New efforts!

[Arcana]

I rememebr it mostly from the AE mobs. As soon as you entered their aggro radius you would have a crap load of throwing knives or whatever flung at you even if they were melee mobs. It used to mean I had plan accordingly when I first charged in before I had my buffs up.

I remember thinking it was cheap, actually. I mean it makes sense that they used their available attacks first.. but one would imagine at least SOME mobs would charge you. The majority of them just flung crap at you until you got in melee range or their powers were on cooldown.

Some fraction of them should have charged forward if they had no ranged attacks available, unless they were ranged preferred critters.  It is possible that the devs made *all* AE critters range-preferred to prevent them from being easily exploitable, but that would then be less an issue of the AI and more an issue of the devs actually wanting that behavior.  I do not off the top of my head recall if this is what occurred.

[Twisted Toon]

I think most of the players that noticed the change in difficulty, adapted to it fairly quickly then forgot that the difficulty had changed.

I'm sure the group I usually ran with did that. But then, they might not have noticed at all, since they usually ran missions with the settings at +4/x8.

It wasn't all that helpful for us the low leveled characters getting sidekicked up to join in those missions when they did that. However, I think all the defeats per mission was offset by all the ding-rez's in those same missions. I even had a few simultaneous defeat rez's. Those were humorous.


Oh look, a tangent.  :P

[Golden Aurora]

I also remember a certain distinctive pleasure in herding enough and abusing the geometery to make the mobs stack. Like on the corners of fences or dumpsters. It made those close cones (shadow maul) and burn patches lethal.

Speaking of burn.

I remember back in i4 how pissed off I was over the burn nerfs. It went from super useful to useless and build breaking overnight.
Actually I never really understood the point of the aggro cap. I mean if you legitimately have the skill and build to kill it... why shouldn't you?
It added significantly to the challenge and fun factor for me back in those days. It used to be a mini game in itself to see if I could wipe them out before the freakshow tesla cages stacked and broke my mez prot.

Ah the old days... South east Brickstown will forever remain fond in my memories. I really felt like a heroine then.

[darkgob]

Actually I never really understood the point of the aggro cap. I mean if you legitimately have the skill and build to kill it... why shouldn't you?

Because it's not skill, it's exploiting game mechanics and basically breaks the game's design intentions.  Game devs usually frown on that sort of thing.

[Arcana]

Actually I never really understood the point of the aggro cap.

There's the point, and then there's the point.  The obvious point of the aggro cap was to prevent players from aggroing unlimited numbers of NPCs into easy to kill clusters.  But there's a deeper point that is necessary to understand that one, and it goes to your statement about skill.  You say if players can do it, they should be allowed to do it.  There's a presumption that the game's design reflects that standard: that if it is possible, players should be allowed to do it.

But game developers don't obey that standard, and there's a good reason for that.  Sometimes, the only reason something is possible is because the devs made something else possible, and the other thing is an unfortunate side effect.  If you forced the devs to live by the rule "if the players can do it, you should let them do it" you force them to design and implement gameplay rules that are far more difficult than intended.  To avoid allowing players to do ridiculous things, the devs would have to make it hard to do anything.

You might think that this is anti-meritocracy, but it is not.  The question isn't whether higher skill is rewarded with higher rewards, the question is whether the game does so in a proportional way.  The truth is that if you could build and assemble a team capable of map-herding, then the whole notion of proportional rewards goes out the window.  It isn't twice as hard to herd twice as much with unlimited aggro caps, but you get twice the rewards regardless.  It doesn't take twice the damage to defeat twice the foes when you have unlimited target caps.  Let's face it: if players were only getting proportional rewards for those activities, most wouldn't do them because it wouldn't be profitable to do so.  They were doing them because they were getting radically disproportionate rewards for that activity.

Basically, things like caps and limits exist so that the devs could give us the most freedom possible within those limits.  Without those limits, the devs would have to resort to making everything more difficult, to fit the standard of "if the players can do it, they should be allowed to do it."  Under that standard, everything would have to be a lot more difficult.  With limits and guardrails you can make things very easy for casual players without allowing experienced players to gain unlimited advantages within the game.

In City of Heroes, you have the extra problem of the teaming system.  You can basically team anyone with anyone in almost any content.  If you create a system where stronger characters have wildly disproportionate ability to earn rewards, you disincentivize playing at any level other than that one.  You incentivize compelling players to participate in degenerate play like herding even when they *aren't* built to do it, when a higher percentage of teams are doing it.

[Remaugen]

Things got awful quiet around here. . . .

[Joshex]

Thanks Codewalker! that's helpful even if it's only a guess, the reset at tick actually helps me visualize this thing (I had actually been thinking of such a tick reset for a particular stat) but to have the whole stat system recheck per tick, has triggered the very things I needed to think about to move forwards.

And to think I'm just going back through and reconsidering the single target logic atm adding in checks and balances so only one 'user' can access a stat field or list at a time and the rest wait their turn without too much activation lag. still gotta reorder a couple things and add in a power state recheck before the final 'activate animation and -hp/+effect' kicks in.

also gotta redo the hip rigging, finish the character designer, loop the first mission, fix the enemy motion AI (also what was talked about concerning the ranged and melee AI triggers helped) fix the spawning, and oh yeah that chat system will be crucial. then theres finishing the tutorial zone.. so much to do.

[Biz]

also gotta redo the hip rigging, finish the character designer, loop the first mission, fix the enemy motion AI (also what was talked about concerning the ranged and melee AI triggers helped) fix the spawning, and oh yeah that chat system will be crucial. then theres finishing the tutorial zone.. so much to do.

And you have to unveil your scientific model about dark matter and gravity that will change everything

[Ankhammon]

A fully depreciated asset?

ewww...

Now I have this terrible vision of some bean counter setting a 25 yr depreciation cycle for NCsoft and that being the only reason we can't get the game back. 



Lordy! Accounting jokes?!?!?
I have been working at HR Block too long.

[LaughingAlex]

I keep wanting to eventually buy dark souls.  Why?  Because the game is challenging while also making the AI also follow a consistency of rules.  It doesnt bend the rules all the time to try and make itself harder.  No invisible walls for example.  But i think its also cause the game was designed with that philosophy from the ground up.  And its a design philosophy i look forward to in city of titans.  While i get a feeling it might have an agro cap, odds are if done right it'd be very silly to try when your enemies have nearly the same stats as you do.

City of heroes didn't have that from the getgo, sadly.  Which is why you had things like agro cap.  It was made in a time when ai was still limited, with smarter AI still somewhat limited to shooters.  And moments when you did have smart AI, it was often accidental.  Even FEARs AI was more accidental; its efforts to out flank you and catch you from behind was merely from it seeking cover with which to destroy you.

Or descents AI and its tendency to also like to shoot you in your back.  It would hide until your ship was turned away before popping out to fire a homing missile.  It was very smart, but still limited and accidentally smart.  CoH didnt have that.  The ai was more the "charge, charge, keep charging" ai.  It only fled when it was so outmatched that it had to, and it ignored other mobs fleeing without some script or another.

Course today ai can be better coded, but back then artificial brilliance that often brings challenging not punishing gameplay was a rarity and accidental.

[CrimsonCapacitor]

Re:Aggro cap...

I joined the game late, right after I16 launched, so I only heard (herd?) about the zone herding exploits.

That said, I always assumed the aggro cap actually required MORE skill than not having it.

Let's say for the sake of discussion, that you're a tank teamed with a squishy of some flavor, let's say Blaster from I23.  You have the mob size cranked up, and you leap into a pile and throw a taunt, run around a corner, grab another group, and herd them into a corner for the blaster to AoE to dea... er, arrest.  But you've now exceeded the aggro cap, so not everyone is focused on the tank. 

So rather than the blaster having to sit there and just "pew, pew, pew" the mob to death, they have to deal with thugs in their face wanting to do bodily harm to that toon.  The tank can't just sit there and take a smoke break while the blaster blasts away.  The tank needs to now worry about getting the aggro off the blaster and back onto the tank.

The aggro cap makes (made?) the game less simple and actually more fun for everyone, at least to me.  Everyone now had a part to play and players couldn't just take a tank, wait for them to pull all the mobs on a floor to a spot and have the ranged character take them out.  Rinse, repeat, hit the 'vator.

[Solitaire]

If they can do it no reason why the people behind getting CoH back can't, articles like this give me hope that one day we will have something to cheer about.

http://massivelyop.com/2016/02/03/darkfall-reboot-is-a-go-starts-audit-of-vanilla-code/

[Codewalker]

So rather than the blaster having to sit there and just "pew, pew, pew" the mob to death, they have to deal with thugs in their face wanting to do bodily harm to that toon.  The tank can't just sit there and take a smoke break while the blaster blasts away.  The tank needs to now worry about getting the aggro off the blaster and back onto the tank.

Or the blaster needs to be smart and defeat the weaker enemies as quickly as possible, whether they're the ones attacking them or not. A nuke to take out minions is ideal if available, otherwise take out the ones with the lowest health first. Once the number of enemies falls back down below the tanker's aggro cap, then they're no longer as big of a threat.

But yet, herding a whole floor is yawnworthy and not fun for anyone except the tanker (who probably really couldn't handle all that aggro if they were all in range and attacking anyway).

If I were hypothetically designing the system from scratch, I would implement it something like this:

Aggro cap number the same as the live game. Enemies beyond the cap will prefer anyone else on their threat list. However, if you are solo and there is no one else on the threat list, they won't just stand there like idiots, but will attack an overaggroed target in "range preferred" AI mode. Most importantly, when attacking a target that is overaggroed like that, they will not follow it out of range, but will let the rest of their buddies take care of it.

That prevents exploiting the aggro cap in situations like AE ambush farms, while still preventing mass herding and keeping the team behavior.

It's also conceptually valid. If 16 of your friends are already beating up on one guy, are you going to try to push your way past them to add one punch, or are you going to go after the other guy standing in the back?

[Baaleos]

While talking about agro and AI.
I would have liked the AI to intelligently target players who were vulnerable.
Eg: If a player is at 2% health, one hit away from death - give them slightly more priority/threat, opposed to the chunky tank who still has 95% health.

This could potentially make it more challenging, since it would decrease team members active, opposed to having entire team active, but at varying hp %.

A team member at 5% health, is just as dangerous as a team member at 95% (with exception to AT passives like scourge)
So, it makes sense for the AI to prioritize removal or destruction of the weaker elements of the party, in order to achieve a decrease in opponents.

Keeping it % based, means it shouldn't penalize against level gaps.
Eg: A level 41 player won't be targeted more than a level 44 player. (Unless they actually do get injured more before hand)


Maybe even a swarm AI type.
When a Player is held - swarm over that target.
Imagine how difficult it would be if you suddenly got held, and shields all toggled off. Then within a second all the enemies re-target you.

[LaughingAlex]

While talking about agro and AI.
I would have liked the AI to intelligently target players who were vulnerable.
Eg: If a player is at 2% health, one hit away from death - give them slightly more priority/threat, opposed to the chunky tank who still has 95% health.

This could potentially make it more challenging, since it would decrease team members active, opposed to having entire team active, but at varying hp %.

A team member at 5% health, is just as dangerous as a team member at 95% (with exception to AT passives like scourge)
So, it makes sense for the AI to prioritize removal or destruction of the weaker elements of the party, in order to achieve a decrease in opponents.

Keeping it % based, means it shouldn't penalize against level gaps.
Eg: A level 41 player won't be targeted more than a level 44 player. (Unless they actually do get injured more before hand)


Maybe even a swarm AI type.
When a Player is held - swarm over that target.
Imagine how difficult it would be if you suddenly got held, and shields all toggled off. Then within a second all the enemies re-target you.

I like this line of thinking, making the ai smart rather than deliberately designed around out dated trinity play.  Actually i loom at games like total war at times.  The ai is smart enough to try to flank you, making it important to sometimes keep reserve units or trick it into exposing a flank.  It guards its flanks when it can, to.

But smart ai often goes for weaker points to.

I think that ai stupidity in mmos is due to deliberately designing with holy trinity in mind, though, rather than letting trinity only happen because of bad faults in the ai and then fixing those faults.  But i also feel many players fear a smart ai.  But thats cause those players never played and outwitted a smart ai, cause they just never saw smart ai in the first place.  Id often hear "it'd be impossible to play against" when reality is, they whether they realise it or not, just dont want any challenge.

They want games to be mindless and just use the same thing over and over again.  Course to me, that isnt a game.  Because of how predictable it gets.  Because many games, specifically mmos, are made that way.  Hell even non mmorpgs i cant help but notice how the ai in some games is actually rather brain dead.  From the very stupid companions in fallout 4 spoiling a stealth players approach to terrarias very simple "just walk towards the players virtical position regardless of the hazards", ai has taken a nose dive in games.

[Shibboleth]

I am thinking it is less players who fear a smart AI than game company executives who want people to play their game and thereby make the game development and support profitable.

[LaughingAlex]

I also notice how many players tend to go by a checklist of "must have this" rather than actually relying more on strategy.  They never slow down to ask themselves or others what people can do and what their enemy mobs can do.  The whole reason AI to many mmorpg players has to be bad is because they don't want to have to apply any level of strategy to anything.  Even though every other genre actually has some level of strategy to it(including the games mmorpg players claim have none, especially the older shooters).

They never ask
"Whats everyones full capabilities?"  This can be answered very quickly, mind you.
"What are our enemies capabilities?" Many players assume every enemy is the same, and often ask for this in games.

More in depth:
"What's our teams biggest strengths?"
"What's our teams biggest weaknesses?"
"What are our enemies strengths?"
"What are our enemies weaknesses?"

Instead, what many players do, is they have a checklist.  You could even see this in bad teams in CoH, but you'd see this anywhere you went in an mmorpg when teams were forming:
Tank check.
Healer check.
Damage dealers check!

Notice I did not include question marks in there.  Or even quotations.  Because they are not asking questions, they are running a routine checklist.  Which smart opponents can generally beat.  This is what separates pve mmo players from many other players of other genres(at least, good games anyways).  Players in games from classic shooters to strategy games like the total war series have to, in order to do well, grasp the situation, rather than forming a rigid plan all the time.

Players who do well in games consistently all the time are always asking questions while playing, subconciously.  The ones who struggle tend to just think "there has to be a secret way to always win that i can form a checklist based on it!".

What ultimately draws me into games is when I see how differing approaches could be taken to how to win the game.  Even games where it's all combat, if there is some variety in how your strategy plays, there can be variety in them if the game focuses more on whether the player simply forms a checklist vs analyzing what an opponent can do besides just half knowing what you can do.

[LaughingAlex]

I am thinking it is less players who fear a smart AI than game company executives who want people to play their game and thereby make the game development and support profitable.

You'd be surprised how many players I met who literally feared a smart AI.  And even more those who had a mindset of "I'm going with this plan no matter what!".  Which to me is a good way of saying "I don't want to have to think".  I think you have a point about game company executives who want the "EVERYBODY WINS!" philosophy to be in play, but the making the game approachable by merely making it easy isn't really the best way to get those numbers.

Think of the games that actually are hard yet still sell.  They sell because they are able to teach players to actually ask questions, even snap questions, and make snap decisions.

[ivanhedgehog]

Or the blaster needs to be smart and defeat the weaker enemies as quickly as possible, whether they're the ones attacking them or not. A nuke to take out minions is ideal if available, otherwise take out the ones with the lowest health first. Once the number of enemies falls back down below the tanker's aggro cap, then they're no longer as big of a threat.

But yet, herding a whole floor is yawnworthy and not fun for anyone except the tanker (who probably really couldn't handle all that aggro if they were all in range and attacking anyway).

If I were hypothetically designing the system from scratch, I would implement it something like this:

Aggro cap number the same as the live game. Enemies beyond the cap will prefer anyone else on their threat list. However, if you are solo and there is no one else on the threat list, they won't just stand there like idiots, but will attack an overaggroed target in "range preferred" AI mode. Most importantly, when attacking a target that is overaggroed like that, they will not follow it out of range, but will let the rest of their buddies take care of it.

That prevents exploiting the aggro cap in situations like AE ambush farms, while still preventing mass herding and keeping the team behavior.

It's also conceptually valid. If 16 of your friends are already beating up on one guy, are you going to try to push your way past them to add one punch, or are you going to go after the other guy standing in the back?

the question is, is this a game that a large number of people would pay to play? probably not. Most people dont want to have a second job once they come home and log on. wildstar made a game that few people actually wanted and now most of their dev staff is laid off. COH had the ability to change difficulty which worked very well. you still had people trying to play at way over their ability just to maximize xp gains. many groups spent way more time running back from the hospital at +4x8 than they should have. +2x6 or 8 would have been better returns in the long run. many gamers are convinced that they want a bleedingly frustrating game until they actually try playing it and realize that it just annoying, not fun.

[Arcana]

Or the blaster needs to be smart and defeat the weaker enemies as quickly as possible, whether they're the ones attacking them or not. A nuke to take out minions is ideal if available, otherwise take out the ones with the lowest health first. Once the number of enemies falls back down below the tanker's aggro cap, then they're no longer as big of a threat.

But yet, herding a whole floor is yawnworthy and not fun for anyone except the tanker (who probably really couldn't handle all that aggro if they were all in range and attacking anyway).

If I were hypothetically designing the system from scratch, I would implement it something like this:

Aggro cap number the same as the live game. Enemies beyond the cap will prefer anyone else on their threat list. However, if you are solo and there is no one else on the threat list, they won't just stand there like idiots, but will attack an overaggroed target in "range preferred" AI mode. Most importantly, when attacking a target that is overaggroed like that, they will not follow it out of range, but will let the rest of their buddies take care of it.

That prevents exploiting the aggro cap in situations like AE ambush farms, while still preventing mass herding and keeping the team behavior.

It's also conceptually valid. If 16 of your friends are already beating up on one guy, are you going to try to push your way past them to add one punch, or are you going to go after the other guy standing in the back?

When I thought about this problem, I thought about making the NPCs more (pseudorandomly) random in their behavior the more aggro you acquired.  The idea was that if a significant component of the difficulty level of the game was how predictable the NPCs were, then perhaps we could modulate predictability to increase difficulty.  Three minions would act very predictably.  Twenty would surround you doing the chicken dance.  Good luck trying to herd them.

But whenever I thought about these issues, my thoughts kept returning to the question of what kind of game was City of Heroes supposed to be.  Is it valid, for example, to believe that being attacked by twenty things and winning was part of the charm of the game, at least for players capable of reaching those levels of power?  I keep thinking paradoxically that it is valid, but it is also game-breaking.  And I keep thinking that the way out of that paradox is to decouple rewards from situational combat.  If we really wanted a City of Heroes where you could be that superhero or superheroine that could take on an army, then the game needed to be designed to allow that scenario, make it even more available to more players, and correctly reward that scenario with a reasonable amount of in-game rewards.

I keep thinking that if we rethought the whole "risk/reward/time/gameplay" equation, a lot of problems get simplified or eliminated.  Suppose we just eliminate rewards for minions.  Period.  No more XP, inf, or drops for them.  We can then put as many of them as we want in a mission, and we can make them as strong or as weak as we want, because they aren't exploitable anymore.  We can let casual players with crap builds still have the feeling of fighting off a room full of minions if the difficulty is downscaled enough.  Put those rewards into the bosses and mission completions.  Stalkers can stealth past those worthless minions, tanks can herd them, blasters can vaporize them, and the game doesn't care.   Simply adjust the leveling curve to match.

Now, the players who all claimed that herding was fun get to indulge themselves all they want - if they want to herd minions.  But what happens when they try to herd LTs or bosses or higher?  Well, rather than use aggro caps we give those higher ranked foes anti-herding powers.  Consider what happens if Lts obeyed Codewalker's aggro rules above with a twist: if they are aggroed but the player is at the cap, instead of shooting from range they sometimes go into a special stance where they get more powerful.  You'd implement this with special powers that they would use in this circumstance that was conceptually valid.  Think DE eminator.  Imagine if Malta Lts above the aggro cap started deploying special power boosters around the player like tiny versions of the Recluse towers.  Maybe some Lts would switch to very long recharge but very high powered sniper shots. 

The idea would be that it would be dangerous to herd up higher ranked foes and just leave them alone to think up mischief.  If you want to herd up fifty Freakshow including Lts and Bosses, be prepared for them to set up something wonderful to take you out.  You'd practically be creating your own zone event.  Survive that, and you deserve the XP.  And yes, it did occur to me that herding up the clockwork should cause the excess clockwork to try to build the Paladin.

Of course, I'm nuts, so my solution to the problem of herding is banana-hat crazy.  But I'm also very serious.