the actions available to it based on the information available to it at the time including the actions of the players, the positions of the players, and conditions that are being applied to it. In a situation like this, the AI might initially determine that the mage in the back of the group is the highest priority target due to its raw damage. As the mob attempts to make its way to the mage, it is impeded at every step by the efforts of the tank, and, as it continues to reevaluate the utility of its actions, it may decide to attack the tank when it realizes that it will never make it to the mage with the tank on it. More complicated fights may see more complicated tactics – maybe the AI will decide to CC the tank after a few rounds of that and then successfully make it to the mage. The developers have stated that mobs will have varying levels of intelligence determining how clever they are in their use of tactics, so it’s possible that the difficulty of tanking and the ways players accomplish it may vary wildly from encounter to encounter.

Without getting our hands on the game, it’s impossible to know at this point whether or not they’ll be able to pull it off. At this point, I am confident though that if they don’t pull it off, someone else will. The trinity is not the only way to have roles; it’s only one way. A way that relies on simplistic and anachronistic mechanics to function.

Leave your comments or thoughts here.