Wednesday, August 07, 2013

Stable PvE Strategies

Continuing the discussion on Trinity gameplay, I'd like to take a step back and survey the situation from a higher viewpoint.

So far, in my experience of MMOs, there are three stable strategies for dealing with PvE. The reason I call these "stable", is that all other schemes I've seen are "unstable". These unstable strategies always break down and devolve into one of the stable strategies.

The stable strategies are:

1. Zerg - The enemy's target cannot--or it is not worth the time to--be calculated ahead of time. Whichever player is targeted is reactively healed or protected, or forces the enemy to turn away, or just takes the hits until death.

2. Kiting - The enemy's target cannot--or it is not worth the time to--be calculated ahead of time. Whichever player is targeted proceeds to run away, out of attack range or utilizing line of sight, forcing the enemy to chase after. The chasing enemy does minimal damage, while the other players can attack the enemy. When the enemy switches targets, the new target runs away.

3. Tanking - The enemy's target can be identified ahead of time, and the situation manipulated to force the enemy to attack a specific, hardened target. While the enemy is focused on the hardened target, the other players attack the enemy freely, while the hardened target is healed and protected.

Of these three strategies, Tanking is the most interesting, the structure capable of the most variation. Zerg and Kiting are far less fun if they are the major strategy in a game.

I do not think these are the only stable strategies in existence. But I have not yet seen an MMO that has successfully executed a fourth strategy.

A "rock, paper, scissors" strategy might be stable. The classic "infantry, artillery, and cavalry" is one example. Artillery destroys the slow-moving infantry before the infantry can reach it. Calvary moves too fast for the artillery, but gets broken upon the defenses of the infantry. But "infantry, artillery, and cavalry" generally requires a great difference in movement speed, which I don't think will work in a single-character MMO .

A game that is entirely ranged might have a strategy based on cover and firing lanes. This is typical of shooter games. This strategy is rarely suitable if melee is an important part of the game.

If you want to convince me that a non-Trinity PvE MMO will work, you have to outline a fourth stable strategy. You can't simply assume it exists. Describe it, and show that it will not devolve into Zerg, Kiting, or Tanking.


  1. I had to leave Gw2 due to the zerg mechanic. It just revolved around people dying all the time and rezzing.

    Without trinity you cannot balance gameplay experience for ranged and melee characters. I was literally forced to play ranged even though i always played melee before. Melee classes just took immense amounts of damage.

    Just as you said there may be something else viable but i would want it to cover everyone (both ranged and melee classes). Jumping around to avoid near instant death attacks is a no-no.

  2. One thing that may create an alternative strategy is "collision" which would allow the mobs freedom to be restricted - either by characters or by player created terrain effects.

    It is, after all, a major factor in RL conflicts.

    It does, however, create other non-combat implications for a game.

  3. Thinking about it, I'd say that there's really only two stable PvE strategies, since Kiting is just a subset of Zerging, where kiting is simply another method that the targeted player can use to mitigate the enemy's incoming damage.

    So then the difference really comes from whether or not the enemy's target is determinable or worth determining beforehand. That means the pertinent question to whether getting rid of the trinity is viable is:

    Is there a situation where it is worth determining who the enemy's target is, and yet where the most effective strategy is not having that target specialize somehow with the expectation of being the enemy's target?

    [As an aside, both PvE strategies can currently be seen in WoW, with the first in Scenarios and the 5.2 and 5.3 rares, and the second in any of the more traditional PvE content.]

  4. Most people who view this Rohan’s blog are invariably those who play WOW or a similar variant of it, so maybe the target audience is going to be biased towards the Trinity. If all believed the one way is the only way to go and that never changed, then we would never evolve. I’ve played WOW for seven years and naturally understand the Tank strategy and that is how we have been programmed to understand raids and instances.
    There was a boss in Karazhan (Shade of Aran), may be mistaken here that ignored the Tank mechanism, in fact he was untankable, he randomly hit everyone yet people managed to survive it.
    Purrfect mentions in his comment that there is another strategy called “collision”. Restricting the Bosses freedom, it’s almost a cc type mechanic; you could possibly call it a variant of Kiting. This for me is a viable alternative and maybe one that could be explored. Since most bosses are immune to cc’s, Blizzard is almost forcing the Trinity strategy and not allowing clever use of all the abilities available to you. How cool would it be if you could cyclone a Boss?
    I can’t talk for GW2 as I’m not at a level to do dungeons but their mechanic seems to work for them. Mashing buttons in a tank and spank type scenario is also not fun, yes don’t stand in the bad stuff, that is common sense not a game mechanic.

  5. Rohan, try to imagine Mass Effect-based MMO, something like ME3 multiplayer but with the persistent world and story instead of unconnected MP maps.
    The decision of who to attack is more dependent on tactical opportunity (i.e. if the mob flanked the heavily armored Soldier, it will attack him rather than fire futily at squishy Adept who is in cover) than on threat calculation.

  6. Some of the most interesting small group sessions I've had in an MMO came from early LotRO where we had an LM, a Hunter, and DPS Captain. Each non-trivial fight was diagnosed, CC targets set, kill order established, and everyone went to the fight ready to do what was necessary once the plan hit reality. That was a much more intensive play style and one I'm not sure translates well to current expectations. It's also something that works with a pack of signatures or elites but not really well with a big internet dragon.

    In the end, you are correct. If the expectation is that the "boss" is a huge monster that can destroy any individual player you will need to use one the three tactics you outlined. If, however, you can shift the idea of a good fight to a one with multiple enemies that require coordinated player action to defeat then I think you can add the CC/mutual support play style as a valid fourth option.

  7. Even with single large bosses, use of a "threat table" approach isn't the only one.

    "Tanking" in GW2 is a very different experience than in WoW. Having done both, I'll readily grant that. But the core concept -- protect your friends by preventing your target from attacking them effectively -- is very, very alive. You just don't do it by pushing "threat buttons" -- instead you do things like knock the enemy down, or hamstring him, or interrupt him with a shield bash. You also don't tend to just stand there and get punched in the face, you have to dodge, actively block, and survive.

    It's a different experience. I'm currently doing both flavors (EQ2 for "threat gaming" and GW1 & GW2 for "non-threat").

    Oh... and here is a thought. GW1 is a trinity game that doesn't really use a threat table the way WoW does. Dedicated tanks / healers / DPS, but no threat moves at all. You have to control threat via proximity and body blocking. Pondering that, there might be another dimension to this topic that we haven't hit yet.

  8. I disagree with your use of the term "Zerg" here, if only in the sense that you're trying to lump more styles of play into what is commonly considered a negative.

    I do not believe that combat mechanics whereby individual players have their own tanking and/or healing capabilities, so (most) any member of the group could handle the boss to be inherently a negative. And even with the lack of a "trinity", that doesn't necessarily mean that people can't or won't spec towards certain roles.

    You don't need to have predictable control over the boss to have an enjoyably designed fight. WoW has plenty of memorial boss fights that have literally no threat table, or are otherwise effectively untankable, afterall. And that's not even including that many games have positional attacks; Games like Neverwinter and STO have bosses where almost all of their damage comes from telegraphed or identified attacks, even if you want to describe it as "zerging". The fact that you can actually avoid basically every single attack on a Neverwinter boss just by paying attention and changing your position, even if you're in melee holding threat, allows for a lot of really interesting encounter design, even if the game was designed in what you would call a "Zerg". Neverwinter, though, does have a proper trinity, so theoretically the tank is the one holding threat, but a lot of bosses in the game doesn't really have threat tables as we would otherwise know, mostly because D&D doesn't REALLY.

    To put it simply, I also disagree with the notion that a "true" trinity-less game must have some undefined fourth encounter style. You can have an engaging and fun encounter without defined roles in any one of these methods, if you look at the way you're going to handle combat mechanics first. Zerg and Kiting styles would involve more positional and reactive mechanics (standing in the way of a monster, moving out of the way of attacks), while Tanking simply requires more then the traditional styles of absorbing damage (classes that work around damage absorbing/reducing barriers, classes that work on lifesteals or self-heals, classes that work on avoidance, classes that control minions, etc)

  9. @rimecat: you're absolutely right in what you describe is fights with a different subset of a trinity that increases difficulty. That can be because you're down some people as in your case (not a not specifically to you, rimecat, but more general: note that I said people and not classes, because, while you're also down a proper tank, you would face similar challenges in a zerg scenario). Or it could be because you want extra challenge.

    In any case. Rohan's scenario is a 2/3 trinity scenario, as is yours. Rohan has tank and heals, you have, cc and (ok, maybe only half) heals. Because of course, DPS isn't even part of the trinity, it's just something you bring in addition.

  10. flosch, that group wasn't really down people - we weren't doing the 6 player instances but open world content, 3 player instances, and the like. The main draw was that each character brought multiple capabilities to the fight, even the Hunter who was the closest (at least then, I've no idea now) to pure DPS. It's more a case of each character brings a set of skills and how you combine them is what makes or breaks the encounter. Much more interesting than the Tanking/Healing I've done in Rift and WoW or the zergfest in GW2.

    But it doesn't work against a single, large enemy who can squish you if you lack tank mitigation and incoming heals.

  11. Every time I think about this subject I keep coming back to Skyrim and how combat breaks down when you and your companion walk into a room of bad guys. It seems to work fine in Skyrim, so why is that? Well, it's not a tab target/cooldown rotation RPG and it's got collision detection.

    I'm pretty sure the Elder Scrolls Online is expanding on that combat style and it looks pretty cool from the gameplay video I've seen.

    So, using that gameplay style, if your party walks into a room with a big bad boss creature, who does it attack? Probably the first one it sees, so you better play smart and keep the squshies in back and let that guy in full plate with a shield go in first. It's just sound tactics. Does the boss stick to that first guy for the whole fight? Maybe. Depends on how the fight is designed. I'm sure a good design team can figure out how to make the fights dynamic and interesting.

    So, to answer your question Rohan, IF the game has Elder Scrolls style combat, then I'd call it Controlled Zerg.

  12. @rimecat: Oh, I see. I missed that point in your post. I remember hunters being more or less pure DPS, captains being DPS with either off-tanking or off-healing abilities (probably enough to at least fill the healer role in a small group), and LM being the much-loved (by me) support.

    Come to think of it, I think it's important not to mix classes and roles in this discussion of the trinity. Classes and roles can be the same in some games, but if you take games like WoW or Rift, you still have a role trinity (or something resembling it at least), but people can pick and choose a role in a group.

    In my opinion, that's the difference between filling one role well (either forever, or potentially changing from group to group), and doing everything at once, but half-assed.

  13. Control - Mobs are prevented from being dangerous by being controlled in a coordinated group effort. Your group spreads protective short time buffs on themselves (damage reductions, shields, blocks and reflections, ...) and debuffs on the enemy (blind, weakness, stun, knockdown, disarm, ...). It doesn't matter who the mob attacks because they will no longer be dangerous.

    A bad PUG might either fail or revert to zerg strategy. But then again, a bad dungeon PUG in WoW doesn't really play the trinity correctly but reverts to zerg too.

    It depends on the difficulty of the game if reverting to zerg is possible. And if it is, PUGs will often revert to zerg.

  14. Sorry for asking, but I wonder why is this:
    "If you want to convince me that a non-Trinity PvE MMO will work, you have to outline a fourth stable strategy."

    Is there any reason I'm missing why a non-trinity model which, for example, follows your "tanking" model is not acceptable.

    It's also possible for a players to use different strategies (including but not limited the ones you listed) depending on situation and the player attacked.