The role of aggro and tanking in Em8er

[FenrisAmarok]

The purpose of this thread is to discuss the inclusion of aggro management mechanics in the game and the pros and cons of different methods of doing so. Please note that this is my own take on the discussion that occurred in discord and my position is biased in favor of including aggro management in the game. Also note that this is going to be a loooooong post. I’ll post a tl;dr at the end.
First, a couple of points:

1. I’m aware that the Em8er team have not discussed this yet. I’m also aware that this was not in firefall. The purpose of this thread is to discuss what we, as a community, want to be in the game and to brainstorm ideas of how the Em8er team could go about it.
2. When I say aggro, I mean the mechanics by which an enemy decides who it is going to target. Aggro management, therefore, are the mechanics by which players can affect this system. The only way I know to do this back in Firefall was stealth, as this essentially reduced your aggro to zero. When specifically talking about the numerical value of an enemy’s attention on any person, it can also be called “threat”.

This is meant to be a continuation of a discussion that occurred in Discord, so I’ll also summarize that here:

• Aggro management was first broken into 2 types: hard taunts, which forced the enemy to target you regardless of the amount of threat another player had accumulated, and threat modifiers, methods by which a player could increase or decrease the amount of thread they gained.
• • There were a good amount of people who did not want hard taunts included at all, but most seemed at least okay with thread modifiers.
• Another idea to manage enemies was brought up that did not include aggro mechanics was something akin to area denial or crowd control. This would include abilities that blocked off an enemy from an area (keeping them away from allies) or using a heavy’s shield to block for allies
• • I personally don’t like the shield idea as is. The general consensus is that the base shield arm would only be big enough to protect the heavy itself, and any others who wanted to hide would have to do so right in his shadow. I don’t think this encourages a support role, and would rather see shields that are or become wide enough to provide protection to allies next to and behind him, like a Reinhardt shield in overwatch.
  • I’m not opposed to enemy management through area/crowd control. I think area control would be more conducive to a “support” role than simply slowing, rooting, or even knocking back enemies.

My position is that players should be able to manage their threat levels and, through skilled play, be able to affect who the enemy targets. I think it would be best to look to older MMOs, like WoW, and how they handled aggro, despite this being a shooter and not an MMO.

In WoW, each player in combat had a position on the “threat table” and a value on it. Damaging enemies and healing allies were the main ways that a player would increase their threat. Thus, the players doing the most damage or most healing would be the ones targeted by enemies. This makes logical sense, especially for the intelligent life forms we’ll be fighting. Most DPS classes had a talent or ability that allowed them to lower the amount of threat they generated or temporarily remove themselves from the threat table (like by becoming invisible). Tanks, then, had abilities that increased the threat they generated or abilities that forced enemies to target them (a taunt) for a few seconds.

I specifically want to talk about WoW’s classic “taunt” ability. This had two effects: first, it forced a single enemy to target you for a few seconds. Second, it increased the user’s threat value to be equal to the enemy’s previous target. This allowed the tank to immediately peel enemies off of their allies while also getting their threat value to a point where they can surpass their ally’s.

Distance also affected aggro management. If an enemy was next to a DPS and a tank and was currently attacking the DPS, the tank would have to exceed the DPS’s current threat value by 10% to change the enemy’s target to himself. Alternatively, if the tank had control of an enemy and a ranged ally was attacking said enemy, the ranged ally would have to exceed the tank’s threat value by 30% to change the enemy’s target. Thus, even in a game where “tank and spank” was the general way to play, positioning was still important.

So what do I want to see in Em8er? I want to see an aggro system similar to what I described above. You deal or heal more damage? Targeted. Standing closer to an enemy that you should be as a squishy ranged DPS? Targeted. I then want to see ways for players to manipulate this. I’m currently working on an omniframe build idea that would demonstrate this, and hopefully I’ll post that soon, but here’s just one ability I think would be a balanced and skill-based adaptation:

• Affect/hit enemies in a semi-circle in front of you: Raise your threat to be equal to the target with the highest threat among affected enemies. For ten seconds, deal 10/20% increased damage to enemies that aren’t targeting you.

This ability allows you to catch up to the threat values of your allies. Because it raises your threat to the highest among current targets, most of the enemies should target you, with only one or two that are still targeting that high threat ally. That’s where the second effect comes in: Dealing increased damage to enemies not targeting you certainly helps kill them, but the idea is meant more to generate more aggro on specific enemies by increasing your damage to them.

While this skill can definitely be used on distant, ranged enemies by running up to them until they’re in range, I also see it having a special place keeping melee enemies from attacking your allies. Again, hopefully a fleshed out version of this will be coming soon.

I’m not sure if I’ve made it completely clear enough, but I think these types of mechanics are completely viable in Em8er. If the plan is to really let players build their own classes, then don’t let DPS and healer players be the only ones satisfied! Tanking is a support role that was sorely missing from Firefall, and along with the lack of weapon diversity in it, this was one of my biggest complaints. Em8er has fixed one, please fix the other!

I think that’s all I’ve got for now! Since this is such a long post, please either quote or state which section you’re responding to to keep confusion to a minimum.

tl;dr: Aggro mechanics and tanking can and should be adapted to Em-8er, and the tanking role is something that was not supported in firefall and should be!

 

[Omnires]

As a person who loves to play stealth having different ways to manage aggro is key. Both tanks and DDs (damage dealers) use aggro management in different ways. Where tanks are trying to draw as much aggro onto themselves as possible. DDs will often try to reduce aggro or redirect to something else allowing them most time to set up their attack or combo for massive damage. Even non-stealth DDs do a lot better with one or two ways to lower aggro or move it away from themselves. For example, let's say you are a mage but you need 10 seconds to cast you large AOE spell that does a ton of damage. If you are alone for some reason might never get to use that spell because the enemies always attack you before it is done interrupting your cast. But if you had say a invisibly spell or decoy spell that can keep aggro off of you just long enough to get the spell off, then now you can use even if you are alone.

I've seen a lot of MMOs where some abilities are just useless of you are solo because you never have enough time or space to use it. Because the devs of those game never thought to give aggro management abilities to the non-tanking classes/players. Or worse they do even aggro management abilities but the abilities never work the way that should. Example "How do they keep hitting me, I'm invisible and not making any sound. So why does every enemy always know where I am?"

 

[Mahdi]

First, well composed and thought out.

Two, big challenge for the team to create an avenue with this content.

Personally, I am torn on the issue. Support has always been my favorite roles and tanking is at the top. I want to jump on this wagon but will fight that urge to quietly cheer from the side. I never felt firefall needed a tanking role. And with Em-8ER not being mmo, I dont see threat management being necessary. Please understand I am not opposing your view, just think a new innovative way is needed for this project.

I cant put it out organized in any fashion, but I would think squads and/or formations are more of a threat than individual builds. Say the Tsi-hu can understand the frame types and can access the tactical impact that group would have in either defending a base or (I hope) attacking a Tsi-hu stronghold.

Example off the top of my head. Tsi-hu have a siege weapon of a sort that fits the mold of slow mobility. A specific assault grouping of omniframe types moving in would be a priority target to defend against.

For us attacking them, I lean towards frames that have LRMs or mortar bombardment capabilities that are the threat, this being a focus.

I believe in a form of tactical threat over unit threat for the fact this brings hundreds of players together. Instead of pigeon holing builds, we as a strikeforce can come together more unified and communicative. Think of guilds big enough to have squad leader roles? Think of well known names people want to group with from previous experience?

To me this keeps all builds viable for how a gamer want to play. That is Em-8ER's soul.

 

[Drakin5]

In Em8er, distance, signature (heat or mechanical), frame type (it’s going to be heavy since, well they got a shield)... there’s a lot of variables on how aggro would be handled. Another one is pacing your damage dealing. As in, you would sacrifice your DPS over time so that you wouldn’t get all the attention.

Another thing that I found awkward is that healers don’t get aggro from all the support they do so...if by AI view, shouldn’t they too get targeted first since that healer has been keeping all of his allies from dying?

See this:
https://tvtropes.org/pmwiki/pmwiki.php/ShootTheMedicFirst/VideoGames

the tanking role is something that was not supported in firefall

Yep that was a problem even before 1.6. Case in point. Every time I joined a Kana raid be it normal or HC, I always get fire blasted first. I guess that shotgun also doubled as a draw aggro weapon.

 

[Pandagnome]

What if someone has the biggest carrot!

Chaos
Activated
Release
Reaction
On
Target



This item could be crafted to lure who ever processes the carrot!
Though the carrot does not last long because once digested it is no more and
perhaps even the carrot holder!

I would think squads and/or formations are more of a threat than individual builds

I think so too such as squad numbers, experience, and formations but say you wanted to add that extra threat could this carrot bait be passed on like a hot potato so whomever has this would gain threat maybe this could be just another mini game i don't know... but it doesnt have to be a carrot though could see it with other skins but a carrot just seems good in my mind

 

[FenrisAmarok]

Thanks for the replies everyone

Two, big challenge for the team to create an avenue with this content.

They're going to have to program methods of aggro for the AI anyway. Even if they don't add ways for players to manage aggro, I really hope they make the targeting make sense. I just remember in GW2 how for most enemies or bosses, there's absolutely no telling who the boss is going to go to. Doesn't matter who has the most health, most damage, most healing, proximity, etc. It was almost random and everyone disliked it XD

I cant put it out organized in any fashion, but I would think squads and/or formations are more of a threat than individual builds. Say the Tsi-hu can understand the frame types and can access the tactical impact that group would have in either defending a base or (I hope) attacking a Tsi-hu stronghold.

Example off the top of my head. Tsi-hu have a siege weapon of a sort that fits the mold of slow mobility. A specific assault grouping of omniframe types moving in would be a priority target to defend against.

I definitely get what you mean, and yeah, for large scale battles like that the shift definitely moves away from individual builds. But what about thumping? When there's only, say, 5 people, swarms of tsi-hu and category 0 kaijuu are coming and there's still a category 1 to deal with? I think a tank, or something that can keep aggro off of characters that are solely DPS focused, would be really useful. I don't think every build needs to be super viable in every situation. Just like a sniper isn't going to be as useful in a close quarters fight, a tank may not be necessary with overwhelming AOE.

But even in the siege scenario, I would hope that those mortar teams aren't all it takes to fend off the wave. And if they're not, why shouldn't some players be able to protect them? And if they are able to fend it off, isn't that stifling build diversity? Everyone would just be running mortar builds to end things as quickly as possible, which I don't think anyone here really wants.

I think allowing players to tank expands build diversity exponentially. Not only does it allow players that like that role to be happy with it, but it also allows players that know they'll be protected play as more glass cannons. Players that want to have more survivability can, and players that want to all out on damage can too. Healers are more useful when there's one or two people handling most of the enemies and others have less health.

I'm also not sure where guilds and players being well known fits into this. I didn't play vanilla wow long enough to get deep into raiding (got my pre-raid gear and then had no more time) but as a tank, you felt like a superstar if you handled the enemies well, and most people would compliment you for doing well. Same with healing. Saving the tank from near death saves the party, and everyone cheers when that happens. Again, I think (from what I'm understanding) what you want will become more prominent with tanking included, not without it.

I also want to be clear. I don't want tanks to be necessary for every, or even most encounters. Even thumping I don't want to have to require a tank. But I do want to have that option and I want it to be useful. I'm always disappointed in games (like vanilla maplestory, which I more recently played) where the tank character has no usefulness whatsoever. He won't die, sure, but he doesn't DO anything else, either. I want to have the option, and a system that makes sense, not a necessity.

 

[Thorp]

While I have had entertainment with mmos that employ Threat meters I do not believe Em8er should use Threat.

I understand AI is simulated by giving the computer a string of seemingly random decisions. Ideally the AI is constructed to calculate decisions that can be observed as near human. Video games have used Threat meters as a way to manifest or quantify a weighted decision for the AI to act upon. Giving the AI a measureable value called Threat allows the AI to weigh decisions and provide outcomes that are at least reasonable tactical responses. Though Threat meters have worked well for MMOs to instruct AI to attack healers or the heavy hitters I however do not perceive threat meters as an end-means to make AI act human-like.

I'm no expert in human psychology but if Em8er intends to have human-like AI it needs to avoid allowing players from directly controlling and predicting the decisions of the AI. Spells that "taunt" to guarantee drawing the attention of AI is disruptive to the flow of combat. AI's that are targeted by taunts are often perceived as obviously making a poor decision to attack the taunter which also draws attention to the robotic nature of the AI and the entire game. I would believe it unlikely that a human would change targets amidst an obvious advantage in their favor. Furthermore it would be unlike humans to break a chase to attack a taunting enemy who may be far away. The ability to forcibly call an AI to change targets breaks the realism flow of combat. Additionally having knowledge that a player can purposefully taunt the AI removes the real-time fear of making poor decisions or falling victim to unfortunate random "wrong place at the wrong time" scenarios. If Em8er wants to have human-like AI, the AI should change targets because the computer perceives a tactical advantage or a reasonable calculated choice to change targets.

In the effort to give players some control to save their teammates from peril, players can be granted skills/spells that affect the AI without the use of a Threat meter. Grapple hooks to literally pull the AI away from an endangered player provides the players an opportunity to save their teammates and not make the AI be observed as making an obviously poor decision. Once the AI is relocated the human-like response would be to reassess the choice of continuing pursuit or calculating if the remainder of their predicated lifespan would be best spent with a new decision. AI needs to be perceived as making on-the-fly decisions and not directly following a Threat spreadsheet.

Ultimately I believe in shooter games, players need to be tactically responsible for who gets shot. Obviously players who do not take cover are subject to being shot. Though the AI is most often going to see the "tank" first and thusly target the tank the AI should be programmed to maneuver to attempt harassment of the support/healers. I suspect a human-like AI would target the first enemy they see and will be capable of calculating a maneuver to expose the support/healers to danger. If the support/healers are never exposed to danger because the AI never targets them or the players can taunt the AI away I predict for myself, the game will get boring quick.

I do not believe Em8er would benefit from a Threat meter. Players need to be tactically responsible for avoiding bullets and covering eachother. The AI needs to be built smart enough to weigh whether changing targets would predictably produce a favorable outcome for the AI (more damage, longer lifespan, ammo/energy drain of the players). I believe gameplay is more immersive and compelling when the positioning and maneuvering keeps the players alive rather than a taunt spell that can forgive poor tactical decisions. Players need to play the arena, not a Threat spreadsheet. (This is why I loved Firefall PvP; the human element keeps combat scenarios new and always changing whereas the AI is often inherently predictable thus exploitable).

Em8er needs AI that uses weighted decisions because Threat will create an AI that is observably robotic and inherently more exploitable.

 

[EvilKitten]

Given that the Tsi-hu don't really care about anything other than wiping out humanity and also probably wouldn't understand what humans are saying or have knowledge of human body language, I don't see traditional "taunts" as really being a thing in Em-8er. Some of this may overlap with the OP and others but I will put it all together anyways.

I think that "aggro" should have a base modifier based solely on hit point changes. Incoming damage will generate aggression, modified by distance, and whether it is direct or AoE (AoE being significantly less). Abilities not directly tied to a player (hard to spot grenades, claymors or trip mines etc) and attacks from players unnoticed by the target who is already in combat should not generate any aggro.

Aggro should have a fairly steep decay rate, say 20% remaining after 10 seconds. Over time of course this will still build up, but large immediate damage will still be very important. If someone hasn't hit a target for a bit, and another player attacks them instead, it isn't that the NPC will forget the previous target, but the one actively shooting at them would take priority.

Aggro should propagate out radially from any mob generating aggro at the square of the distance. Aggro *can* be passed on via "radio" (or whatever Tsi-hu use) to others in their group who are out of sight but this would take action on their part. The range of propagated aggro should also be based on the rank and intelligence of the NPC. AKA a pack of wild animals would not spread aggro nearly as much as a Tsi-hu commander under fire.

Healing should act similarly, pinging out radially from each target healed, but the aggro given to the healer not the healed. A player healing a group of people close to an enemy will generate massive amounts of aggro, a healer far away healing a bunch of snipers would generate almost nothing.

Direct damage, high spike damage should be king of aggro generation. With a reasonably high decay rate large spikes of damage will overwhelm other aggro generators but will need to be constantly applied to maintain focus. A large spread out group will require multiple "heavy" direct damage dealers otherwise the fringe NPC's will target other players.

~~

Sniper rifles will generate large damage spikes, if fired from in close they will almost be guaranteed to pull aggro. Sniping from long range is extremely important.

Players with any Healing abilities should avoid AoE heals of close in targets, players suffering significant damage would be best to fall back so a healer can help them without pulling aggro. On the other hand someone wishing to play as a tank could use healing as a means of pulling focus.

Support setups wishing to avoid focus should avoid direct spike damage weapons and instead use constant DPS (SMG's eg.), AoE mortar/grenade weapons, or DoT type weapons to better manage NPC focus.

Those wishing to play as "Tanks" should always focus on high impact direct damage weapons at close range, Melee being particularly effective. Abilities that deal direct damage to multiple targets (say swinging a sword in an arc or using an ability that leaps from target to target) will generate the most aggro in the shortest amount of time.

 

[Omnires]

One of the real test of a good aggro and counter-aggro system are things like snipers and trappers. Because stealth players and people who can do long range remote killing can test the limits of how enemy A.I. work.

For example one of things I hate in games is when I'm hiding in the shadows somewhere with a soundless or sound suppressed long range weapons and when I kill an enemy in one hit with a head shot every other enemy within like 3 miles instantly knows where I am and attacks me. The same goes if I place something like a landmine somewhere without being seen then move to a new area, but again as soon as an enemy NPC get hurts or killed by my landmine every other enemy in the area instantly knows it was me how places the mine down and know exactly where to find me even if I have moved to a different area.

So if we can avoid having things like this happen. I'll be happy.