Dynamic HM enemy builds

HM PvE suffers, in my opinion, from a lack of real surprise. I would love to see the following suggestion implemented, to add some challenge to HM areas.

To start, the game would need to start recording player group builds used in general areas of the game. For instance, a database would be kept of all the group builds used in HM Kourna for the most recent two days. This data would be comprised of primary and secondary profession, attribute point distribution (including values from runes), and skill selection (including PvE skills).

When a group entered any area in HM, 2-4 groups of enemies would be spawned, each specced to the exact teambuild found in the database for that region. For instance, if I entered Turai's Procession, I might get a group of Kournans and a group of insects, each comprised of eight members, and each utilizing a recently used player teambuild. I would know when I encountered such a group, because instead of the traditional enemy names, they would be named Kournan Elementalist, Kournan Monk, etc.

The groups would be comprised of level 20 enemies, but they would have the same radar range as a player, meaning that as soon as a player saw the group, the group would also see them and engage. Additionally, any enemies in these groups using minion skills would have a full group of minions ready.

This would change the environment of HM PvE, as players could not rely on simply building for known enemies in the area. Additionally, players would have to be able to quickly assess the build of an incoming group and implement a strategy to combat it. While you may get the luck of the draw and face a poorly put together team, you would be pretty likely to encounter Discordway, Sabway, Rayway, Physicalway, etc teams with regularity, and would have to know how to fight them.

It seems to me this would make HM more challenging, while keeping it as fresh as PvE can be in a game like GW. Opinions?

Nice idea, but monitoring players' builds used in an area sounds like a lot of work, and there's a good chance that some of those teams will just be bad.

I would rather see that each creature has two to four build variants, and monsters within mobs would have a random chance of being given each variant. Even then, that might still be asking for a lot.

2 problems:

1. Bad team builds by bad players would make that a joke.

2. PvE skills which are very often used and very overpowered,enemies would either rape ur party with skills like PI,YMLaD and FH!,or be entirely frustrating taking down an imbagon team with SY/TNTF with strong healers especially for eles which doesnt have armor ignoring skills.

I believe recording player builds'd be bad,maybe if they made a database with various tipes of team builds and random select them it could work better.

With that in mind,lvl20 would be easy especially on HM,so i'd buff them to 22-24 on NM and 26-30 on HM for a bit of challenge.

Good idea tho

Good point about the bad builds.

Maybe the solution would be to only record team builds that successfully vanquished an area. This should hopefully weed out a lot of inferior builds, especially after the system got rolling, and players running the bad builds had to successfully defeat better player builds to complete a vanquish and get recorded.

As far it being difficult to fight a team running PI or SY/TNTF, I guess that's what's appealing to me. The advantage will still be on the players' side, in terms of being able to make tactical decisions on the fly. But it would definitely be more difficult than just rolling the same old mobs.

ha i like this idea, but sounds like alot of work too.

i'd love to see the enemy have 3 necros, 1 /rt healer, 1 bomber, 1 spiker xD would be amusing trying to vanquish with them teams dotted around.

or a bunch of discord necros.

one question...would they have my pain invertor or any other PVE only skills ?

So every time i vanq ill come accross D-way/Sabway/Paraway.
Sounds bad.
If you did want this, the best idea would be to have a few varied skill sets for each prof and species of enemy, the skill set would be randomly chosen when you enetered the area.

Keep in mind that AI often doesn't understand how to use a lot of player combos.

this wouldnt work whatsover imo.

what ever build/variants that are run by enemies, a player will always overcome them with a gimmicky build.

enemies dont have the AI to run player made builds. what about people who clear areas with a perma sin and echo cryway on a regular basis? can you see a sin just smashing skills and cryway enemies echoing all sorts of crazy shit?

regardsless of what builds you put on a foe, they will not be able to run it efficiently. all HM is, is making enemies hit you harder. because when you get down to it, thats all they can do. cast quicker, hit faster and harder. but as far as co-ordination, foes havent got shit.

a foe hits you with PI, call it, you cast PI on that foe and watch him die as your monk out heals their monk. you face 8 foes all packed with searing flames, you flag your heroes out of aoe while they all ball up.

so yeah.... not matter what you would try and impliment to mimic player builds/tactics, it would never work, as foes cant change to adapt to what your doing.

the only thing AI can do better than most players is interrupt on a mes... thats about it

If Anet could implement such a feature, they'd be rich. The amount of coding involved for such a thing would require not only tons of man hours to comlpete successfully, but would require a severe upgrade to their servers and a monthly fee to be charged.

As noted, builds that are common would be even more annoying to face than the current ones. Running into groups of 3 Necros with Discord and hexes/conditions and a A or /A with AP would get old fast.

When HM was first introduced, I had hoped for what we got in GW:EN, not what we got. I thought a good plan for making the game more difficult would have been to give monsters a secondary class and skills from it. Hard rezes alone would make things more challenging in certain areas, but even a Warrior with Plague Touch would be neat to see.

Sadly, nothing will happen anymore, except work on GW2, skill updates for balancing, and bug fixes.

1. Anet alrdy can see everystep we take, but saving builds in a database takes space, which they don't want to use.

2. AI doesn't always understand player builds, this will be abused.
Farming anyone? AI doesn't understand keeping up enchantments.

3. Just give mobs about 4 builds which are random chosen will work fine. (anti farm)

4. You would get weird shifting in builds. as everybody runs something with minions everybody would take Anti minion skills such as [Verata's Gaze] and [signet of creation] and corpse exploid skills which are faster then the minion masters. After the builds change enemies have those anti necro skills, so everybody won't take minions making enemies build quite useless. one shift later it's safe to take minions and we are back at step one.

I agree to a point. Some 8-player builds would not be run well at all, and would provide less of a challenge than hoped.

I would expect to see quite a few builds, however, that were meant to be run by a player and 3H/h or 6H. These builds, if they are put together well enough to vanquish, are designed with the limitations of the AI in mind, are are already proven to be effective in the AI's hands.

A build like Sabway hardly needs player input at all. A build like Physicalway is similar. The Orders Derv knows what to do, as do all of the attackers. As long as the AI Imbagon spams his skills, the build works. A team like Discordway typically utilizes a caller, but really, the AI just heals and spreads hexes, and when it sees an enemy that meets Discord's conditions, they activate it.

The nice thing about a good deal of the builds you'd encounter is that they'd already be designed and tested to work with the AI. As for Perma/Cryway teams, do people really run those much in regular explorable areas? My impression is that were used more in Elite areas like DoA, which is where they would then pop up in this system.

Agreed, but that's not the point. The point is to make the players adapt to what the enemies are doing. Which is exactly what you illustrated above. The intention is not to stop players in their tracks and make them fail, it's to provide a bit of uncertainty that forces a bit of on-the-fly adaption.

Continuous shifting of the PvE meta is one of the benefits of the system. I doubt it would go as far as you illustrated, because I'm probably not going to change my build just because I may run into one or two MM's on a map. However, I can see something like this playing out in larger ways. For example, one thing the enemy team will certainly have are multiple rezzes. This is something that players do spec against, by bringing FS. Eventually, I could see this becoming more common is HM explorables, which would mean that enemy groups would start carrying FS more and more as well. It would certainly not be something a player team couldn't overcome (just kill the spirit), but it would add an element that's missing from most battles.

Sounds like you want a team based version of the Doppleganger. No thanks, Dopple is easy enough to beat, why make an entire explorable zone the same?

that is the point though. players already adapt to what the enemies are doing. your not talking about mimicing builds here, your talking about completely advancing and changing enemies AI.

i cant see a difference as it is between hero AI in NM and HM, except in HM it means you may have to take a few skills to counter overpowered lvl 30 skills.

but the foes still act the same way.

think of skills being used on enemies in terms of layers. a player may be loaded with an enchantment, so a foe may use enchant stripper, because the AI tells them that you have one on. the foes aren't just spamming that skill in hope you may have an enchant on you.

same goes for foes with interrupts. they arent looking for a specific skill in your build that they know is dangerous. they will just cast it as soon as they see you attempt to cast something, even if its flare. they have no conception of what skill you are usuing, just that it has to be interrupted.

so regardless of what skills you give the foes, pve, overpowered builds, or whatever, they will not act differently, i think what you want is a change to the foe AI... which isnt going to happen

I don't mind having creatures like 'humanoid', demi-humans, elemental and demonic with variable builds.

But creatures like beasts, insects, arachnids, dragons, skales, etc... those can't think of builds, those should stick with what they have. A spider can't learn to change builds.


Now, I don't think that making variable 'evolutive' builds for monster would be possible in GWO, given the time and resources they have.
But I think that it would be possible to give enemies a skill that once used is replaced by a random skill, much like when they use' copy skills'.

So, they could have 2..4 basic skills (the ones they can't be without, like monster skills or resurrecting skills if they had one) and then the rest of the skills, mainly the damage-dealing ones, would be replaced by random skills that depend on:
- The weapon they are carrying: for example, melee weapons would get more touches and point black AoE skills, and ranged weapons more range skills, and a creature without a sword won't get sword attacks.
- The attributes they have with higher values.
- The skills they already have, so they don't get the same skill twice.
That way the same skill could be used for all creatures. And Anet would 'only' need to create one single and very complicated skill, and then change all monster builds to replace all non-essential skills with randomizer skills.

Then, since the AI seems to be the same for most creatures, they should be able to use the skills right away.
And this would have the 'extra' benefit of making different enemies of the same type have different builds. One hydra could attack you with fireballs, while other attacks you with infernos.

And there you are, random skills.

Well, if you read what I'm suggesting, that's exactly the opposite of what I'd like. I'll walk you through it:

Why is the Doppelganger easy? Because you choose his skills. Since you choose his skills, you know exactly what skills he'll use. Combine that with predictable and exploitable AI, and he goes down easily.

Why are explorable areas easy to Vanquish? Because you know exactly what skills the enemy will be using, every time. Combine that with predictable and exploitable AI, and they go down easily.

What we already have are essentially areas of Doppelgangers, with the notable exception that the player doesn't actually choose the enemy's skills, he only has complete knowledge of them. Despite that difference, the player can build for any area with full knowledge of the enemy's strengths and weaknesses.

With my system, the player could build for the majority of the enemies he'd face in an area, but there would always be 2-4 groups of unknown enemies. They'd still have predictable and exploitable AI, but they would at least force the player to make a team build that could adapt to a variety of unknown enemy group builds.

I should note again that while the enemy AI is the enemy AI, the groups suggested would have one difference. With the ability to have the same radar awareness as the player, the battle field dynamic would be very different from the usual. The typical aggro dynamic means that the player can control, at least initially, which allies are visible to the enemy. This is one thing that allows tanking to work. Also, the typical aggro dynamic means that the player can see the enemy, get into position at his leisure, and choose the exact timing of engagement. Neither of these two things would be true with the new groups in this system.

First of all, when the special enemy group appears on the players radar, they will immediately engage, regardless of whether the player is ready, or already in the middle of another battle. This forces the player to give up control of the battlefield flow, and to explore more carefully. Secondly, tank and spank would not work, unless serious snaring/body-blocking could be quickly implemented.

I'm not talking about a change to the AI, other than the radar range issue discussed above. I understand your point about the AI. But keep in mind, there are so many players playing with H/H that a great many of the enemy groups one would encounter in this system would have already been fine-tuned to work with the current AI. AI minion bombers work well. AI necro healers work well. AI paragons work well. AI smiters work well. That's why players use them in H/H teams: they do an efficient job with minimal micro-management. So while some builds would not be as challenging due to AI limitations, many of the builds encountered would be, due to the fact that they were built for the AI in the first place.

A player can C-space through a region like Pongmei Valley HM, if his team's build is suitably prepared for the area. Once you introduce a few teams with unknown builds who are in control of the moment of engagement, the player has to actually start playing. It may be that the player gets lucky, and faces the Perma tank / Cryway team that doesn't know how to echo and spike effectively. No biggie. But when that team of Discord necros with 20-30 minions starts running toward him, he'd better have his plan ready.

Fair enough. It could be implemented that way if the lore aspect were a concern.

The point isn't to have random skills at all. The point to is face thought out and tested team builds, that don't necessarily follow the typical build pattern of a particular foe's group.

ok, i dont know if you dont understand game mechanics, or are confused, or are having a bad time getting your point across... but, i really cant see where your going with this. where is the balance in facing some perma that cant run the build and some echo cryway team that cant co-ordinate itself. where is the fun in that?

if you came up against a discord team it still wouldnt work... do the foes just start with 30 minions? or are they killing people for you to get minions? the only reason people can h/h so effectively with the AI is because you can give 3 of them builds to use.

back in the day when there were no heroes, people used to pug alot pre nightfall. towns were always filled with people looking for groups for missions quests etc, because heroes who had a set build werent all that good for the most part.

but the main point is, whatever build you give to foes, in any part of the game, people will find a gimmick, or overpowered build to counter it. dungeons are probably in the hardest line of things to overcome in pve, and elite areas in HM... if you use a balanced build.

if you want a real challenge where the foes builds are changable, the enemies are unpredictable, and react to every different situation, go to PvP... thats what its there for

So I'm wrong in reading your OP that you want the monster builds to be based off of what players use most often? Sounded like you wanted the monsters to pull their builds from the commonly used player builds. Sabway, Discordway, RoJway, etc. would all be the commonly used player builds for most areas, and so you'd see monsters with those builds frequently.

Sounds like Doppleganger to me. Player based builds being used by monsters..... exactly what the Doppleganger does, although you could face this version of the Dopple with a different build than he has.

No, in fact. That's not what the Doppelganger does, at all.

He doesn't pull a random build from a pool of recently used successful player builds. He uses your build. Every single time. Do you not see how that difference profoundly affects how a player prepares to fight him? Really?

Common advice for a Ranger facing the Dop would be to bring a bunch of long-casting nature rituals and a couple of attacks. This advice is given based on the fact that the Dop will have that exact set of skills, and will activate every single one of the nature rituals while the Ranger attacks him.

If the Dop instead grabbed a random Ranger build that had recently been successful in an area in the Crystal Desert, would that advice make any sense at all? Of course not. The Ranger would instead have to prepare a build that could deal with being attacked, could deal damage to the Dop, and could stop the Dop from healing that damage.

Do you not see how very different those two scenarios are?

What I am proposing is not a situation where players spec for known opponents. We already have that, in every explorable area. I am proposing a situation in which players spec for many known opponents, and a few groups of unknown opponents. Will players face popular builds more than snowflake builds? Of course. But a player will not know the exact build of every enemy he is going to face when he is speccing for an area. His gameplay experience will therefore be different.

I guess the fun for me would be seeing eight enemies appear at the edge of my radar, and seeing them running toward me. I would see an Assassin, a bunch of Eles, and a couple Monks, and I would have to figure out very quickly what I thought their build was so I could prepare for the fight. Is it a Cryway team? A Searing Flames team? Some other non-meta build? Do I need to flag my team apart, anticipating AoE damage?

Once I engaged the enemy, the build would become apparent. Even if the Sin wasn't really useful for anything, the Eles may all CoP my first ally suffering from a Mesmer hex, and I'd have an AoE spike on my hands. Was I ready for it? If so, this will be an easy battle. Cool. If not, I'd better recover quickly.

That sounds like fun to me, even if the team wasn't handled as well by the AI as by the human team that originally ran it. The battle itself may not be hard, but I'd be going through the entire area on guard for an unknown team, and would have moments to figure out what exactly I was about to face when I finally came across them. Because, depending on the build, the AI could handle it just fine, and I'd always have to assume that was the case.

That's the appeal to me. Currently, PvE does not have any situation like this.

Wow, really? You do understand that the player made build that the heroes handle so effectively would be the exact build the enemy would have, right?

And the minion question was answered in the OP: yes, they would spawn will a full battery of minions. After that, they'd be on their own.

Doppleganger is easy to beat because a) you know exactly what he's going to have, and so you can ensure he has a build that he can't use properly, b) it's a 1v1 match, and c) he can't use PvE skills. The OP's suggestion has none of those attributes.

so your just going to go ahead and assume that everyone has all 3 campaigns...

I'm all for more dynamic mobs in HM, but not exactly implemented like this.

Think about it, there shouldn't be a cave full of about 30 easy to kill enemies+boss each with the same skills just outside a major town with nothing between them and you. Why can't there be a few monk raptors or give them something to hit with that bypasses typical farm stuff(like why aren't there angorodons right in front of the cave or in it instead of just patrolling near it).

Take the Hydras in the crystal desert, they already have boss hydras of each class, why not disperse non-boss versions among the ele hydras, give them elites and send them on their way. Same deal with the trolls, vermin, tengu and other easily farmed crap, de-bossify the boss versions and spread them around.

Then it's just a matter of randomizing the spawns, which does already happen in most of tyria.

it would also kill the difference of each expolrable. the only change would be number of enemies.

So that meens to beat this strategy, is go inn with a surten team build a few times, go out and take the build you actual want and go in the area again.

Yep that would make it more inoing but not harder.

Just to clarify, every zone would still contain all of its normal enemies, with the addition of 2-4 of these groups.

That would indeed work, if you were the only player completely vanquishing areas in a particular region, like the Echovald Forest, over the most recent two day period. And to get your build saved, you'd have to complete a vanquish.

So the exploit would be:

1. Pay off the entire active playerbase not to vanquish in an entire region for two days.

2. Complete a vanquish with a build you wanted to face.

3. Re-vanquish, content in the knowledge that you will know the enemy builds for the area you'll be re-vanquishing.

I guess I'm not too concerned about anyone going to such ridiculous lengths to vanquish something twice.

Since it's HM, I'm going to assume players are OK with a challenge. I also don't really buy into the idea that players of one campaign are powerless against skills from a campaign they don't have. But that could be a difference of opinion.

So you think Anet will suddenly abandon its campaign setup for skills? Monks in Factions can't use Healer's Boon, but if HB is used enough by people vanquishing in Factions, the monsters will suddenly be using this skill? I don't see Anet doing that. And I still don't see how Anet's servers could handle the amount of info this would entail. How many people vanquish, how many areas are vanquished, how many various builds are used, how many types of monsters are there, etc., etc., etc.

I think the idea of a random monster generator would be decent, but this idea is not.

Keep in mind that regular enemies would have exactly the same skills as they do now. The only enemies affected would be 2-4 groups per area. That's it.

That said, what exactly would the problem be for a player playing in HM Cantha to encounter a Monk using Healer's Boon, or an Ele using Searing Flames? I understand that a Factions only player can't use those skills, but what exactly is the fundamental problem with a Factions only player encountering enemies with those skills (in HM only)?

honestly, think about what your saying... i cant be bothered to keep typing the same thing and over and over.

it would be extreemly unbalanced for a player who owns prophecies only to face mobs with skills from all 3 campaigns + eotn + pve only skills... in HM

Stating something doesn't make it true.

Why would it be imbalanced? Can you think of a situation where a Prophecies only player would be unable to defeat a group utilizing skills from all campaigns, assuming the player was skilled? Weren't you talking about how bad the AI was a few posts back?

And isn't HM optional?

I'll agree with you there. For a long time I had only Prophecies and EotN. As I went around the various zones in EotN I ran into enemies with skills from factions and NF. Obviously, I wasn't familiar with these skills; but this didn't stop me from owning every red dot on the radar.

As for your actual suggestion, I agree with those who say that it would be much easier to have multiple builds for enemies and have builds randomly assigned to individual enemies. That would spice things up a good deal without being anywhere near as complicated as your suggestion. The only thing I would add to that idea is that the builds should be significantly different. For example, the possible builds for a warrior enemy should not be:
[fgj (pve)][executioner's strike][body blow][dismember] and [fgj (pve)][galrath slash][body blow][sever artery][gash]
In that example, there's almost no difference in functionality.

I think I would have a helluva problem vanquishing areas with monsters that are using Searing Flames, God Avatars and Shadow Form, with only henchies.
So it's okay to wreck a players experience because it's optional, as in 'if you don't like it don't play it?'.

I would too. But then, I can think of more than a few areas in Proph that I'd question my ability to vanquish as a Proph-only player using all henchies. Lornar's, Twin Serpent, Ice Floe, Eastern Frontier, etc.

I could see the wisdom of this feature not kicking in if a team were comprised of more than half henchmen. This would eliminate the threat to hench only players in any campaign, but provide the challenge to players using heroes or human team-mates.

For the record, I don't think there's anything special about any of the skills you listed that couldn't be handled using only Proph/core skills. But I do see the point about all hench teams.

I guess I already see HM as 'if you don't like it, don't play it'. It's an optional mode, meant for those who are looking for more difficulty than NM.

I know I vanquished Tyria using [[Broad Head Arrow] to shut down casters en masse, and I suppose it would be 10 times harder when not able to reliably shut down casters (aka as 3 Dolyak Masters...). [[Concussion Shot] is kinda meh because of the fast casting of foes in HM.

Make an UM (Ultra Mode). It would not be fair to make HM harder for the players that haven't started HM yet, compared to the people that already have completed HM. This way, your idea would be truly optional.

It would be nice to have random builds on the monsters, but have ones that are created by anet, not recorded from peeps. and it would be hilarious if there was a mode where monsters had max pve skills =D

ANet can create a pool of builds for the enemies to have. Not crap builds but ones the enemies can use. The community could give ideas for the builds.

When the zone is instanced the groups are already "static" with professions but the builds are pulled from the pool of builds. It would be random each time you instance. Mind you it would be a little more work for ANet but they would have to match weapons with the builds as well. It would be random. It would give players a bit more variety.