Threat-Based PVE is Lame and How I’d Fix It


The traditional model of PVE tanking, which revolves around the “threat” (aggro) mechanic, severely limits the gameplay experience of MMORPGs. I’ll share an alternative model that would provide a much more dynamic combat experience.

PVE tanking and the “holy trinity” (tank/healer/DPS) themselves are not the issue – the problem is that they tend to be implemented with threat-based mechanics. In this article, I’m not going to advocate ditching the “holy trinity” as ArenaNet (AN) did with Guild Wars 2 (GW2) last year.

Why Threat-Based PVE is Lame

Here’s what’s wrong with threat-based PVE:

  1. It defies common sense. Why would an intelligent boss and his buddies attack the tank, who is the toughest, most-armored, highest HP target who is also the least capable of inflicting meaningful damage?
  2. It dumbs down PVE encounters. You have threat, boss attacks you. You don’t have threat, you can do whatever you want
  3. It creates an environment where where the focus tends to be on the UI not the combat: watching threat meters, tracking cooldowns, using abilities is optimized rotations, and listening to addons/mods tell you what to do and when – as opposed to having to continually read and react to situations and what your opponents are doing

The last two points combined have severe implications in most games where the vast majority of content and boss fights are simple tank-n-spank:

  1. The majority of PVE content fails to prepare players for much more challenging boss fights later in endgame which require not just threat-management but also high coordination and awareness. The hard boss fights are often referred to as “guild breaking” but the fault here lies IMO not so much with the talent in the guild but rather the lack of an experience provided by the developer to train players to be more skilled and aware
  2. Players are trained to expect predictable opponents – this becomes very problematic for PVE players foraying into PVP, since human opponents are very intelligent and far less predictable than threat-dumb bosses

My Fix for Tanking: Think NFL Offensive Lineman

If we ditch threat-based tanking, what would it look like to have a model where tanking is relevant but that the combat is much more dynamic and based on rich coordination?

Let’s consider how tanking works in the National Football League (NFL), which is American pro football.

In the NFL, the quarterback (QB), who is a squishy RDPS, is protected by five offensive linemen, who are the tanks. An offensive lineman (OL) keep defensive players from reaching his QB; the OL blocks, shoves, knocks down, and holds opponents. An OL is therefore both a protector and a bully.

But here’s the thing. In order for an OL to “tank” effectively, he needs his QB to work with him. Both players have to be on the same page as to what the QB and rest of the team are doing, and the movement and positioning of the QB has to be synchronized with the protection scheme of the OL. Moreover, when things break down in real-time, the QB has to move in such a way that his OL can continue to shield him. An OL also has to adjust his blocking assignments in real-time to counter the defense.

What if PVE combat works like this, where the tank uses abilities to control mobs and protect friendlies, and the trinity is constantly working in real-time to address situations as the boss and mobs fluidly switch targets and tactics, because there is no concept of sticky aggro? Combat shifts from staring at your hotkey bars and UI elements to paying attention to what the boss and mobs are doing and what your friendlies are doing, and taking appropriate action as an individual and group. Tanking becomes a team effort: the healers and DPS have to pay attention and work with the tank.

With such a  model, developers could make bosses and mobs smart, and encounters would be much more fluid, challenging, and fun. Combat would keep you on your toes, so to speak. Boss fights could still have stages, if that’s your thing. But the days of boring tank-n-spank would be gone, or the days where you watch TV while raiding as a RDPS, because let’s face it, you don’t need to watch most of the combat these days, and that’s rather sad.

Aside from improving the combat gameplay experience, there are other positive implications of the “Offensive Lineman PVE Tank” model:

  1. Players would have the expectation of working together. I saw this in GW2, where the collaborative nature of the game mechanics encouraged players to help one another
  2. PVE players who foray into PVP would be much better prepared – they’ll have gotten reps in PVE and learned how to pay attention / avoid tunnel vision, counter unpredictable opponents, move properly i.e. not backpedal, etc. Simply put, this model of PVE would make players better
  3. Tanks would have the same role in PVP as they do in PVE (control and protect). Unfortunately in many games, tanks simply have no functional role in PVP
  4. Tanking can be implemented in terms of abilities not numerical factors. With threat-based tanking, tanking successfully is dependent on numerical factors: threat generation, mitigation, avoidance, and HP. With the proposed model, tanking can could be implemented based on weapon selection / abilities (bolas, nets, shield bashes, shield walling, Vulcan death grips, etc) and character customization, which means that any class could play the tanking role. This would allow for more flexibility for group compositions – and no more waiting in Ogrimmar or Ironforge trying to find a tank.

Follow Taugrim

Taugrim on Twitter Taugrim on Facebook Taugrim on Google Plus

About these ads
Posted in Game Design, Guild Wars 2, PVE, PVP, World of Warcraft
57 comments on “Threat-Based PVE is Lame and How I’d Fix It
  1. Conwolv says:

    Hey Buddy! Nice to see you posting on something not World of Tanks.

    As to your suggestion; I could see it work. However, it really depends on execution.To be honest, I think little bit of column A and B might be better though.

    By making the threat/aggro mechanic smarter, you could do what you suggest while not changing the overall PVE combat mechanics players are used to.

    The tank still tanks the mob, but the mob will actively either try to attack the DPS or Healers in the party. The tank has to react to these situations and control the mob with things like knock backs, abilities that turn the mob away, stuns, etc.

    But during the fight, threat still needs to be maintained.. the mob has to “deal” with the tank, who’s actively preventing them from getting to the others.

    Clearly this would require fine tuning, but doesn’t drastically change the mechanics people know and like.

    • taugrim says:

      The tank still tanks the mob, but the mob will actively either try to attack the DPS or Healers in the party. The tank has to react to these situations and control the mob with things like knock backs, abilities that turn the mob away, stuns, etc.

      Exactly.

      But during the fight, threat still needs to be maintained.. the mob has to “deal” with the tank, who’s actively preventing them from getting to the others.

      In that case, there really isn’t threat, is there?

      The tank is being a pain in the ass and locking down the mob, so that the healers and DPS can do their job.

      Maybe the mob gets mad and decides to start whacking on the tank, which is preferable to having the mob attack a squishier player.

      • Conwolv says:

        It would be to a point. Mob attention would be ranked, much like it is now (first on threat, etc). However, instead of the mob only focusing on the tank like they do now, they’ll go after the next highest on the threat list from time to time. Tanks job is to prevent that from happening through CC abilities.

        • taugrim says:

          Mob attention would be ranked, much like it is now (first on threat, etc).

          Ah, but why have threat in the first place? Is there really a need for a numerical indicator of who the mob wants to attack?

          Dispose of threat, have your eyes tell you what the mob is doing :)

          • There can still be threat, but a boss/mob thinking the tank is the highest threat is ridiculous (as you stated as well). It can be based on what player is the most threatening to the mob’s existence. So threat should be based on who does the most damage or who is able to keep up or protect the damage dealers the best. But the mob should also factor in what threat he can take out the easiest.

            It could be something like this: There is a certain number based on the damage a person does (or other threatening factors like crowd control and stuff). This could reflect in a number X.

            Then there is a number Y that reflects the killability of the player. The higher the number Y the harder it is to kill it. This should be based on proximity (farther means higher value of Y), toughness (the more defence a player has the higher the value of Y).

            Now the mob/boss should pick the target where the X/Y ratio is the highest. Different mobs can have different values of X and Y, based on there vulnerabilities and abilities. If the boss got a gap-closer then that makes it easier for the boss to target someone far away and that should reflect in a lower value of Y.

            If a mob has a crowd control ability he can control the player with the highest value of X and go for the target with the highest value of X/Y. Note that if the controlled player also was the player with the highest value of X/Y that it does not necessarily have to target him still. The fact that the player is controlled means his value of X will instantly decrease which will lower the ratio X/Y. You can have mobs who are vulnerable to certain attacks and that will change the value of X for certain players.

            You can have smarter mobs who are more capable of determining who is the highest treat and you can have mobs who can’t really determine it and are just smashing whatever is within reach. You can throw in some little random factors to prevent a boss from becoming too predictable. Randomness can vary from mob to mob. Even a boss that is completely random can be fun.

            This kind of threat system could result in all kind of different situations, even with the same mob.

  2. Reminds me a bit of how Warhammer Online attempted to address this. Though yes, there was still a threat/taunt mechanic, tank classes could also obstruct opponents via collision detection. Additionally, taunts worked in PvP by debuffing adversaries who attacked anyone besides the tank(s).

    That said, I think the main obstacle is (and perhaps always will be) player inertia. Every effort I’ve seen to change the status quo is usually resisted, unfortunately.

    • taugrim says:

      Additionally, taunts worked in PvP by debuffing adversaries who attacked anyone besides the tank(s).

      WAR went partially there, by making PVE abilities have an application in PVP. And as you said the collision detection and such. And the Guard mechanic was innovative.

      That said, I think the main obstacle is (and perhaps always will be) player inertia. Every effort I’ve seen to change the status quo is usually resisted, unfortunately.

      I think that’s because many efforts to innovate haven’t delivered sustainably compelling experiences.

      Carbine with WildStar is thankfully challenging a lot of the silly assumptions that have girded combat mechanics in MMORPGs. Their take on aiming and CC look wayyy better than anything the market has seen.

      • Conwolv says:

        I’m really looking forward to Wildstar. I’ve had a couple of opportunities to play Wildstar so far. While PVP has been fairly ‘spammy’ the PVE has been rock solid. Hopefully PVP gets fine tuned.

        • taugrim says:

          I’ve had a couple of opportunities to play Wildstar so far. While PVP has been fairly ‘spammy’ the PVE has been rock solid

          Good to hear!

          I haven’t played WildStar yet but it looks like the most interesting of the upcoming games, because the designers are rethinking combat from the ground up.

  3. Blegvad says:

    I think it’s a great idea. GW2 was a major letdown on the PVE side of things, they might not have had a hard trinity but that loss was not replaced with compelling gameplay. I think defined roles are essential in terms of creating interesting gameplay and when you homogenize to the extent that GW2 did it all starts to not matter: “We’re all DPS?”

    To extrapolate on your concept further I think a system with no threat (or without taunts) could actually create a “quadrinity” where a support role that CCs and manipulates the environment which helps the offensive lineman style tank stay in control could be really great not only in terms of gameplay but would truly create the fourth pillar that we’ve been missing.

    Nice read as always!

    • taugrim says:

      I think it’s a great idea. GW2 was a major letdown on the PVE side of things, they might not have had a hard trinity but that loss was not replaced with compelling gameplay. I think defined roles are essential in terms of creating interesting gameplay and when you homogenize to the extent that GW2 did it all starts to not matter: “We’re all DPS?”

      Yep, a lot of the boss fights in GW2 ended up being one long fight of simply wearing down the boss’s HP bar, without particularly distinct stages or mechanics. And a lot of the dungeons were farmed by very limited compositions.

      What I propose still keeps the concept of tanking around – because I personally love tanking – but it makes it far more interactive for both the tank and the other players.

      Combat should be fluid and engaging. I PVE tanked WoW in Vanilla and TBC, WAR, etc…hundreds of hours, but I found the threat-based models actually limited the development of my guildees to the extent that I played well. E.g. we had a guild healer who loved healing, who gave it up after she had a hard time in a heroic TBC instance (one in SSC with a hydro boss and round room), because in order to stay alive she had to keep moving the whole time, but that was something she wasn’t used to at all. She ended up re-rolling to a Rogue. But the game failed to prepare her, and that’s a shame.

  4. Baffle says:

    Pretty good article. I agree, I’d love to see devs move away from threat-based encounters.

    Have you taken a look at the new PvE raids in SWTOR? They offer more mechanic-based fights that require coodination and a lot of teamwork. They’re definitely moving in the right direction. Oh and I know you’re aware but TOR pvp is a ton of fun for tanks- they have purpose!

    • taugrim says:

      Have you taken a look at the new PvE raids in SWTOR? They offer more mechanic-based fights that require coodination and a lot of teamwork

      It’s good to hear SWTOR has moved in that direction. RIFT had very complex raid fights that were a lot of fun, from what I heard from my guildees who were in Maximation.

      People on Twitter commented to me that EQNext is planning on threat-less PVE, so we’ll have to see how their model works.

  5. Blegvad says:

    Exactly – to go even further I think the concept of limited action sets (LAS) that we’ve seen in TSW, GW2 and coming up in Wildstar and EQNext might be able to create further freedom from the class/roles paradigm that we’ve been stuck with forever.

    Abstract example: Every class can perform any of the 4 roles (Support/DPS/Tank/Heal) depending on which “deck” or set of abilities you have loaded. What defines your class at that point isn’t what you can do but HOW you are able to perform that role.

    You can still be a specialist (melee tank, wizard control, priest healer, etc.) but that becomes just one possibility inside a much larger game where it’s all about how the team works together to defeat challenging content and not about how your initial choice pigeon holes you into one function forever. It also opens up new possibilities (wizard tank, melee control, priest dps) that would provide a much deeper experience for the players especially if those abilities and synergies can be mixed and matched. I think that is where modern MMO games are going to make their mark.

    • taugrim says:

      Exactly – to go even further I think the concept of limited action sets (LAS) that we’ve seen in TSW, GW2 and coming up in Wildstar and EQNext might be able to create further freedom from the class/roles paradigm that we’ve been stuck with forever.

      Yea, I feel that classes have historically been too limited in their capabilities.

      GW2 got things moving in the right direction by tying abilities to weapons, so that you can swap weapons to have different capabilities.

      You can still be a specialist (melee tank, wizard control, priest healer, etc.) but that becomes just one possibility inside a much larger game where it’s all about how the team works together to defeat challenging content and not about how your initial choice pigeon holes you into one function forever. It also opens up new possibilities (wizard tank, melee control, priest dps) that would provide a much deeper experience for the players especially if those abilities and synergies can be mixed and matched. I think that is where modern MMO games are going to make their mark.

      Ah yes, I might finally be able to play a melee mage!

      But seriously, opening up choice to the player, instead of having limited archetypes, seems like a lot more fun.

      It sounds like the next wave of upcoming games are far more flexible in terms of class customization.

  6. Hi Ed, Great seeing this article.

    It should be fun seeing how the next patch of MMOs attempt to address this issue.

    The idea of the smart boss is interesting, but so many possible pitfalls in encounter design.

    Especially when interesting encounter design is defeated by class stacking. For example, a boss that is defeated by brining in all rogues, who use smoke bombs kicks and stuns to keep the boss off balanced the entire fight.

    Or simply ignore all game mechanics and burn the boss down as quickly as possible.

    It would be so interesting to see the encounter design for a threat-less encounter.

    • taugrim says:

      The idea of the smart boss is interesting, but so many possible pitfalls in encounter design.

      I think even a limited set of AI rules for how a boss behaves would it make it feel miles more intelligent than the threat dumb mobs today.

  7. Ailozh says:

    Amazing article.

    What are your thoughts on Wildstar and their skill shots, break out gameplay and limited action skills? I really love what I’m seeing from Wildstar so far. I think they’re doing a lot of things right and it looks like it’s going to be challeging!

    • taugrim says:

      What are your thoughts on Wildstar and their skill shots, break out gameplay and limited action skills?

      Big thumbs up. They are re-thinking combat mechanics to make combat more interactive.

  8. Crow Born says:

    Hello Taugrim. I am glad to see more posts about mmo mechanics, especially some that don’t deal with the WoT – as that game never really was my cup of tea. They were good posts though and definitely worth the read, so keep them coming.

    As for your suggestions to alter encounter design, I quite agree that the current setup in most mmos is boring, borderline stupid, and ultimately doesn’t make sense from a real-world point of view.

    For months, a tankish guildmate of mine in WoW has complained about how he felt like a pinata in raids, and that nothing he did except from popping defense-increasing cooldowns really made a difference in encounters. His suggestions for spicing it up are quite similar to yours – remove threat so he actually has to be aware of who the boss is hitting after the first frantic five seconds of doing his rotation, ‘remove’ rotations all together and give tanks more cc abilities to help protect his allies, and just make the damned encounters more reaction-based and less about knowing strategy and facerolling your keyboard. This could stop players in pugs from blaming every single wipe on the tanks, and perhaps open up for the idea of needing more tanks in raids to keep the bosses from chasing down the 12515344211k dps mages, which they definitely would if there was no threat mechanic.

    ‘My’ point of view on this however, is from a healers and ranged dps perspective. I am more or less used to standing at max range from the boss, clicking on whichever name on the raidgrid that needed the most healing, or just spamming q, w, e to keep up my dps – and I am not sure any of your or my guildmate’s suggestions would really change my playstyle. If the boss was trying the whack down anyone else than me, I would simply focus my healing on that guy instead of the tank, relying on the tanks and dps to slow down the boss. If ‘I’ ever ended up on the boss’ radar, well… Bubble and run away at max speed, once again trusting in others to keep the boss away from his target. Regardless, my rotation would mainly remain intact.

    The solution would certainly not be to give everyone more cc abilities, as that would homogenize the classes as in GW2 – and at a certain point the amount of kicks, stuns and slows being thrown around would just be ridiculous. So I have no idea how to really make encounters more interesting for everyone… Except.

    In Siege of Orgrimmar, the most recent raid to be added to WoW, there’s a certain fight that tests every player to their limit – the Kor’kron Dark Shaman. This is a fight where success does not lie with the tank, healer or dps, but rather with the individual, and his ability to react to danger. It is a ‘very’ movement intensive fight, where everyone constantly have to be aware of ground-effects, or soon end up a very moist puddle on the floor. The fight forces most classes to give up their rotations, and rely on instincts to survive, and I really love that fight. For once I were not just staring at a grid or the boss’ healthmeter, but on what was actually going on in the fight. To me, this encounter was a vast improvement over any other fight I have ever done before, and I hope to see more like this.

    While adding more movement intensive encounters, perhaps coupled with no-threat-mechanics, may be a good idea, I am not sure it would really be the way to go. But since this is already turning out to be a humongous post, I won’t bore you further with my ideas and concerns. Have a nice day.

    • taugrim says:

      His suggestions for spicing it up are quite similar to yours – remove threat so he actually has to be aware of who the boss is hitting after the first frantic five seconds of doing his rotation, ‘remove’ rotations all together and give tanks more cc abilities to help protect his allies, and just make the damned encounters more reaction-based and less about knowing strategy and facerolling your keyboard

      I agree 100% with your guildee.

      This could stop players in pugs from blaming every single wipe on the tanks, and perhaps open up for the idea of needing more tanks in raids to keep the bosses from chasing down the 12515344211k dps mages, which they definitely would if there was no threat mechanic.

      And therein lies one of the huge flaws with threat-based tanking – the tank has to have on-par gear with the DPS of the tank won’t be able to hold aggro. So outcomes are dictated by gear, not skill.

      If the boss was trying the whack down anyone else than me, I would simply focus my healing on that guy instead of the tank, relying on the tanks and dps to slow down the boss. If ‘I’ ever ended up on the boss’ radar, well… Bubble and run away at max speed, once again trusting in others to keep the boss away from his target. Regardless, my rotation would mainly remain intact.

      The issue right now is it’s rather easy to stay alive as a healer or RDPS if you temporarily have aggro. You can bubble and run as you say until the tank can taunt the boss off you.

      But what happens if the mobs are smart? Maybe you have to use LoS more than you do now, standing physically behind your tanks and melees so that the mob archers can’t shoot you. Maybe you have to counter the mobs CC which are designed to help trap and shut you down.

      It think even the whole concept of having standard rotations is applicable mainly in a world of predictable encounters.

      It is a ‘very’ movement intensive fight, where everyone constantly have to be aware of ground-effects, or soon end up a very moist puddle on the floor. The fight forces most classes to give up their rotations, and rely on instincts to survive, and I really love that fight. For once I were not just staring at a grid or the boss’ healthmeter, but on what was actually going on in the fight. To me, this encounter was a vast improvement over any other fight I have ever done before, and I hope to see more like this.

      And that in a nutshell is the issue with most threat-based PVE today – they lack the dynamic combat and thrill factor you described.

      Granted, not all encounters need to be that hard, but that is in the right direction.

  9. Liang Nuren says:

    Hey,

    One thing that isn’t immediately clear is how the ‘lineman’ approach works when you factor in hostile RDPS. I think that once you factor in the idea of RDPS as a whole you end up seeing that *Tank* isn’t actually very much of a role. What I think is inevitable in situations like that is that everyone starts to focus on DPS or CC while still maintaining enough tank to do not be vaporized before the heals or reps land.

    BTW, I think you should try out Incursions and Wormholes in Eve. The NPC AI there constantly shifts and looks for weak points in the player’s defenses, and failing that attempts to kill what it perceives to be the most effective players. And, of course, there’s always the added fun of what happens when players drop by while you’re plexing. ;-)

    • taugrim says:

      One thing that isn’t immediately clear is how the ‘lineman’ approach works when you factor in hostile RDPS. I think that once you factor in the idea of RDPS as a whole you end up seeing that *Tank* isn’t actually very much of a role

      Sure it is.

      The tank can throw a net on the RDPS to disable their firing, or whack them in the arm to make their shots miss.

      The tank and the friendlies could position themselves so that they are not in LoS of the hostile RDPS.

      That is, make positioning matter.

      What I think is inevitable in situations like that is that everyone starts to focus on DPS or CC while still maintaining enough tank to do not be vaporized before the heals or reps land.

      It depends on the class design.

      For example, you could give tanks the most CC options, and other classes CC more intended for oh-crap situations.

      Games tend to lose class balance when some classes can do too much DPS *and* CC.

      The NPC AI there constantly shifts and looks for weak points in the player’s defenses, and failing that attempts to kill what it perceives to be the most effective players

      Sounds like fun!

  10. Scottpoet says:

    To some extent for those of us who play Firefall (fps) are already seeing this in open world PVE right now. Red5 studios have done an amazing job with their AI, to the extent I am still not entirely sure how threat works in their game.

    The trinity is mostly intact and all the roles still have their niche while remaining universally useful.

    • taugrim says:

      To some extent for those of us who play Firefall (fps) are already seeing this in open world PVE right now

      How would you rate that game compared to others we’ve played? I haven’t given it a look yet.

      • Scottpoet says:

        ehh I keep starting to write a short reply and it continues to turn into a book. So short version.
        It has the best open world PVE I’ve ever played by far. It is just no contest. Progression is a good mix of horizontal and vertical progression. PVE has a much higher skill cap thanks to it being FPS. Grouping is pretty organic, and mission style questing pretty great as solo and amazing as a group.

        I’ve probably put in 200+ hours of play time into a starter zone O_o now that I think about it.

        The pvp was taken out last time I played to be worked on, so I can’t really comment. However the controls are so solid I have trouble imagining it being terrible.

        • taugrim says:

          I’ve probably put in 200+ hours of play time into a starter zone O_o now that I think about it.

          I installed and tried Firefall last night for about an hour and a half.

          The combat was less….tactical?…than I was expecting, but I’ve been playing a lot of WoT so maybe that’s influencing my perception.

          • Scottpoet says:

            Yeah I guess I should have warned it is a true, halo esque, first person shooter. Most problems can be solve with good shooting and tactics aren’t much of a thing if everything is dead.
            However as you keep playing, you will find yourself dying over and over again from certain situations. That’s where tactics, gear and planning become a thing.

            Myself… sometimes all I need is that clean headshot to keep me going.

            • taugrim says:

              Myself… sometimes all I need is that clean headshot to keep me going.

              I must be too low-level and geared to see this. Even with a Recon battleframe I have to shoot NPC mobs like 20 times to kill them.

              • Scottpoet says:

                Accord recon runs the R36 assault rifle, you won’t get a sniper rifle until you unlock a Nighthawk or Raptor battle frame.
                The only battleframe i wouldn’t suggest to start with is the recon suits tbh. They unfortunately are incredibly gear dependent, squishy and require the highest skill. It is just easier to progress with other frames and come back to it when you already have gear access.
                My ingame name is Clikk and my peeps & I will run with you if ur on.

  11. Whiskers says:

    I love hearing about new concepts for MMO gameplay. Removing threat while maintaining the trinity could make for some really fun raids. Some of my most enjoyable moments in a wow raid was to see how far our group could get on a boss after the tank died. Essentially there was no threat control and we had to kite and utilize all of our class mechanics rather than obey our rotations. Good times.

    • taugrim says:

      Some of my most enjoyable moments in a wow raid was to see how far our group could get on a boss after the tank died. Essentially there was no threat control and we had to kite and utilize all of our class mechanics rather than obey our rotations. Good times.

      Yep, those are seat-of-your-pants fun, especially when you managed to finish off the boss’s last few % HP with no tanks.

      Or in my case, tanking without sufficient healing support while kiting and using every possible tool to buy precious seconds for the raid DPS.

      I used to love finding windows in boss fights as a Feral Druid tank in TBC where I could innervate a low-mana healer or RDPS and shift back to tanking without missing a beat.

      • Gameldar says:

        Feral druid in Cataclysm was awesome for that same reason – except it was about squeezing out as much DPS in those breaks from being the focus and getting back into bear form before you got squished again… then they completely broke that with Mists…

        I was initially sceptical to your idea – because I love tanking (and is the one main thing I miss from only playing GW2 these days – although you can tank as such because there is a threat mechanic still – but I do very little PvE in it anyway). However then I was thinking about it – the most fun I’ve had tanking is where the threat mechanics have been iffy. I’ve always preferred the off tank role because it was often more about picking up hundreds of little mobs rather than sitting there an timing my cooldowns. My all time favourite tanking experience is where I was a couple of tiers in gear behind the DPS and so my hold on the threat was very tenuous and I was having to be very active in switching targets and taunting/mangling them to maintain the aggro (which was also the case in WoW at the beginning of Wrath when DKs were OP and I was tanking on a warrior).

        I played a mud back in the day (Xyllomer) that had this sort of tanking paradigm. The steps the ‘tank’ had to do was pretty simple, just execute the guard command – although it was a bit more work because you had to “guard x from y” (if I remember correctly) but the target switching certainly made it so you had to pay attention.

        The thing I’d still really like to see even where threat was not a factor was that tank DPS was still a factor (and not just a given) – such that it is not just about chucking out the CC at the appropriate times but you have a dual focus on also trying to do as much damage while still doing the tanking role. But with all these things – they all make it harder to balance as well (particularly where you want the cross-over into PvP – tough to kill/able to do decent damage/has very good CC…

  12. Matt Herrera says:

    I like the concept of “Tank-as-lineman” but it would all have to come down to implementation. Maybe a hybrid system could be implemented? Like instead of traditional Threat, enemies have something like a “Frustration” meter built into the UI?

    Maybe most monster AI is programmed to attack the first thing that attacks it or the closest enemy for the first few seconds. After that first few seconds it starts zeroing in on the healers or DPS, as it realizes that the tank isn’t dying and/or something else is hitting it far harder than the tank is.

    From there, the tank’s job is to body-block, mini-stun, interrupt, snare or whatever to keep the monster from reaching the squishies or turning around to stomp the melee DPS. As he does so, the monster builds up “Frustration” (which should be build into the UI) because it can’t get to what it wants to kill. You can add mechanics as the monster’s Frustration builds, like it gets movement speed increases or more damage, or starts to do knockbacks or cleaves.

    Once the monster reaches max Frustration you can add other mechanics to test coordination. Maybe the monster has had enough of the Tank’s BS and decides to end the annoyance once and for all. It could use knockbacks, disables or good old fashioned burst damage to get the tank out of its face. Or maybe at max frustration the monster starts thrashing about wildly, striking everything in melee range for massive damage until it calms down. Or maybe it starts using a breath weapon or starts lobbing ranged attacks at the RDPS and healers.

    • taugrim says:

      Hi Matt,

      I think you just answered your own question about how it could work, with your description of the Frustration mechanic.

      It’s really important to not make it measurable though. Threat meters make combat so faceroll easy. Your eyes tell you what the boss is doing and who it’s trying to squish.

      • Matt Herrera says:

        Well, maybe not a meter, but some sort of system to let players know that something is about to happen. SWTOR did a lot of audio cues for phase transitions, but there was no threat meters or indicators. Often the only indication that you had pulled aggro was that the boss turned around and pasted you (especially if you were Assault Vanguard, like I was).

        A system like this would need, if not a meter of some kind, an audio/video cue (other than target of target). Since collision detection would be a must for a system like this, perhaps strictly-enforced turn rates on mobs, and “taunts” act to force the target to face the tank. Otherwise melee dps is going to have a rough time of it.

        I think a frustration system combined with collision detection and something like WildStar’s telegraph system would be a blast to play.

        • taugrim says:

          A system like this would need, if not a meter of some kind, an audio/video cue (other than target of target).

          That would be like the reticles on the ground for boss AOE in GW2, before the AOE actually landed.

          Since collision detection would be a must for a system like this

          Yes, CD would really support the experience.

          “taunts” act to force the target to face the tank. Otherwise melee dps is going to have a rough time of it.

          I would say rather than implementing taunts, provide abilities to make the tank a bully, which enables MDPS to attack with impunity since the tank is controlling the mob. If the mob starts attacking a player, the tank could “intervene” and take the hit, or apply some CC to the mob that allows the MDPS to get away then re-engage.

  13. Hi Ed,

    I’m going to echo what others are saying here. It’s always fun to see a new post show up, and equally exciting to see you theorizing on how to shake up traditional MMORPGs again.

    I loved the initial ideas TERA pitched, especially the Lancer tank. Positioning is always a key to success in tanking, but I feel like it’s easy for some to over look that aspect. Actually having collision detection and the ability to block the movement of the mob you’re tanking was really exciting coming from WoW/Rift/SWTOR.

    As I understand it, the Lancer is only half way to what you’re describing. The blocking seems to be more about how you mitigate personal damage intake since threat mechanics still apply heavily.

    Still waiting to see the kickstarter for Age of Taugrimcraft: the horizontal scaling, open world PvP MMO with well designed stealth, viable pvp tanking, collision detection and no-threat PvE mechanics.

    • taugrim says:

      It’s always fun to see a new post show up, and equally exciting to see you theorizing on how to shake up traditional MMORPGs again.

      Yea I’ve missed blogging and interacting with y’all.

      Can you believe it’s just over 2 years since we played SWTOR Beta at NY Comic Con?

      As I understand it, the Lancer is only half way to what you’re describing. The blocking seems to be more about how you mitigate personal damage intake since threat mechanics still apply heavily.

      One of the interesting questions with an “Offensive Lineman as Tank” model is how beefy do the tanks really have to be? You’d figure fairly durable since some of their CCs might only work in melee range, but not so durable as it’s not assumed they’re going to be taking the vast majority of the damage.

      With this model, instead of face-tanking everything in sight, tanking would be more of an art – shield bash this mob over here, hamstring this other mob running at my healer, then charge a squishy RDPS.

      Still waiting to see the kickstarter for Age of Taugrimcraft: the horizontal scaling, open world PvP MMO with well designed stealth, viable pvp tanking, collision detection and no-threat PvE mechanics.

      ROFL.

      You know my game mechanics preferences well!

  14. “this becomes very problematic for PVE players foraying into PVP, since human opponents are very intelligent and far less predictable than threat-dumb bosses”

    PvP players also jump around like monkeys on crack for no apparent reason, which bosses JUST DON’T DO and which drives me nuts, because they need to HOLD STILL while I’m embarrassing myself/attempting to put a dent in them, plzkthx.

    In any case, I almost feel like the scenario as described DOES exist in some way in WoW … mostly when crap goes wrong. Say the paladin doesn’t have Righteous Fury on! Perhaps the main tank goes down! Then the juggling of the mob between players of varying squishiness is totally present.

    • taugrim says:

      PvP players also jump around like monkeys on crack for no apparent reason, which bosses JUST DON’T DO and which drives me nuts, because they need to HOLD STILL while I’m embarrassing myself/attempting to put a dent in them, plzkthx.

      LOL!

      In any case, I almost feel like the scenario as described DOES exist in some way in WoW … mostly when crap goes wrong

      It does, but not in a structured way. Also, bosses are tuned to deal massive damage that will 1- or 2-shot anyone except a tank, and I think this is part of the problem.

      • I can always tell when a tank in LFD is more accustomed to PvP than PvE, because he or she will literally jump all the freaking time, and generally never, ever hold still in one spot for more than a second, which makes the mob turn every whichaway and is annoying as eff if you’re a class that has to worry about positioning (I give up on rogue, I can’t handle it). It’s not the chain pull technique that people have perfected as of the past few patches – it’s that spaztastic “HIT ME I DARE YOU! YOU CAN’T, CAN YOU!? I’M BORED LOLZ” sorta thing.

        “Also, bosses are tuned to deal massive damage that will 1- or 2-shot anyone except a tank, and I think this is part of the problem.”

        …………… … … ..

        /has sudden, painful flashbacks to getting horribly, awfully and almost instantly maimed/murdered/daed by Garrosh Hellscream and his ninety million adds after the tank died

        Yes, as of this moment my only option is to run screaming and pray to Gawd some other living DPS manages to kill whatever is going to hit me before it actually does.

  15. Hey Taugrim! great to see you posting and finding time in between your busy schedule to game. Definitely miss your PvP guides man. Hope all is well though.

    As far as your topic goes. I think there are several things that have to go along with this in order for a non-trinity system to work. I applaud Guild Wars 2 for doing this and trying to break the mold a bit. However, there was a specific area that I think they sort of failed at when they tried to do this. This area being how class builds and boss mechanics tied into one another.

    Since this was touched on a bit above, I won’t go into too much detail. But in dungeons you ended up with monsters with very simple mechanics, yet extremely high health pools…and that was pretty much it. This meant that at the top level, a good party who knew the fight, could just take minimal cc and go full berserker builds and steam-roll content. Effectively making defense oriented specs or support specs useless in PvE, if you wanted to be the most efficient. Even at medium level play, you could still end up hurting your group if you didn’t have enough damage with certain boss fights, which is the case in most games, but you never really hurt your group if you didn’t have a tank-oriented player or a support. It was simply a matter of knowing what to avoid. And while Arenanet seems to be buffing a few other builds as of late, I still feel like they have a long ways to go before both tanks and support players really feel completely useful in PvE.

    Everquest Next is going with a similar non-holy trinity system, so it’ll be interesting to see how they go about doing it. I definitely think a system like this could work, but the player builds and monster mechanics need to go hand-in-hand for something like this to feel good for those types of players I think.

    On the topic of the holy-trinity though. I do think there are ways to have the trinity and make combat feel more organic. WildStar seems to really be trying to do this with their game, which is one of the reason why I’m pretty excited to play it.

    I actually go a chance to sit down with WildStar’s lead class designer and lead combat designer for ZAM. (Here’s the link: http://www.zam.com/story.html?story=33110 )

    They gave a few examples of dungeon mechanics where there are certain mechanics where the bosses ignore player aggro. Making it so that tanks have to actively use their CC’s intelligently and position correctly instead of just hitting the taunt button to hold threat. I think this, combined with their free-form targeting and complex telegraph systems are really going to make holy-trinity based PvE a lot more involved and exciting. Obviously, what sounds good on paper doesn’t always end up being the case. We’ll have to actually get in there and see how it all works out, but It looks promising so far.

    Here’s a dungeon video they showed at gamescom with some dev commentary highlighting the boss mechanics in one of the earlier dungeons. In case you haven’t seen it just yet. (http://www.youtube.com/watch?v=QG_82r0RzfI)

    All in all, I feel in order for both systems to feel more “organic”, it’ll really come down to games in the future getting their class design and combat design to synergize well with their boss mechanics. And while I don’t think we’ve seen that just yet, I think WildStar and Everquest Next are going to be the ones to pay attention to in this regard for both sides of the spectrum.

    • taugrim says:

      Hey Taugrim! great to see you posting and finding time in between your busy schedule to game. Definitely miss your PvP guides man. Hope all is well though.

      Thanks. It feels really good to post again and get some discussion going on game mechanics.

      This meant that at the top level, a good party who knew the fight, could just take minimal cc and go full berserker builds and steam-roll content. Effectively making defense oriented specs or support specs useless in PvE, if you wanted to be the most efficient.

      And while Arenanet seems to be buffing a few other builds as of late, I still feel like they have a long ways to go before both tanks and support players really feel completely useful in PvE.

      Yep, you hit the nail on the head.

      Everquest Next is going with a similar non-holy trinity system, so it’ll be interesting to see how they go about doing it

      They haven’t yet shared how it’s going to work in detail, right?

      Will be interesting to hear.

      I’m not for ditching tanks or tanking necessarily, just the current model of tanking is so limited.

      I actually go a chance to sit down with WildStar’s lead class designer and lead combat designer for ZAM. (Here’s the link: http://www.zam.com/story.html?story=33110 )

      Thanks for posting your interview, that was a great read :)

      I’ve been meaning to post about the WildStar combat mechanics but haven’t gotten around to that yet.

      Here’s a dungeon video they showed at gamescom with some dev commentary highlighting the boss mechanics in one of the earlier dungeons. In case you haven’t seen it just yet. (http://www.youtube.com/watch?v=QG_82r0RzfI)

      Interesting to watch.

      I really like the interactivity introduced by the telegraphs. Combat looks much fluid. And I really appreciate their take on training players starting at level 1, which you touched on in your article.

  16. Pander Beers says:

    Thanks for the awesome article – we’ve been talking about this in our guild’s VoIP all evening. Our first thought was to when SWToR gave tanks the Guard ability which was wonderful in PvP. If tanks were able to intercept projectiles by physically blocking with their body and positioning, we could see formations, too. The whole concept is eye-opening and has a ton of potential. I hope devs are reading.

    • taugrim says:

      Our first thought was to when SWToR gave tanks the Guard ability which was wonderful in PvP

      I preferred WAR’s implementation of Guard over SWTOR, as you could rotate it on demand. Both in both games as you said it gave tanks a meaningful role in PVP.

      If tanks were able to intercept projectiles by physically blocking with their body and positioning, we could see formations, too

      Yes, when you scale the combat, that is one of the implications – a bunch of tanks locking shields to hold a chokepoint, especially when outnumbered. Think Sparta / 300.

      Games with collision detection also tend to promote tactical positioning, although that technology seems expensive in terms of server resources, so I don’t know how long it will be before we see mass combat that scales gracefully even with collision detection.

      I would love to see more tactical combat, especially in World PVP. But that would only exist to the extent that the combat mechanics encourage it.

      • Lymain says:

        I didn’t play WAR, but I just remembered how much fun it was to PvP tank in early SW:TOR before they nerfed the burst damage. Swapping Guard to the player the enemy was focusing/bursting, taunting to mitigate bursts, CC’ing to disrupt burst…that was a blast. The gameplay was very dynamic, non-rotation-based, and it required good awareness.

        I did kinda feel like an offensive lineman. Make PvE combat like that, and I’ll probably be entertained for quite a while.

  17. Lymain says:

    Unless I’m mistaken, EQN is dropping the trinity, so I doubt they’ll be a good case study for this type of combat.

    It’s an interesting idea, but I’m not sure we’ll ever have combat mechanics that are still interesting after hundreds (if not thousands) of hours of gameplay, which is what people expect from MMOs.

    • taugrim says:

      I’m not sure we’ll ever have combat mechanics that are still interesting after hundreds (if not thousands) of hours of gameplay

      GW2 proved to me that more actiony combat creates a more consistently engaging experience. Action combat requires constant real aiming, especially for melee classes, and it was up to the player to manage line-of-sight and range.

      This concept will be extended even further with WildStar and Carbine’s take to counters when CC’d.

  18. Scottpoet says:

    I like the offensive line idea, but I also think the threat mechanic isn’t really to blame for rather lackluster tanking and fights. I mean the reason threat was created was to fix other problems, and those problems haven’t gone anywhere. In my mind the way to fix it is by adding other simple mechanics already out there.

    Mechanics that come to mind are from a table top game I used to play, Warhammer 40k. It had a similar problems with heavy armored units being too good. Anyways, they added 2 mechanics to fix this.

    The first was the ‘Swarm Rule’. Basically when a single unit was incredibly outnumbered, they started taking massive amounts of damage that is unavoidable. Before that rule, you could run a single tanky creature into a large mass of troops stalling them indefinitely. After the rule was introduced, it looked more like the offensive line concept. Your tank had to be supported by meaty melee units to keep him from being swarmed and killed.

    The other mechanic that was introduced was the Choppa… excuse me the CHOPPAARH! A very simple and cool mechanic. The idea is that no matter how good your armor is, if you are hit with a 1000 pound weapon or idk slapped by a dragon, it may not be able to cut through your armor but it always hurt. Basically choppa units do the same damage to a tank as they would to a regular melee dps. What they are not is a hard counter to tanks.

    Both of these mechanics were aimed at the same thing, making meaty melee and short ranged units useful in soaking up damage in order to keep your tank from being overwhelmed. It gave rise to tactics similar to Taugrim’s opening post.

    Mechanics like this in an MMO might mean that at 5 adds on your off-tank your melee dps pick up adds 6, 7, & 8 to keep him from being swarmed while range dps kill them. Or having a melee dps off-tank a mini-boss who hits with a choppa is perfectly fine. He takes no more and no less damage than a normal tank would, but soaks up extra damage the dedicated tanks can’t survive, doing much more dps than a tank could to the mini-boss.

    IDK maybe it wouldn’t work, but I’m think threat still has a place as a viable mechanic. I would certainly like to see fights without it, or aggressive AI’s. However it is also a good tool and should remain in use.

  19. Casey M. says:

    The idea is interesting, but unfeasible. If the raid group didn’t have a dedicated tank to take the majority of the damage from a boss, the healers would freaking have conniptions trying to keep everyone alive. It’s tough enough when the boss has AOE, but to have the party’s squishy DPS taking more damage then normal would be paramount to mental self-flagellation for healers in the party. No one would ever want to be party healer unless the developers buffed the healers to the point where to broke the game in all other parts of the game, especially PVP.

    And you say it like EVERYONE doesn’t do anything but pay attention to DPS meters. If I’m playing as a tank in a raid, which I do quite often, my rotation is down to the point where I don’t need to pay attention to a meter to know that my threat is high. And if a DPS DOES pull threat away, I use a taunt. The threat/”holy trinity” system is a very simple mechanic that works for a variety of reasons:

    1) Tanks have a modicum of control over the boss and can use it to help the raid as a whole (e.g. turning the boss away from a group to avoid conal AOE)
    2) Tanks are tanks, DPS are DPS, and healers are healers. Obvious, but tanks are DESIGNED to take damage because of their stat distribution, and the other two groups are not. Roles are specialized for a reason. GW2 does a good job of breaking the mold, but you do run into cases where no one really excels at anything and class differences begin to blur. (e.g. in GW2 I noticed that there’s not a ton of difference between playing a tanky guardian and a tanky warrior)
    3) It makes boss fights predictable to an extent. Many boss fights already have various random factors that players already have to worry about, and worrying about who the boss might one-hit next would not make the instance any more fun, it would make it frustrating.

    The threat/trinity mechanic is still around because it still works. If it didn’t work, games wouldn’t implement it and/or people wouldn’t play them. There’s still plenty of challenge in boss fights because developers always are making up new and challenging boss fight mechanics that people have to overcome.

    • I really think the trinity/threat mechanic is still around because it worked in previous games and publishers don’t like to let their developers take big risks. It worked before, now make me a game that is mildly similar.

      I find games where threat generation is the main focus of the tank leaves me looking at the environment only as a dps. I hardly ever focus on what the boss is doing, because I know I have a snowball’s chance in hell of actually pulling aggro with a competent tank. What’s left for me to do is manage cooldowns and not stand in red circles on the ground.

      Even though I do enjoy PVE content like this, I sometimes wish there were more to it.

  20. Ben says:

    Specialization is a core element of team play. The lack of that element was probably why the implementation of GW2 felt somewhat lacking. So yes, keeping player specialization is important.

    Blocking access to vulnerable group members was how it worked before threat, and still does on tabletop RPGs. The concept is certainly more interesting than threat, but the difficulty of getting that right probably deters game developers for now. Allowing more CC on bosses, introducing CD, and other ways of blocking or controlling the pathing of bosses might increase the danger players abusing the path finding algorithm and avoiding most of the damage, resulting in a trivial boss fight. Interestingly, a system of blocking access to the vulnerable group members might shift the numerical composition of groups, e.g. you might need as many tank melee characters to control the boss movement as you have other classes.

  21. I was wondering what you had been up to.

    I don’t think you really need the trinity or threat. This sort of idea could work in GW2 if they got rid of the defiant mechanic and put in some more interesting mob AI. It is essentially how PVP works, with support, CC, and damage being important. Watch small-scale WvWvW fights or even larger scale GVG fights andyou see this.

    However, defiant on bosses in PVE really takes away the entire concept.

  22. Beetle says:

    Sounds good in theory, but in practice, won’t really work if you ask me. To much that has to be thought over, to much stress for the people that would play the tank class. And if you think of it, it doesn’t sound very different from a threat based system. It is still threat that has to be managed by someone / some player(s).

    GW2 does it different, and is it so much better? Not really. I liked it at first, but after a while, the constant dodging, the constant red circles … it actually made me miss the trinity.

    I like the concept of threat, I do. I remember playing super early vanilla WOW, when threat was an issue for tanks. Back in those days, you knew as a DPS (hopefully) that the tank should get a few sunders in before you started whacking the mob. Threat was something to respect (drama-overkill-alert!). And it was cool. Sometimes things don’t have to be over complicated.

  23. Medicine Man says:

    This is definitely one of my top-3 issues with current gen MMOs (my other two being the similarity of post-leveling content across games and the shift to playing the damage meter/healthbar-whackamole metagames, rather than playing the encounters) and one of the primary reasons that most of them feel like WoW’s long lost cousin. This drive to iterate on the formula established by Blizzard has led to a lost-generation of sorts with regard to MMO developers.

    I tried out Tera and was impressed by how they iterated on bog-standard MMO combat –until– I tried some group based content. The limitations of the “holy trinity”-style class structure have never been so obvious to me, both for the constraints on gameplay and the way it creates suppy/demand issues for pick up groups. Such a glaring flaw.

    Re-examining the class roles and redistributing responsibility across group participants would go a long way to addressing these problems. Thanks for the post, Taugrim.

  24. TL;DR -> Lower the passive threat generation, give the tank more active CCs, and DPS more Mitigation CDs

Comments are closed.

Follow
Taugrim on Twitter Taugrim on Facebook Taugrim on Google Plus Taugrim on YouTube Taugrim on RSS

Enter your email address to receive notifications of new posts by email.

Join 676 other followers

Post Categories
Archived Posts
Copyright © 2009-2014 taugrim.com
All rights reserved
Follow

Get every new post delivered to your Inbox.

Join 676 other followers