
![]() |
1 person marked this as a favorite. |

That is something to worry about. In many games, Pen and Paper and Electronic both, there are enemies and encounters designed specifically to make life difficult at best, or outright cock-block at worst, certain Heroes and their abilities.
I think, however, with a Skill system in place, it will be less a chore for players to deal with. This Hex is home to Undead who are immune to your Charm and Cold spells? Skill up with a friend to learn Fire and Turn Undead spells and skill up new skills to compensate.
That is one of the things I am really looking forwards to with the Skill System of Pathfinder Online, in that people will build what is fun to play, and their skills will increase depending upon what they do (I believe), hence you go from decapitating Goblins to blasting fireballs into the Zombies to curing a town of diseased NPCs and PCs to sneaking past Hobgoblins ..... as you do all these things, succeed or fail, your skills will grow and change depending upon your actions, how you deal with these enemies and these situations has the potential to grow truly Organic Characters.
Or you can just go "WAAAAAAAAAAAAAAAAAAAAAAARGH!" and just bulldoze your way to your envisioned build with a dedicated group of friends.
Either way is great fun, or at least I think they can be.

![]() |

Some characters will be able to absorb more damage, but deal out less (or be much more situational about the damage or effects they deal); Call those characters fighters. Other characters will be able to deal more damage, but absorb less. Call those characters mages. Some characters will be able to deal out quite a bit of damage, and also absorb quite a bit. Call those clerics.
While being engaged by a group including fighters, clerics, and mages, why would an intelligent opponent attack the fighters, if given free choice in who to attack? The fighters' job is to give the enemies no choice but to attack them, first. It's the mages' job to kill the enemies before the fighters die.
One way to do this is the give the fighters an "attack me" button. A better way would be to implement mechanical penalties to the enemy for certain tactics (provoking AoO for moving, for example) and implementing a robust AI which prioritizes targets and judges the cost/benefit expectation for various tactics. One of those has been exemplified in a finite game; the other one is exemplified in the tabletop game, except with the GM instead of an AI.

Hudax |

Some characters will be able to absorb more damage, but deal out less (or be much more situational about the damage or effects they deal); Call those characters fighters. Other characters will be able to deal more damage, but absorb less. Call those characters mages. Some characters will be able to deal out quite a bit of damage, and also absorb quite a bit. Call those clerics.
While being engaged by a group including fighters, clerics, and mages, why would an intelligent opponent attack the fighters, if given free choice in who to attack? The fighters' job is to give the enemies no choice but to attack them, first. It's the mages' job to kill the enemies before the fighters die.
One way to do this is the give the fighters an "attack me" button. A better way would be to implement mechanical penalties to the enemy for certain tactics (provoking AoO for moving, for example) and implementing a robust AI which prioritizes targets and judges the cost/benefit expectation for various tactics. One of those has been exemplified in a finite game; the other one is exemplified in the tabletop game, except with the GM instead of an AI.
Your mechanic with an intelligent GM running the show:
1) The mob moves away from the Fighter.
2) The Fighter gets an AoO and hits the mob.
3) The mob moves away from the Fighter faster because it's getting stabbed in the back.
I take it you would prefer:
3) The mob turns back to the Fighter, angered by the AoO.
All your suggestion does is re-skin threat mechanics without really changing anything. Giving the mob "no choice but to attack the Fighter" is exactly what I don't want to see anymore: static fights and static roles.

![]() |

Daniel Powell 318 wrote:Some characters will be able to absorb more damage, but deal out less (or be much more situational about the damage or effects they deal); Call those characters fighters. Other characters will be able to deal more damage, but absorb less. Call those characters mages. Some characters will be able to deal out quite a bit of damage, and also absorb quite a bit. Call those clerics.
While being engaged by a group including fighters, clerics, and mages, why would an intelligent opponent attack the fighters, if given free choice in who to attack? The fighters' job is to give the enemies no choice but to attack them, first. It's the mages' job to kill the enemies before the fighters die.
One way to do this is the give the fighters an "attack me" button. A better way would be to implement mechanical penalties to the enemy for certain tactics (provoking AoO for moving, for example) and implementing a robust AI which prioritizes targets and judges the cost/benefit expectation for various tactics. One of those has been exemplified in a finite game; the other one is exemplified in the tabletop game, except with the GM instead of an AI.
Your mechanic with an intelligent GM running the show:
1) The mob moves away from the Fighter.
2) The Fighter gets an AoO and hits the mob.
3) The mob moves away from the Fighter faster because it's getting stabbed in the back.I take it you would prefer:
3) The mob turns back to the Fighter, angered by the AoO.
All your suggestion does is re-skin threat mechanics without really changing anything. Giving the mob "no choice but to attack the Fighter" is exactly what I don't want to see anymore: static fights and static roles.
Seconded, the fact is the tank cannot be a perfect tank. Now there can be mechanics, in P&P it was not always impractical for the tank to set himself up to discourage people from attacking others. Most enemies for instance would not often consider it worth the AoO to go after the wizard, and fighters had other tricks up their sleeves say reach + combat reflexes etc... so that when the enemy does turn it's back on the fighter, he will be badly hurt, or even trips/grapples etc... All enemies acting the way you want them to is the system we've grown accustomed to, but not necessarally the best system. It is also why pvp/pve builds are usually so drastically different in most theme park games. Your pure defense tank just isn't so good when your opponent realizes, oh he's a defense tank, blitz past him and chop off the pure healing priests head first.

![]() |

Daniel Powell 318 wrote:Some characters will be able to absorb more damage, but deal out less (or be much more situational about the damage or effects they deal); Call those characters fighters. Other characters will be able to deal more damage, but absorb less. Call those characters mages. Some characters will be able to deal out quite a bit of damage, and also absorb quite a bit. Call those clerics.
While being engaged by a group including fighters, clerics, and mages, why would an intelligent opponent attack the fighters, if given free choice in who to attack? The fighters' job is to give the enemies no choice but to attack them, first. It's the mages' job to kill the enemies before the fighters die.
One way to do this is the give the fighters an "attack me" button. A better way would be to implement mechanical penalties to the enemy for certain tactics (provoking AoO for moving, for example) and implementing a robust AI which prioritizes targets and judges the cost/benefit expectation for various tactics. One of those has been exemplified in a finite game; the other one is exemplified in the tabletop game, except with the GM instead of an AI.
Your mechanic with an intelligent GM running the show:
1) The mob moves away from the Fighter.
2) The Fighter gets an AoO and hits the mob.
3) The mob moves away from the Fighter faster because it's getting stabbed in the back.I take it you would prefer:
3) The mob turns back to the Fighter, angered by the AoO.
All your suggestion does is re-skin threat mechanics without really changing anything. Giving the mob "no choice but to attack the Fighter" is exactly what I don't want to see anymore: static fights and static roles.
Nope- I want "The mob decides if the AoO is worth the reposition, and either takes it or maneuvers differently." The best tank would have very powerful AoOs, which were never used because nobody would provoke them.
IF the PvE mechanics worked the same as the PvP mechanics, but with some mechanism for denying or increasing the cost of enemy maneuvering, then the only thing in the way of this is AI and the cycles to process it. The only benefit to the 'hate' mechanic is that it can be done, and the only downside to a good AI is that it is prohibitively expensive.

![]() |

Some characters will be able to absorb more damage, but deal out less (or be much more situational about the damage or effects they deal); Call those characters fighters. Other characters will be able to deal more damage, but absorb less. Call those characters mages. Some characters will be able to deal out quite a bit of damage, and also absorb quite a bit. Call those clerics.
While being engaged by a group including fighters, clerics, and mages, why would an intelligent opponent attack the fighters, if given free choice in who to attack? The fighters' job is to give the enemies no choice but to attack them, first. It's the mages' job to kill the enemies before the fighters die.
One way to do this is the give the fighters an "attack me" button. A better way would be to implement mechanical penalties to the enemy for certain tactics (provoking AoO for moving, for example) and implementing a robust AI which prioritizes targets and judges the cost/benefit expectation for various tactics. One of those has been exemplified in a finite game; the other one is exemplified in the tabletop game, except with the GM instead of an AI.
Actualy there is another more basic and simple reason why opponents must deal with the fighters before attacking the mages in PnP games....it's called Physics. A solid object may not pass through another solid object.
Attack of Opportunity or no...if you have a high priority target but can't get to it, you are not going to be attacking it. There is no reason why todays computer games can't impliment such a system... the first computer games that came out didn't because thier physics engines couldn't handle (for purely technical reasons) collision detection between moving objects...hence the cheap hack of "aggro" was born.

Hudax |

Nope- I want "The mob decides if the AoO is worth the reposition, and either takes it or maneuvers differently." The best tank would have very powerful AoOs, which were never used because nobody would provoke them.
Now you have one of two scenarios:
1) The Fighter has an awesome ability that he rarely gets to use, and mob behavior is made predictable, or
2) Basic group strategy is to pull aggro off the tank (however that can be done) as often as possible so he can use his awesome AoO.
IF the PvE mechanics worked the same as the PvP mechanics, but with some mechanism for denying or increasing the cost of enemy maneuvering, then the only thing in the way of this is AI and the cycles to process it. The only benefit to the 'hate' mechanic is that it can be done, and the only downside to a good AI is that it is prohibitively expensive.
But enemy maneuverability is part of the solution, not part of the problem. A fight where the mob's movements are predictable or controllable is a static, scripted fight. That's what I don't want. I want the enemy to behave unpredictably and uncontrollably, so fights are no longer static or scripted.
You're still thinking in the context of tank death/aggro loss equals wipe. I don't think that should be the case.

![]() |
1 person marked this as a favorite. |

But enemy maneuverability is part of the solution, not part of the problem. A fight where the mob's movements are predictable or controllable is a static, scripted fight. That's what I don't want. I want the enemy to behave unpredictably and uncontrollably, so fights are no longer static or scripted.You're still thinking in the context of tank death/aggro loss equals wipe. I don't think that should be the case.
Your point is valid, I believe the biggest issue of the modern MMO is that tanks are too good at tanking. tanks are scaled up to the point where they have 10-20x the defense/HP as others, the mobs are scaled up in damage to match, ensuring that anyone who isn't a tank is guaranteed to die. Don't get me wrong just like in P&P tanks need good defense, but it should be 20-50% better, not 400%+

![]() |

Daniel Powell 318 wrote:
Quote:IF the PvE mechanics worked the same as the PvP mechanics, but with some mechanism for denying or increasing the cost of enemy maneuvering, then the only thing in the way of this is AI and the cycles to process it. The only benefit to the 'hate' mechanic is that it can be done, and the only downside to a good AI is that it is prohibitively expensive.Enemy maneuverability is part of the solution, not part of the problem. A fight where the mob's movements are predictable or controllable is a static, scripted fight. That's what I don't want. I want the enemy to behave unpredictably and uncontrollably, so fights are no longer static or scripted.
You're still thinking in the context of tank death/aggro loss equals wipe. I don't think that should be the case.
In the scenario I presented, the only way to get the enemy to provoke that awesome AoO is to make that the best choice for it- "I take a hard hit, but I eat the mage who casts the fireballs." The fighter is using the deterrent abilities ALL THE TIME, especially when they aren't activated. If a rational and fully-informed party decides that presenting a piece of bait to provoke that AoO is worthwhile, a rational and fully-informed enemy will decide that taking the bait is not worthwhile (assuming that every outcome which is better for one opposed group is worse for the other).
The enemy can maneuver by default, and will target the squishiest opponent first by default. The hardened players job is to keep the squishy players alive for long enough to kill the enemy.
The enemy behavior is either non-deterministic or predictable. There is literally no middle ground. I doubt that non-deterministic behavior is appropriate for anything other than an embodiment of chaos. For PvP, enemy behavior is presumed to be predictable-they will always do what is most likely to cause them to do better. They will most want to attack the members of your side that you least want them to attack, and they will act to make it difficult or costly to attack the members of their side that you most want to attack.
Oh, and a tripping, bull-rushing, grappling fighter might be the ideal way to model the tank; he makes it difficult or impossible for the enemy to get to the people behind him, by reducing their ability to maneuver. An illusionist can do the same thing, by causing the enemy to engage things that aren't there. The role of tank is filled by anybody that increases the time that it takes the enemy to incapacitate somebody.

Hudax |

In the scenario I presented, the only way to get the enemy to provoke that awesome AoO is to make that the best choice for it- "I take a hard hit, but I eat the mage who casts the fireballs." The fighter is using the deterrent abilities ALL THE TIME, especially when they aren't activated. If a rational and fully-informed party decides that presenting a piece of bait to provoke that AoO is worthwhile, a rational and fully-informed enemy will decide that taking the bait is not worthwhile (assuming that every outcome which is better for one opposed group is worse for the other).
Now you have a cold war of unused tank and boss options. What is the point of this? And who says bosses ever perform the "best" action? As Ryan stated upthread, it is trivially easy to make the AI unbeatable. Currently bosses only do what the hate list and script tell them to do. Those are NOT optimum choices--quite the opposite, in fact. They are precisely the choices that get them killed.
The enemy can maneuver by default, and will target the squishiest opponent first by default. The hardened players job is to keep the squishy players alive for long enough to kill the enemy.
This doesn't have to be the case. The "squishy" players can be developed with survivability tools. The game in general can be made so that the squishies don't fold like wet kleenex when the boss glances at them sideways. "Squishiness" doesn't have to exist.
The enemy behavior is either non-deterministic or predictable. There is literally no middle ground. I doubt that non-deterministic behavior is appropriate for anything other than an embodiment of chaos.
Read upthread to see various suggestions on how to create that middle ground.
Even if you were right, I would still opt for chaos. I've played the orderly game.
Oh, and a tripping, bull-rushing, grappling fighter might be the ideal way to model the tank; he makes it difficult or impossible for the enemy to get to the people behind him, by reducing their ability to maneuver. An illusionist can do the same thing, by causing the enemy to engage things that aren't there. The role of tank is filled by anybody that increases the time that it takes the enemy to incapacitate somebody.
One thing we agree on--tanking is nothing more than glorified crowd control. I want it to become more like regular crowd control. I want the concept of "main tank" to disappear. I want the tin can and the illusionist to be equally adept at controlling the boss when it's necessary for them to do so.

![]() |

Daniel Powell 318 wrote:In the scenario I presented, the only way to get the enemy to provoke that awesome AoO is to make that the best choice for it- "I take a hard hit, but I eat the mage who casts the fireballs." The fighter is using the deterrent abilities ALL THE TIME, especially when they aren't activated. If a rational and fully-informed party decides that presenting a piece of bait to provoke that AoO is worthwhile, a rational and fully-informed enemy will decide that taking the bait is not worthwhile (assuming that every outcome which is better for one opposed group is worse for the other).Now you have a cold war of unused tank and boss options. What is the point of this? And who says bosses ever perform the "best" action? As Ryan stated upthread, it is trivially easy to make the AI unbeatable. Currently bosses only do what the hate list and script tell them to do. Those are NOT optimum choices--quite the opposite, in fact. They are precisely the choices that get them killed.
Quote:The enemy can maneuver by default, and will target the squishiest opponent first by default. The hardened players job is to keep the squishy players alive for long enough to kill the enemy.This doesn't have to be the case. The "squishy" players can be developed with survivability tools. The game in general can be made so that the squishies don't fold like wet kleenex when the boss glances at them sideways. "Squishiness" doesn't have to exist.
The boss AI already determines what course of action the boss decides is best. I want that AI to be informed by actual consequences rather than abilities which directly modify it (taunt adds hate).
Either all players will be equally squishy, or some will be less squishy than others-for a given opponent. If the opponent is targeting will saves, or touch AC, then the dynamic changes. In general, anything that can be called a boss should be able to fold at least one type of character easily.
Crowd control is a different role than tanking- CC is the task of taking a large number of weaker creatures and making them less of a threat (either by killing/defeating them, or disabling/distracting them) Tanking is the task of taking a single creature and making it less of a threat to anybody else (by making it focus on the tank). Wizards make the best crowd control, starting with color spray and sleep.

Hudax |

The boss AI already determines what course of action the boss decides is best. I want that AI to be informed by actual consequences rather than abilities which directly modify it (taunt adds hate).
As long as this doesn't lead to the boss making the same decision every time, or allow players a constant ability to manipulate the boss, I'm ok with this. What I mean is, make the boss "irrational" by having it sometimes make random decisions about who to target and what ability to use.
Either all players will be equally squishy, or some will be less squishy than others-for a given opponent. If the opponent is targeting will saves, or touch AC, then the dynamic changes. In general, anything that can be called a boss should be able to fold at least one type of character easily.
I'm fine with rock-paper-scissors as long as it doesn't mean instant death. More difficult to survive, sure. But having bosses be able to one-shot people would stop this whole thread in its tracks.
Crowd control is a different role than tanking- CC is the task of taking a large number of weaker creatures and making them less of a threat (either by killing/defeating them, or disabling/distracting them) Tanking is the task of taking a single creature and making it less of a threat to anybody else (by making it focus on the tank). Wizards make the best crowd control, starting with color spray and sleep.
I meant in the general sense. Tanking the boss and kiting the boss, for instance, are only different in execution. The end result is the same.
In EQ, we had a tank die on Avatar of War once, and one of our rangers kited him for several minutes until he made a mistake and died. The only thing that changed for the rest of the raid was we ran around in a circle instead of standing in the corner. (And the healers got to dps.) That is the sort of change-up I want to see multiple times in a fight (minus the requirement for tank death to make it happen, and minus the unforgiving nature of one person's mistake wiping the raid.)
To that end, all forms of crowd control would be equal. Some players would tank in the traditional sense--arms and armor. Some would summon illusions to occupy the boss. Some would have a pet to protect them. Some would use slow, fear or root mechanics. Some would debuff the boss to hit like a kitten. Some would buff themselves into tankability. All done on cooldowns (if the boss turns on you again before you're ready--now you need significant help), and all at the boss's whim rather than planned swaps.

![]() |

I meant in the general sense. Tanking the boss and kiting the boss, for instance, are only different in execution. The end result is the same.
Kiting the boss IS tanking the boss, as is rooting him or making him fight an illusion. Getting the sole attention of a group of adds and soaking their damage with hit points IS crowd control, as is confusing them or blowing them to bits with a fireball. Turning undead can be crowd control and tanking at the same time, or sometimes a win-button, depending on the quality of the cleric and the opponent.
I agree that the role of a character should be flexible, and some characters should be able to fill more than one role at once, while other characters should be able to fill one role at a time, but better.
If, however, you need to kite an enemy, then getting slowed or rooted will result in your death. That means anyone who slows or roots is basically immune to that form of kiting, just like anything with true seeing is immune to illusion and most things with touch attacks are basically immune to 'plate mail and tower shield' tanking.

Hudax |

Kiting the boss IS tanking the boss, as is rooting him or making him fight an illusion. Getting the sole attention of a group of adds and soaking their damage with hit points IS crowd control, as is confusing them or blowing them to bits with a fireball. Turning undead can be crowd control and tanking at the same time, or sometimes a win-button, depending on the quality of the cleric and the opponent.
Exactly what I'm trying to say. All of those things could be called tanking or crowd control interchangeably in my mind.
My thoughts in general could be summed up as "crowd control on steroids without threat control."
If, however, you need to kite an enemy, then getting slowed or rooted will result in your death. That means anyone who slows or roots is basically immune to that form of kiting, just like anything with true seeing is immune to illusion and most things with touch attacks are basically immune to 'plate mail and tower shield' tanking.
That's part of the rock-paper-scissors.
I would advocate that everyone has at least two different methods of CC. Otherwise raids will exclude builds that are vulnerable to the boss. And I think bosses should have limited ways to shut down CC. Most raid bosses shut down all kinds of CC except tanking. I think they should only be able to negate one or two types.

Stormanne |

An idea to avoid some of the typical in combat AI limitation would be the introduction of an active environmental awareness AI. Whereas in current MMO's, NPC activities are limited to scripted activities that are tethered to a specific area of patrol. Remove the tethers and give a basic set of command protocols for the NPC to work within. We've all seen the monsters in other games standing around waiting for the PC to encounter them. Creation of an AI that is attempting to accomplish something and is in a realistic manner trying to pay attention to their environment would eliminate a lot of the ability for PCs to create patterned responses and strategies and keep encounters a lot more organic.
At the very least, a field of vision AI that doesn't rely of an arbitrary boundary being crossed before an NPC reacts. This would open up the possibility of realistic ambush scenarios and true to life attacks of opportunity. The FoV AI, when combined with a non-tethered operational theater, would also eliminate "skirting" (sticking to the outside edges of NPC operational boundaries in order to circumvent encounters or exploit limitations of AI)by PCs. An FoV system would also allow for more varied responses by the NPCs. Depending on the situational awareness of a particular group of NPCs, a mix of ranged to melee strategies could be engaged to take out perceived threats at a distance and bypass the trappings of typical "Holy Trinity" encounters.
I'm also well aware that all these are fantastic and grand ideas that don't amount to a hill of beans without someone with the know how to pound out that type of functioning code.

![]() |

Daniel Powell 318 wrote:Kiting the boss IS tanking the boss, as is rooting him or making him fight an illusion. Getting the sole attention of a group of adds and soaking their damage with hit points IS crowd control, as is confusing them or blowing them to bits with a fireball. Turning undead can be crowd control and tanking at the same time, or sometimes a win-button, depending on the quality of the cleric and the opponent.Exactly what I'm trying to say. All of those things could be called tanking or crowd control interchangeably in my mind.
My thoughts in general could be summed up as "crowd control on steroids without threat control."
Quote:If, however, you need to kite an enemy, then getting slowed or rooted will result in your death. That means anyone who slows or roots is basically immune to that form of kiting, just like anything with true seeing is immune to illusion and most things with touch attacks are basically immune to 'plate mail and tower shield' tanking.That's part of the rock-paper-scissors.
I would advocate that everyone has at least two different methods of CC. Otherwise raids will exclude builds that are vulnerable to the boss. And I think bosses should have limited ways to shut down CC. Most raid bosses shut down all kinds of CC except tanking. I think they should only be able to negate one or two types.
How about, everyone can fill at least three roles out of: Tank, controller, DPS, healer, facilitator, debilitator? We also need to stop thinking about THE raid boss, instead of the finale ENCOUNTER. A lich is one thing, but when he's backed up by a pack of mummies, some yaun-ti retainers, and an adamantine golem everything changes.
The sorcerers try to lock down the lich's spellcasting with counterspells (improved counterspell, dispel magic). The rogues slip into the yuan-ti, assisted by the wizards. The tin cans delay the golem while the clerics incinerate the mummies. The mummies need to get taken down fast, so that the clerics can tend to the wounded and replace protection spells, then start on the lich. Once the yuan-ti are thinned out and badly wounded, the wizards start assisting the fighters while the rogues take the stragglers. Once either of the heavies drops, everyone goes for the dogpile.
There's a lot going on there- tanking by counterspell, crowd control through Greater Invisibility, crowd control through turn undead, and damage through "my allies in melee with you have improved evasion".
Also, no plan survives contact with the enemy: the clerics might not turn the mummies off of the sorcerers, giving the lich a chance to hit the rogues and fighter with a Mass Hold Person. After a few coups de grace, the yuan-ti push through the blade barriers and apply more pressure on the survivors. The arcane casters dimension door back to the other side of the barrier, but feel hopeless about saving the clerics. They panic, and use an old trick: they drop walls of force to block in the enemy and incendiary clouds to burn them out. Burning the wishes intended to finish the adamantine golem, they manage to kill everyone else, only to be pounded into jelly by the unstoppable object.

![]() |

Games that offer the possibilities that each character may fulfill two roles have not managed to cut the shortage of tanks and healers. This is because you still need the appropiate gear but mainly because tanking and healing is much more demanding than nuking.
Also even if you do not hardcode the roles into your encounters, there will be chars that are much more resilient than other chars and, to balance this, should do less damage. So even if you do not use these roles in your encounters they are still there and the WILL BE a best setup to beat each and every encounter or else these encounters are simply not challenging at all and this setup will likely determine which chars will be wanted and which aren't.

![]() |
1 person marked this as a favorite. |

Rather than have a cut and dry system, I say let the skill based stuff go, and let the player community figure out which builds work best for what role, which skill combinations net the best DPS, Resiliency, healing, etc. Let us experiment with what works, don't design or hardcode it for us. It's skill based, let's see what works and fits.

![]() |

Rather than have a cut and dry system, I say let the skill based stuff go, and let the player community figure out which builds work best for what role, which skill combinations net the best DPS, Resiliency, healing, etc. Let us experiment with what works, don't design or hardcode it for us. It's skill based, let's see what works and fits.
Yup, I think the other major thing to note, PVE raids are not likely to be a large factor of the game. There may be a handful of instances, but I highly doubt they are going to be a large factor in how the game runs. Roles, tanking etc.. will be pretty moot points, because the invading army has it's own tactics that do not rely on a hate system.

![]() |

This is the best and most constructive thread I have read on this whole site for months.
I too am irked by the rock-paper-scissors aspect of most MMOs. A cleric should be able to have just as good of defense as a fighter.
Also, it isn't about making the players more intelligent, it is about making the monsters more intelligent. Opponents should have different attack priorities based on their type/intellect.
Even out the playing field a bit and reduce the scripting. A game can be just as chellenging without forcing players into a narrow set of roles.
[Edit: Now we just have to wait for the haters to show up.]

![]() |

Based on a number of the posts in this thread, this is what's knocking around in my head:
1) Characters have a number of (invisible) stats based on what they appear to be. Examples might be "armored", "spellcaster", "melee damage", "ranged damage". Visibly carrying a bow would give points to ranged damage, visibly wearing robes would boost spellcaster, carrying a large weapon would improve the melee damage appearance, and so on. Basically, anything that changes what you look like (not what you actually do). Things like gloves of storing, or having the mage carry a bow they never use would affect this, so you can "fool" things... but only to a point. Putting the wizard in fullplate might fool things, but it'd also make him very ineffective (unless he had built the character for casting in fullplate...). (For example purposes, I'll assume it's a scale of 0-10)
2) Monsters have a "target equation" which factors in each of the above things to figure out how much of a target a particular character is. For example, dire bears might prioritize things with low "armored" scores, not weigh spellcaster or ranged damage at all, and want to avoid the melee damage characters. (A*(10-3) + S*0 - M*1 + R*0 = Target). Meanwhile, a highly intelligent dragon might prioritize spellcasters, then ranged damage, then melee, then armor.
3) Based on what the characters do, their "appearance" numbers would change, which then changes their targeting priority - pulling out a sword would increase the melee damage stat, while casting a spell would add to spellcaster based on how strong it was. Shooting and missing, or shooting and not getting through DR would lower ranged, while failing an attack would increase armor (which leads to "Oh, that guy's not wearing any armor - he should be easy to hit... except he's dodging every attack. I'll go gnaw on the tin can instead. He's easier to hit.")
4) Based on the current targeting priority, the monster will pick a preferred target. All else being equal, it will attack the preferred target exclusively. However, all else is rarely equal. There would be additional factors like "closeness", "visible wounds", "turns spent chasing", "damage done", "damage healed", and so on (all the usual aspects of Hate) which would be combined with the priority to pick an actual target and a tactic to use.
To see this in action, I'll use the dire bear: It wants to eat unarmed things, is wary of melee, likes to attack things that are already injured or are close, and really wants to kill anything hurting it.
Also assuming a standard adventuring party of fighter, rogue, wizard, cleric. So the bear starts out wanting to attack the party's wizard (least armored) and avoid the fighter, and to a lesser extent the cleric. The rogue doesn't have any visible weapons, but is wearing a chain shirt, so is in the middle. That makes the starting priority: Wizard, rogue, cleric, fighter.
Based on having a clear shot to the wizard, the bear charges at him and attacks. He hits, does damage to the wizard, and the party starts reacting. The fighter moves up and attacks. The rogue tumbles around to flank. The cleric heals. The wizard moves away. The fighter's closeness and damage done works to balance his having a melee weapon and armoring. The rogue is still not a threat (although to a smarter creature it might be). The wizard is out of reach, and the cleric isn't relevant. In a purely Hate based system, it'd either switch to the fighter (damage done), or go after the wizard (initial aggro). In this case, it might do either, depending on the relative values, or it might turn on the rogue because it's unarmored enough and close enough to be a target, and the fighter wasn't a big enough threat to distract it.
Then the rogue pulls out some knives and does a sneak attack for lots of damage. Now the rogue is a big threat, and it might even choose to follow the rogue if the rogue tries to run away, with the fighter running behind.
...
A more complicated creature like a shadow knows that many things can't hurt it and if it kills something, it gets an ally, so it might put a huge priority on attacking injured creatures. That could lead to it choosing to ignore all the damage it's taking from the fighter and wizard to go touch the unconscious rogue, then will turn on the best target left. On the other hand, it fears clerics, so as soon as the cleric casts a positive energy spell (and gains +1 "positive energy" factor), it will try to avoid the cleric. Whether that means abandoning the plan to touch the rogue would depend on how much it fears positive energy.
In the end, it does boil down to just a more complicated version of a Hate system. But you have to have an algorithm if you're going to program it, so it's impossible to totally do without one. This system has enough factors that the correlation between Hate and what's going on is complex enough that it won't feel like a Hate system. You can "tank" by making yourself the biggest threat, but only if the creature cares about what the biggest threat is. Each monster will "learn" as the combat goes on what the characters can do, and will behave slightly differently in response to that. This would work just fine in a classless system, because it's entirely based on what the character looks like and does, not what they are. If a character never heals, they won't "pull aggro" from the creature that targets healers above all others (even if that's detrimental to its own health). Conversely, an ooze couldn't care less whether there's a whole army of healers standing there while there's something hitting it with a stick.
.....
I think I've rambled enough. It's basically a codification of what I use for running monsters in my own games. "Who looks squishy? Who is the biggest threat to me right now? Can I ignore this guy to go attack that other guy I'd prefer to be attacking?"

![]() |

That's good stuff. It would have to be tweaked to account for non-visual senses (how would blind-sense work?), but all-in-all this is a type of system that would make the encounters work differently than the same-old MMO fight scripting.
I also like the idea of making the enemies do something other than stand around, waiting for someone to gank them. How about critters that use scent to actively seek out new foes? How about ones that wait to attack until you're wounded or already engaged? What happens if you want to kill something that doesn't want to fight (kobolds in trapped caverns for instance?)
Keep it rolling. I hope the future developers take this and run with it.

![]() |

Based on a number of the posts in this thread, this is what's knocking around in my head:
Well all of that in mind, it would be awsome, but now we are talking a more sophisticated AI that has ever been considered in an MMO, that is not primaraly about combat with NPCs. AI programing is not cheap or easy, and while it would be awsome, I strongly can't imagine it being done in an MMO with a small budget, and an emphasis on PVP, rather then PVE. That's along the lines of wishing your house had a better security system than fort knox. It sure would be great, but it probably should be filed under not going to happen.
Simpler equasions are possible, i would believe that an AI could note the armor etc... possibly base things on damage delt by, and healing to current target done by, as the current AI's that I think should be packaged with the middleware do, but expecting extreme depth, and multiple categories of AI's to do, then we aren't looking at a small tweak, we are looking at a PVE investment on a WoW budget scale for a game that isn't intending PVE to be it's selling point.

Hudax |

I'm already here...carry on.
Lol :)
@MicMan: What I'm suggesting is that hard and fast roles be done away with entirely, not merely that people be able to fill multiple roles. Because you're exactly right--tanks and healers do have more responsibility--and that's part of the problem. What I'm suggesting is to spread those roles out over everyone, so that everyone has an approximately equal level of responsibility regarding tanking (and other equally effective methods of CC) and healing (and other equally effective methods of mitigation). An individual's role at the moment would change randomly based on circumstances outside their control (namely, when the AI decides to attack them, and when someone is in dire need of aid), and over a long period of encounters they would find themselves performing each role to about the same degree as everyone else.
All CC would be equal in theory, but situationally more or less useful. The ranger's slow spell would be equally capable of controlling the enemy as the fighter is of tanking the enemy. But the enemy's root spell would force the ranger to use a different tactic (maybe the ranger has a weapon shield ability or a short, powerful mitigation buff).
Further, the various mitigation tools would also be equal to each other, and situational. The ranger's buff might make it unnecessary to heal the fighter for a short while. The mage's false life might give someone else the time they need to get a heal or use a potion. The fighter's AC/HP might be great against physical attacks but useless against magical attacks. The mage might take serious physical damage but have great resilience against magic.
Resilience is relative. In other games, fights are tuned to the tank, making the tank's resilience paramount and everyone else's woefully lacking. Hence they get dps and healing to compensate. It doesn't have to be like that.
Group setup is also relative. Provided fights have enough diversity (whether via adds or the boss's own abilities), anyone could earn a spot in the group. I don't think it's difficult to avoid making archetypes that are one-hit wonders and making encounters that render their one good ability useless. The key would be giving everyone equal opportunity in every role (but in different ways).

Hudax |

@Bobson: Very nice. Very complex, but I like the general idea a lot.
@Onishi: As Ryan said, it's easy to make the AI too good. To extrapolate from that--it would be easy to make the "scissors" enemies always and exclusively attack the "paper" players with precisely the best attack until they are all dead, etc.
This, or something like it, could be much easier to do than we think.

![]() |

CCP has actually started doing that in EVE of late. They started with the "Sleeper" special Mobs and have been slowly extending that more complex Hate algorithm to other AI. It made Sleeper fights much harder and a bit more like dealing with a moderately intelligent player. While they would agro the first thing to show up, they would switch to more attractive targets like healers (or my Command Ship) if given the option.
Personally, as complex as Hate would be harder to program as a PvEer I'd enjoy it. It also would help make an idea I had about put non-PvP seeking people into the category of Mobs a bit more reasonable.
====
@Ryan's comment about people playing FPSs. I'm sorry but most FPSs have absolutely no relevance to the MMO PvP. They are two totally differ horse of two totally different colors... in fact one is camel. FPS matches are quick (rarely more then 30 minutes) sides are often scrambled at then end to distribute "better/weaker" players to help balance play, and you can always quite and join a different server/match. There is no true "loss" of gear/stats/etc.
None of this is applicable to an open-PvP single server world. PvP is often stacked by guilds/corps, death and loss is not a momentary issue, loss of gear can be a major blow (people totally quit EVE when they lose ships like Titans and such), and such "better players" are persistent. Most open-PvP I've seen is more kin to FPS Competitive play, which only a small, but very skilled, minority participate in.
How is a "human" level PvE fight different then real open-PvP? 1) Unless the developers really are out to kill the players you will not have absolutely overwhelming reinforcements join the fight mid way (which happens in open PvP). 2) Ultimatly if needed you can "escape" a bad fight, where real PvP will chase you until they are forces by circumstance to stop. 3) AIs don't hold grudges and plot complex revenge ambushes or retaliatory strikes. You can get away with ditching 1 (that 10 times per whatever guideline in the industry you mentioned), you can even ditch 2 to an extent, but in the end it's #3 that is important. Win or lose you know the AI is going keep hounding you for a month or more and "camp" the town exit.
I don't hate the AI driven Bot in an FPS, although I sometimes question if it's Aiming algorithm is a bit too perfect to be a credibly "human" challenge. There is "human" even "pro human" and then there is "100% accurate head shots".

Uchawi |

I dont care if you are playing RPGs or online games, it will come down to understanding each class and defining roles. Whether it is a close proximity to WoW, or other games, is up to the developer. Or if they keep the roles loosely defined, then the players will do their best to create their own via specific class builds, and most likely bring in the trinity.
In addition, you have protected content (instances) versus open. I actually prefer the old days of EQ when there was competition for resources as long as you had enough people mustered at the appropriate time, but they also had instances for certain events. This takes the elements of PvP into PvE, but also guarantees content will continue to the be developed. The focus of PvP should be to balance the classes from the ground up in cosideration of fighting other players and keeping a sense of balance. But I want a game the focuses on content, versus having a limited world, and expect PvP to keep people interested, as you cater to one audience.

Hudax |

@Dorje: I believe Ryan was just using FPS as an example of how you could make AI too good. It's more obvious in an FPS--just give it perfect headshots and perfect dodging. It will always kill you in one hit and you will never, ever hit it. In an MMO it is slightly more complex. Just give the boss enough HP and an instant complete heal that never fails to go off at say 50%. Sure it's cheating but that's the whole point--it's easy to make the AI cheat.
@Uchawi: PvP balancing actually lends itself to the idea of the thread--very fluid roles. In PvP you are constantly forced to perform all roles. The tools characters will need in PvP are precisely the tools I want them to make full use of in PvE. It would be easier to design the PvE aspect of the game this way, because you wouldn't be in a constant struggle with the PvP aspect.
If one's "max dps" build continues to get him killed because he lacks the oomph in his defensive abilities, he will round out his build. If one's "max mitigation" build continues to get him nuked to death and be unable to kill anything, he will round out his build. If one's "max healing" build continues to put a bullseye over his head and daggers in his back, he will round out his build. These statements would hold true in PvP or PvE equally.
It's the PvP aspect of the game that will enforce well-rounded PvE character concepts, and defeat the so-called "inevitability" of roles devolving into the trinity.

![]() |
1 person marked this as a favorite. |

Another matter to consider would be solidity. One way of "tanking" is simply to get in the way and keep the monster from getting past you.
The same also applies to monsters
In a P&P game the evil wizard summons up minions and puts them between him and the heroes. You can't simply walk through the minions to get at the wizard (or at least not without the minions carving a significant chunk of flesh out of your hide).
Like wise the Heavily armored fighter in the group stands between the Monsters and the Squishies in the back, keeping them from being chopped down while they cast spells and rain arrows down on the Orcs.
Some form of collision detection allows a new array of tactics both for the players and for the developers to give the monsters

Squiddybiscuit |

There's very good argument to be made that the problem with the Trinity is the Tank. And the Tank works because of Hate. Hate was an interesting mechanic, but by metagaming it the Tank became possible, and once the Tank became possible, the Trinity was the only rational outcome.
WoW has gone down a rabbit hole of feedback due to the Tank from which they cannot easily extract themselves. But Pathfinder Online does not have to follow them.
We're going to focus design effort on the Hate mechanic and on other ways that PvE content interacts with players with an eye towards avoiding a feedback loop that would trap us in the Trinity.
RyanD
Glad to hear you're on this already.
I really, really dislike the whole concept of hate - that an enemy would continue to pound at one specific character because he has some "agro" ability while said tank is being constantly healed and DPS-players are attacking the enemy.
It just needlessly makes the A.I look even stupider.
There should ideally not be a single aggro ability, if you want the attention of your mob then you should just attack it hard enough for it to have to focus its attention on you alone - leaving its back open for a sneak attack or something similar.
Not 10-100 attacks from another player(s).
I think Tera Online has a novel system in that the tank has a longer reach with the help of pole/staff weapons, and is with this longer reach able to block mobs and their attacks - thus protecting their allies.

![]() |

Bobson wrote:Based on a number of the posts in this thread, this is what's knocking around in my head:Well all of that in mind, it would be awsome, but now we are talking a more sophisticated AI that has ever been considered in an MMO, that is not primaraly about combat with NPCs. AI programing is not cheap or easy, and while it would be awsome, I strongly can't imagine it being done in an MMO with a small budget, and an emphasis on PVP, rather then PVE. That's along the lines of wishing your house had a better security system than fort knox. It sure would be great, but it probably should be filed under not going to happen.
Simpler equasions are possible, i would believe that an AI could note the armor etc... possibly base things on damage delt by, and healing to current target done by, as the current AI's that I think should be packaged with the middleware do, but expecting extreme depth, and multiple categories of AI's to do, then we aren't looking at a small tweak, we are looking at a PVE investment on a WoW budget scale for a game that isn't intending PVE to be it's selling point.
That's the beauty of it, at least as far as it exists in my head: There is absolutely no requirement to have any I in the AI. There isn't even any random number generating (at least in the scripting - attacks and damage would be random). It simply solves 3-5 multi-variable equations (where every variable has a defined value) using basic arithmetic, then takes the highest result. It's the kind of thing that computers are really good at.
Admittedly, it'd be harder once you moved beyond the "Move up and bite it" level of abilities, and that's where it might bog down. But the basic "choose the threat" algorithm wouldn't be that hard to write, and would only require a small amount of extra effort from the developers whenever they added an object or monster.

![]() |

Kryzbyn wrote:Rather than have a cut and dry system, I say let the skill based stuff go, and let the player community figure out which builds work best for what role, which skill combinations net the best DPS, Resiliency, healing, etc. Let us experiment with what works, don't design or hardcode it for us. It's skill based, let's see what works and fits.Yup, I think the other major thing to note, PVE raids are not likely to be a large factor of the game. There may be a handful of instances, but I highly doubt they are going to be a large factor in how the game runs. Roles, tanking etc.. will be pretty moot points, because the invading army has it's own tactics that do not rely on a hate system.
Well, I know of at least one MMO (DCUO) that feature lots of PvP, and there are classes in PvP which take a lot more damage than others, while having slightly reduced damage output. Their major features in PvP is that they take less damage in general, do more damage than normal to Controllers, and take more damage than normal from Healers. (Controllers restore energy/mana to the group, and do more damage to healers, while healers restore health in addition to being paper to the Tanks' rock.)
Without adding arbitrary boni: Rogue beats wizard beats fighter beats rogue: Alpha damage beats controller beats defense beats alpha damage. The guys with the tower shields out in front prevent a few bow rangers from decimating everybody before the engagement begins, but start to fry when the lightning bolts come out.

![]() |

Onishi wrote:Bobson wrote:Based on a number of the posts in this thread, this is what's knocking around in my head:Well all of that in mind, it would be awsome, but now we are talking a more sophisticated AI that has ever been considered in an MMO, that is not primaraly about combat with NPCs. AI programing is not cheap or easy, and while it would be awsome, I strongly can't imagine it being done in an MMO with a small budget, and an emphasis on PVP, rather then PVE. That's along the lines of wishing your house had a better security system than fort knox. It sure would be great, but it probably should be filed under not going to happen.
Simpler equasions are possible, i would believe that an AI could note the armor etc... possibly base things on damage delt by, and healing to current target done by, as the current AI's that I think should be packaged with the middleware do, but expecting extreme depth, and multiple categories of AI's to do, then we aren't looking at a small tweak, we are looking at a PVE investment on a WoW budget scale for a game that isn't intending PVE to be it's selling point.
That's the beauty of it, at least as far as it exists in my head: There is absolutely no requirement to have any I in the AI. There isn't even any random number generating (at least in the scripting - attacks and damage would be random). It simply solves 3-5 multi-variable equations (where every variable has a defined value) using basic arithmetic, then takes the highest result. It's the kind of thing that computers are really good at.
Admittedly, it'd be harder once you moved beyond the "Move up and bite it" level of abilities, and that's where it might bog down. But the basic "choose the threat" algorithm wouldn't be that hard to write, and would only require a small amount of extra effort from the developers whenever they added an object or monster.
Any system will be subjected to gaming. That's part of the point of the game. Every algorithm I can imagine will boil down to assigning each potential target some number, and them attacking the target or targets with the highest or lowest number. I just think that things like "Is there an extra cost to attack?" and "Is this target almost dead?" should be included as well as the other factors of hate. Ability use is an easier thing to script: Use the available ability with the highest preference value. Ability availability and preference values, on the other hand, falls into the realm of balance and testing, hardly the ideal place for experimentation in a sandbox game.

![]() |

My point was not about the level of difficulty presented to the players but the variety of challenges and prehaps the manner in which difficulty is achieved.
Exactly! I would very much like to see an emphasis on players being able to recognize and adapt to new situations in PvE, rather than players being rewarded for knowing what's about to happen, and being ready for it.
As far as the original topic of the thread goes, I've always thought it made more sense for a Tank to generate Threat by actually being able to do massive damage if the enemy wasn't actively defending against him, rather than by calling the enemy names. It would be a simple thing to give Fighter-type archetypes attacks that did significantly more damage if they were flanking or behind the enemy.
Also, another way of getting away from the Trinity, is to make sure many archetypes have a variety of unique ways to either increase damage out and decrease damage in. A "healer" shouldn't be required as long as you have some player who can either slow down the attacker, or weaken them, or maybe guard the person being hit, or cast spells to increase their armor.
The thing I would most like to see is a balancing where a wide variety of options to increase damage out is balanced with a wide variety of options to decrease damage in. It would be best if many of these were mutually exclusive so that you don't fall into the rut of needing a Healer *and* a Slower *and* a Buffer, etc.

![]() |

GrumpyMel wrote:My point was not about the level of difficulty presented to the players but the variety of challenges and prehaps the manner in which difficulty is achieved.Exactly! I would very much like to see an emphasis on players being able to recognize and adapt to new situations in PvE, rather than players being rewarded for knowing what's about to happen, and being ready for it.
As far as the original topic of the thread goes, I've always thought it made more sense for a Tank to generate Threat by actually being able to do massive damage if the enemy wasn't actively defending against him, rather than by calling the enemy names. It would be a simple thing to give Fighter-type archetypes attacks that did significantly more damage if they were flanking or behind the enemy.
One significant flaw in that solution is how to handle situations in which there are multiples of that archetype in the fight.
Heresy waring: Fourth edition used a mechanic where one character could 'mark' an opponent or opponents. A 'marked' creature takes a penalty to attacks that do not include the creature marking it, and many player classes gain an additional benefit against a creature they have marked that makes an attack that does not include them. A creature be marked by only one creature at a time.
I like the idea of bonuses to attack an enemy when they attack someone else, but adjusted to avoid abuse: Creatures may be choose to be in high profle, normal, or low profile modes. Creatures in low profile take some penalty (like reduced weapon damage; preferably something that wouldn't bother wizards). Creatures in high-profile suffer a different drawback (unable to use stealth? Something that wouldn't bother a fighter) Creatures attacking a creature in low profile suffer some penalty (reduced defense?) against attacks from any creature in high profile. There is some cost with changing profile in the middle of combat (Perhaps retaining all of the drawbacks of both while gaining the benefits of neither for a period of time).

![]() |
1 person marked this as a favorite. |

Nihimon wrote:GrumpyMel wrote:My point was not about the level of difficulty presented to the players but the variety of challenges and prehaps the manner in which difficulty is achieved.Exactly! I would very much like to see an emphasis on players being able to recognize and adapt to new situations in PvE, rather than players being rewarded for knowing what's about to happen, and being ready for it.
As far as the original topic of the thread goes, I've always thought it made more sense for a Tank to generate Threat by actually being able to do massive damage if the enemy wasn't actively defending against him, rather than by calling the enemy names. It would be a simple thing to give Fighter-type archetypes attacks that did significantly more damage if they were flanking or behind the enemy.
One significant flaw in that solution is how to handle situations in which there are multiples of that archetype in the fight.
Heresy waring: Fourth edition used a mechanic where one character could 'mark' an opponent or opponents. A 'marked' creature takes a penalty to attacks that do not include the creature marking it, and many player classes gain an additional benefit against a creature they have marked that makes an attack that does not include them. A creature be marked by only one creature at a time.
I like the idea of bonuses to attack an enemy when they attack someone else, but adjusted to avoid abuse: Creatures may be choose to be in high profle, normal, or low profile modes. Creatures in low profile take some penalty (like reduced weapon damage; preferably something that wouldn't bother wizards). Creatures in high-profile suffer a different drawback (unable to use stealth? Something that wouldn't bother a fighter) Creatures attacking a creature in low profile suffer some penalty (reduced defense?) against attacks from any creature in high profile. There is some cost with changing profile in the middle of combat (Perhaps retaining all of the drawbacks of both while gaining...
Again none of this is really all that neccesary if you simply bring collision detection and physics back into the combat engine (and with todays technology there is no longer a valid technical reason not to do that...like there was in the 80's).
It's kinda irrelevant who the mob WANTS to attack...if they only can physicaly get to a limited number of opponents... they are either going to run away or fight the person who is actualy in thier face and hacking at them with a sword and blocking the path to that Mage they might really prefer killing, but simply can't get to.

![]() |

Simple collision detection would be too trivial to exploit: Provide a long path around the line of defenders to the hated target; when the enemy gets close, close the gap at that end and open one at the other. Issues come up because movement now needs to be acknowledged and permitted by the server, rather than simply acknowledged. If something got in the way due to latency, then movement becomes very difficult for the player, even if it should be trivial for the character.
Blocking all or enemy movement should be one option, but shouldn't be automatic for everyone who exists.
Look at an American Football play- equal numbers of equally capable people can't be blocked for long. Implementing collision detection also requires deciding what to do when people collide- and also making decisions about whether to bull-rush the line or kill it. That hardly simplifies the consideration.

![]() |
1 person marked this as a favorite. |

One significant flaw in that solution is how to handle situations in which there are multiples of that archetype in the fight.
I'm having a hard time understanding how this is a flaw.
For my part, I want my fantasy game to be as close to reality as possible within the genre. That means physics and collision for sure, but it also means logic: In a melee fight, anyone who gets surrounded by experienced warriors who are close enough to hit him is going to die very quickly.
Granted, it's a flaw if you're primary goal is to balance scripted content, but I think we need to try to avoid that mindset.

![]() |

Simple collision detection would be too trivial to exploit: Provide a long path around the line of defenders to the hated target; when the enemy gets close, close the gap at that end and open one at the other. Issues come up because movement now needs to be acknowledged and permitted by the server, rather than simply acknowledged. If something got in the way due to latency, then movement becomes very difficult for the player, even if it should be trivial for the character.
Blocking all or enemy movement should be one option, but shouldn't be automatic for everyone who exists.
Look at an American Football play- equal numbers of equally capable people can't be blocked for long. Implementing collision detection also requires deciding what to do when people collide- and also making decisions about whether to bull-rush the line or kill it. That hardly simplifies the consideration.
It's really no different then most of the issues that currently exist due to lag..for example pop out from behind a wall...attack the target with a ranged ability then pop back..lag prevents the target from "firing back" because you are now out of LOS by the time the server registers the return attack...remember the server can't rely on the client to tell it that it had LOS at the time of attack because the client can LIE... or...RUN through the the opponent, because of LAG by the time his attack is registered...you're already BEHIND him and he either can't execute his attack because of facing OR with auto-facing he is now exposing his back (for a backstab) to the guy behind you.
I would respectfully submit that...
A) Lag sucks regardless of Collision Detection
B) Lack of Collision Detection introduces just as many issues due to Lag as implimenting Collision Detection would.
C) In order to address EITHER set of issues...the server would need to be aware of the relative positions of the objects that are interacting with each other.
D) Lag with Collision Detection would be relatively hard for a player to exploit in the manner that you described. The players "closing the Gap" would have to be aware that Lag would occur at the right timing for thier movements to be registered to intercept the attackers movements but not for the attackers to realize that and switch course...they'd also have to be sure that they weren't Lag'd enough to miss the attackers movements otherwise they might open up a gap in the line for the enemy to run right through that was solidly covered before (i.e. they'd have to be able to predict..without seeing it...that the attackers would path for the gap). Finaly...since you are now talking about a "Line of Defenders" shifting over as one coordinated movement...they would have to have some way to be sure that thier OWN movements weren't LAG'd at different rates otherwise they'd end up bumping into each other with Collision Detection and going nowhere.
In other words I can't see such a manuver being coordinated and executed by a group of players in any sort of predictable fashion that can exploit LAG due to the unpredictable nature of LAG itself. I rather can see some exploits that occur WITHOUT Collision Detection (or irregardless of it) that could happen with Lag in a more reliable fashion.

osuracnaes |

Hm... I'm surprised a thread about getting rid of the Trinity went so long without a mention of Guild Wars 2. :P
They're dropping the tank/healer/dps setup by making a large amount of damage avoidable and giving every class at least one self heal. The action RPG style combat makes this possible, IMO. I'm not sure how you'd manage it with the auto-attack system that seems to be more common.
On the topic of hate/aggro/threat in PFO goes, I think a great mechanic would be to use something similar to the Tactics sections in enemy stat blocks. Just flipping through Stolen Land gives some good examples:
- Rangers preferring to target favored enemies
- Inexperienced bandits chasing players
- Creatures grappling smaller targets
- Vermin and animals just attacking whatever is closest
- Rogues attempting to flank targets
- Killing healers first!
There are so many possibilities other than attacking the person with the highest threat!

Obsidaeus |

I am sorry reading this made me laugh at times...You need meat shields with armour..just like on tabletop..no mage is going to go toe to toe with heavies..thier spells cancel..I hated this in Age of Conan..mages and archers shooting off stuff in the face of attacking mobs...if you want mage tanks and squishy tanks then i figure add some realism ..like your hands are casting spells and a blade from the mob chops them off...I am sorry who in thier right mind in cloth armor or regular clothes and no weapon just chanting and waving thier hands in front of a mob would be sane...as for these hybrid min/max characters they just drive me insane..you get a party of ppl together and this is why a game like Skyrim MMO would not work..everyone would spend hours argueing over who heals and who tanks or who does what role..or when they all died who didn't do healing or such...It creates a mess of characters with no clue how to play thier characters or roles.This idea is made by ppl who want godlike characters to do everything in game..have all the spells..all the fighting..all the rogue abilities etc. etc. on one character...so looking at new games in the pipeline like say Secret World..everyone has open skills and you lock in 7-8 skills and build from you skills..this forces ppl to play healers, tanks, buffers whatever..If i want to play a basic fighter to 20 i don't want someone screaming at me to munchkin my character to heal others or sacrifice my capstone dabbling in multi skill trees to make the rest of the minmaxers happy..just like in war someone has to be on front lines..someone has to support..just my opinion

![]() |

I am sorry reading this made me laugh at times...You need meat shields with armour..just like on tabletop..no mage is going to go toe to toe with heavies..thier spells cancel..I hated this in Age of Conan..mages and archers shooting off stuff in the face of attacking mobs...if you want mage tanks and squishy tanks then i figure add some realism ..like your hands are casting spells and a blade from the mob chops them off...I am sorry who in thier right mind in cloth armor or regular clothes and no weapon just chanting and waving thier hands in front of a mob would be sane...as for these hybrid min/max characters they just drive me insane..you get a party of ppl together and this is why a game like Skyrim MMO would not work..everyone would spend hours argueing over who heals and who tanks or who does what role..or when they all died who didn't do healing or such...It creates a mess of characters with no clue how to play thier characters or roles.This idea is made by ppl who want godlike characters to do everything in game..have all the spells..all the fighting..all the rogue abilities etc. etc. on one character...so looking at new games in the pipeline like say Secret World..everyone has open skills and you lock in 7-8 skills and build from you skills..this forces ppl to play healers, tanks, buffers whatever..If i want to play a basic fighter to 20 i don't want someone screaming at me to munchkin my character to heal others or sacrifice my capstone dabbling in multi skill trees to make the rest of the minmaxers happy..just like in war someone has to be on front lines..someone has to support..just my opinion
I have been playing MMOs with skill systems like Skyrim (Saga of Ryzom) for many years and have yet to see a fight caused by who is going to do what. Normally, in team creation you start with a few people who know what they want to do, then you specifically recruit only what you need.
This said, I otherwise agree with you 100%.

![]() |

This idea is made by ppl who want godlike characters to do everything in game..have all the spells..all the fighting..all the rogue abilities etc. etc. on one character...
It's always best to start from the premise that people are arguing in good faith, and are actually trying to achieve what they say they're trying to achieve, rather than pretending you know their true, hidden motives.
I am convinced that the primary motivation for trying to get rid of the Trinity is to make it possible for a group of friends to play together, without forcing any of them to play a class they don't really want to play.
The way you're describing the system doesn't really get rid of the Trinity, it just allows everyone to assume any role in the Trinity. I don't believe that's what PFO is after. I think they're actually trying to remove the need for the separate roles. I don't have the slightest clue how they plan to do this, and I wouldn't be surprised if they don't yet either, but I'm hopeful.
I would add that another very worthy goal along these lines is to make it possible for long-term experienced players to group up with relatively new, lower level players and still make a viable group.

![]() |
3 people marked this as a favorite. |

Roles are not something that is put into a game. They are something which arises out of differentiation. Every character archetype will be better at healing, or dealing damage, or taking damage, or being attacked without taking damage, or distracting opponents until later, or buffing, or some combination, or something.
There is no way to remove the advantage gained by playing to character strengths and using group members to compensate for weaknesses. As long as characters have strengths and weaknesses, specialization will occur.