In Defense of the Tankmage

“Tankmages ruin games.”  “If people can make a tankmage, that’s all they’ll make.”  “Tankmages are too difficult to balance.”  Why are developers so worried about people making characters they actually want to play?

To be fair, a tankmage, improperly handled, can cause an inordinate amount of damage to a game.  They can terrorize other players, or at the very least, marginalize them.  If you can do tons of damage, rarely get hurt, and be able to heal yourself why have other people around?  You have to make it so that being a tankmage isn’t the same thing as being unstoppable, which is what people tend to immediately jump to when making a class that’s very good at doing more than one thing is brought up.  So let’s dispense with that right now:  I’m not talking about making all-powerful characters.  I’m talking about making characters that are very good in more than one thing instead of being very good at one thing or kinda okay at more than one thing.  Beyond that, I’d go one further and say that the games in the marketplace right now that tried to balance around having dedicated classes and hybrid classes have, in most cases, managed to instead create a bunch of unintentional hybrids and ended up back at the place they were afraid they would anyways.

 WoW’s a good example.  You’ve got, theoretically, a dedicated tank (warrior), a dedicated healer (priest), and a couple of dedicated DPS classes (rogue, mage).  Then there’s the hybrids: hunter and warlock (tank/dps) and shaman, paladin, and druid (healer/tank/dps).  Of course, that’s not how they actually play out in practice.  Specced and geared correctly, a warrior can rival a rogue’s DPS.  Rogues and mages have great damage mitigation through mezzes.  The priest has one of the most effective grinding specs in the game, effectively raising their dps to near, if not quite at, mage levels.  The hybrids are either indifferently good to the point where they’re brought along because of one or two skills (cleanse and buffs) or because they rival the “dedicated” class in some fashion (bear druid tanks, warlock DPS), which usually means ignoring to the point of unsummoning their pets for hunters or wearing cloth armore for paladins in some cases.  It’s a rare major (and even minor) patch where some ability in some class isn’t being tweaked for being over- or under-effective, with major class overhauls being very common.

 City of Heroes and City of Villains offer us an interesting viewpoint as well.  In City of Heroes, the designers originally intended to make a tankmage style of design, but were unable to balance it.  In addition, as Jack Emmert put it in a recent post on the CoH boards, the archetypes in CoH were designed with tank/healer/dps firmly in mind specifically because they weren’t certain it would sell without them.  Of course, the balancing of the archetypes in that game has been one of the major, constant headaches for them, continuing to this day with a coming revamp of the Blaster’s inherent ability Defiance, in hopes that they’ll stop underperforming.  It must be said, however, that some of the balancing difficulties were as a result of a lack of MMO development experience on the CoH team, so it can’t all be laid at the feet of the dedicated/hybrid problem.  Still, every Issue up to and including Issue 6 had major tweaks intended to fix balance issues between the five main hero ATs.

When City of Villains came out, they largely discarded the idea of dedicated tank/healer/dps classes and instead made all but one of the classes a hybrid.  Brutes can DPS and tank in a pinch, Dominators can mez and DPS, Corruptors can DPS and heal/buff/debuff, and Masterminds tank, DPS, and heal/buff/debuff.  Stalkers are a bit of a problem child, as they’re the closest to the pure DPS role, and somewhat underperform at it except in a relatively narrow set of circumstances, much as Blasters do.  Dominators had a little speedbump at first, and still have some problems in high-end encounters, but they’re much better than they were previously.  Ultimately, the CoV archetypes “feel” more in tune with the source material than most the hero ones, particularly when comparing something like the Brute archetype and the Tanker archetype, both of which were supposed to represent superstrong, supertough comic book characters.

So what do these admittedly limited examples show?  First, the idea that building a “dedicated” class is going to help with balance is a fallacy.  This is doubly true if “hybrid” classes are included in the design of the game.  Dedicated classes almost always end up being either underpowered or de facto hybrid classes, simply because people will want to solo some of the time and dedicated classes often don’t do that well, at least if there are hybrid classes in the game.  Second, “hybrid” classes are the norm, and yet if they’re conceived of as “hybrid” from the beginning of design, they’re almost always either underpowered or forced to ignore a large chuck of their abilities for endgame content, where singular focus on specific abilities is the norm, almost entirely because of the idea that “dedicated” classes be the gold standard for a task.  Third, when developers place aside the hybrid and dedicated fallacy, they can create gameplay that is far more interesting than if they design to the tank/healer/dps ideal.  Fourth, letting characters be good at more than one thing, letting them be tankmages, isn’t inherently dangerous.

So how do we make tankmages that aren’t overpowering and that allow for interesting gameplay?  First, create consistent baselines.  If you look at the wonderful City of Data by RedTomax, you can see that, for example, blasters have a 100% damage modifier.  That’s important, because it may give us a glimpse into the mindset of the designers:  the dedicated DPS class for CoH, at least as it was originally launched, is used as the balancing point for the other classes’ damage.  Everything else, from a numbers standpoint, was seen as a “weak Blaster,” at least until some rebalancing happened.  But then it turned out that that didn’t work well, and Scrappers ended up getting 112.5% damage, and Tankers got a damage buff, etc.  At some point, it’s possible that Blasters’ damage will need to be tweaked, and then no one will be at the 100% damage modifier.  Arguably, that would have been a better starting point, divorcing the Blasters from being the standardized damage point.

Second, make sure you know your numbers.  Not just kinda, not well enough, make sure you know them.  It always amazes me how little designers sometimes seem to know the numbers to their own games.  To be fair, that’s often not their department, if they’re story guys or level designers.  However, if you’re the lead designer for the game, or the power designer, or the combat designer, you need to know your stuff about how the innards of the game work, and you need to know it well, or else you will never balance the game, never mind worrying about tankmages.  This is kind of a tangent, but I still remember the furor when Statesman came onto the boards during a round of Scrapper and Tanker balancing and said that Scrappers couldn’t ever get to Tanker level of resistances, then came back later and said that he meant “without dipping into pool powers.”  There were player-written guides on the boards that told precisely how to do what he said was impossible staying entirely in the Scrapper powers.  Know your numbers, or it will come back to haunt you, one way or another.

Third, don’t be afraid to do away with classes entirely.  It’s scary, I know.  The players could do something crazy and create wholly terrible or unbelieveably buff characters.  That’s what your alpha/beta period is for.  For the most part, if you’re smart, you know your numbers and, most importantly, listen to players when they say, “hey, this seems broken,” at least enough to check it out yourself or, possibly even better, have them show you what they can do, you’ll catch most of it, and rules-based sets should can hopefully prevent it from happening in live.  And when someone catches it, reward them!  Being in beta, while still seen as a reward in and of itself, isn’t enough to report some of the really juicy ability combinations sometimes, particularly if they look obscure enough to be make it to live and be exploited.  Give prolific bug hunters badges, a place in the QA credits in the manual, free time, something for helping you above and beyond the call.

 Fourth, pick one thing that everyone is good at, then allow them to be great at it.  In CoV, for example, just about every AT can be good at DPS.  Some are a little better at it, some are a little worse at it, but no matter what team you put together in CoV, you’re going to have “enough” damage.  Brutes, Corruptors, and Stalkers can be great at it, but each in a different way, and each of them with specific pitfalls for their playstyles.  Each of those ways of being great at damage complements the others when a team is rolling along well together, but none of them gets in the others’ way if the team isn’t meshing well.

Fifth, make sure everyone is bad at something.  This is less important, and it’s not necessary that everyone be bad at the same thing; in fact, it’s important that characters each be bad at something different.  It gives them a reason beyond simple community to team.  It shouldn’t be so crippling that players can’t solo, but it should be noticeable enough that they’ll enjoy teaming with someone that can fill that hole in their abilities.  It’s vital that the thing that each character is bad at be important, though.  The UO tankmage was devastating largely because, for a combat role, being supertough and doing tons of damage was all that was necessary, particularly given that mages could heal and buff themselves while debuffing enemies as well.

 As a hypothetical, let’s say that we wanted to revamp CoH to make it a classless game, with the ability of people to turn themselves into tankmages.  How would we do that safely?  First off, we take the various pools that are availble to each class and combine them into appropriate lumps.  That gives us Melee Damage (all of the Scrapper, Stalker, and Brute primaries and the Tanker secondaries), Ranged Damage (Corruptor and Blaster primaries and Defender secondaries), Controls (Controller and Dominator primaries), Pets (Mastermind primaries), Personal Defense (all of the other Scrapper, Stalker, Brute and Tanker powers), and Team Defense (Defender primaries and Mastermind, Corruptor, and Controller secondaries).  Some of these will probably need to be rebalanced, in terms of power placement inside of the pools, given the way powers will unlock in the new system.  The four-power Power Pools like Flight, Leaping, Presence, etc. could either be retained or discarded, or possibly upgraded with additional powers and turned into “real” pools as necessary.  At level 1, all players pick one of the Melee Damage, Ranged Damage, Controls, or Pets pools and one of the Personal Defense or Team Defense pools and designate one of them to be their primary and the other to be their secondary.  The highest-tier power for primary pools becomes available at level 32, and the highest-tier power for secondaries becomes available at 38.  At level 6, they unlock the four-power power pools, which work just as they do in CoH now, in terms of how the powers in the pools unlock.  At level 12, 22, 32 and 42, they can pick one of the other “major” pools, if they choose.  The key to balancing this is that those pools are considered to be ten levels higher than their starter pools and are on the “secondary” power pick track, so some of the higher level picks won’t be available to them.  In addition, picking higher level powers require picking a certain number of minimum powers so that someone can’t simply pick all of the Pets powersets and have a horde of ninjas, thugs, robots, and zombies deployed.  A more concrete example follows:

Captain Smackdown picks Super Strength (Melee Damage) as his primary and Invulnerability (Personal Defense) as his secondary.  At level 12, he decides to dabble in Energy Blast (Ranged Powers).  At 22, he picks Cold Blast (Ranged Damage).  At 32, he picks Leadership (Team Defense).  At 42, he takes Robotics (pets).  On first blush, folks might say, “Wow, that’s overpowered.”  However, remember that just because he has access to the pools, that doesn’t mean that he’s got all of the powers.  He still only gets a total of 24 power picks, spread out across those choices.  If the pools are designed correctly, picking one thing means sacrificing another, to the point where the character is going to be underpowered for some application.  In the case of Captain Smackdown, he has excellent DPS and good defenses, but nothing much in the way of heals.  He merely has more options to build the character that’s interesting to him, one that is, dare I say it?  A tank mage.