EI Mod and my opinions

[Tartantyco]

 

 

 

A simple fix for the "lone tank" syndrome, and one I've suggested before, is to add penalties for each enemy attacking a character that exceeds their engagement limit. So, if your tank has an engagement limit of 3, he'll get penalized if 4 enemies are attacking him, and it will increase exponentially as more enemies attack. This would be in addition to the current flanking system.

My Eder has 4 engagement limit and the secondary melee easily takes care of what is left. This is not a solution.

 

Why should your tanks not be able to deal with the amount of enemies that their engagement limit allows them? I mean, if your tank has an engagement limit of 5, then him being in combat with 5 enemies shouldn't be an issue. The issue with tanks is not them tanking a lot of enemies, it's them tanking all the enemies. It also makes mob enemies more dangerous.

 

I have yet to have a tough fight that needed more than Eder to tank 4 melee and secondary melee to tank up 2 more. Some enemies are also ranged while others teleport and ignore tanks anyways.

 

...And that's not a problem. Go back to page 11 or 12, where Hiro posted a video where his lone tank is taking on 6 enemies at once. Those are the problem areas. If your tank has an engagement limit of 4 and there are 4 enemies total, him standing alone and taking them all on is not an issue. He'll get the flanked penalty, but that's about it. Lone tanks usually just means the party has to rest more often, nonetheless.

 

Tanks tanking is not a problem.

[MReed]

 

 

PrimeJunta, on 13 Apr 2015 - 3:58 PM, said:

 

 

So how about we discuss the various ways the encounters could be made more difficult and varied, preferably within the general parameters of the P:E design?

 

Well, that's my problem. Maybe i lack imagination, but i cannot think Obsidian could do something to improve the gameplay, because the design parameters are the problem.

Well, they should have way less copy paste encounters, and preferable cut the overal combat in the game to half the amound it has. But the actual gameplay would be just as boring, and the same tactic would work on everything.

It's not an accident that the only interesting spells in the game are the realy OP ones, that singlehandely change the flow of battle.

In order for the game to get interesting, it would require a system redesign, abandon the no hard counter policy, and completely remake the itemization (my biggest disapointment with the game currently)

 

 

You start an encounter. Suddenly enemies spawn on a random chance and approach you from behind. That area you thought was cleared you missed a creature (cuz POTD) and now there's an Ogre bearing down on your squishies. Or a party of hardcore looters. Or something like that.

 

Do the same with reinforcements spawning "behind" when X% of foes drop dead.

 

Um, yeah.  I'm sorry, if you believe that DA2 encounter design is a model that should be followed (because this is exactly how DA2 encounters worked), then...  I have a hard time taking anything that you suggest seriously.

[Luckmann]

So comparing a scripted AI cycle of abilities compared to the common AI routines used for all enemies in a game is somehow comparable now?

But.. a scripted AI cycle of abilities is all it is. Yes, PoE:s is arguably more advanced (arguably) but it's the exact same thing. You learn how they behave, and you beat them based on that merit.

 

So of course it's comparable.

[tinderbox]

 

You start an encounter. Suddenly enemies spawn on a random chance and approach you from behind. That area you thought was cleared you missed a creature (cuz POTD) and now there's an Ogre bearing down on your squishies. Or a party of hardcore looters. Or something like that.

 

Do the same with reinforcements spawning "behind" when X% of foes drop dead.

 

[...]

Random enemies spawning at random without any rhyme or reason is the worst kind of lazy game design, all the way up there with inflated numbers. Having enemies pop up from nowhere in the middle of a battle would have me shutting down the game in sheer disgust.

 

It works in a limited capacity when it makes sense, but the way you describe it, it just sounds like DA2:s nonsensical and random teleportation of enemies into battle.

 

 

Well, let me amend that to, more encounters where you get flank attacked. By which I your party is assaulted from more than one direction at the same time, forcing you to make different deployment choices. This should happen more often in dungeons particularly.

[PrimeJunta]

In the end, really, people will always complain about AI. That's how it is, and always shall be until we eventually make actual Artificial Intelligence that can think and act like a human, complete with errors in judgement and acting of course. But I don't see that happening in my lifetime.

 

As an aside, in a real-time game I think it ought to be pretty trivial to devise an AI that beats a human every time in a roughly fair fight. Simply because a computer can react and compute so much faster than a human, and would possess all the data about an encounter. So it'd always be able to hit you first with a debilitating CC attack, always be able to move out of the AoE of the spell you're casting, always perfectly position its AoE's, and so on and so forth. That would not be fun at all.

 

In that kind of situation, you'd have to have all the fights unfairly stacked to your advantage. That would not be much fun.

 

(The stealth system in P:E is this kind of unfair IMO. Other than scripted setpieces, you will always get the drop on the enemy. That's not fun.)

Edited April 13, 2015 by PrimeJunta

[Zwiebelchen]

 

This, however, I strongly disagree with, at least if I'm reading it properly.  If you are saying that the game should use a "cutscene ambush" (where the game overrides my preference for how my party should be positioned prior to the beginning of a combat encounter) to force the player to place armor on someone other than the tank, then...  Well, I totally disagree.  The relative lack of cutscene ambushes is one of the things that PoE gets right, in my book.

 

It would be desirable in my opinion to:

 

1) Improve the foe AI so that it would be rare for only one character to be targeted by melee attacks -- this has to be done first, or nothing else in this list will make the slightest bit of difference.

2) Nerf the mechanics that penalize offensive potential when defensive potential is increased -- while such a mechanic is absolutely necessary (because otherwise, everyone will be max offense and max defense at the same time), the current mechanic (recovery time penalty) is so enormous that it totally overwhelms all other considerations.  A character that tried to "strike a balanced between offense and defense" will be inferior -- not enough defense (due the way DR and deflection works) to significantly reduce incoming damage, but a dramatic reduction in damage output (due to recovery time penalties).

3) Eliminate (or nerf to the point where it might as well be eliminated) engagement -- the primary effect of the current "strong engagement" mechanic is to penalize character concepts who need to go into melee combat and then withdraw.  Note that such character types are already penalized vs a "straight ranged" or "straight melee" character types, as they require far more micro-management (or skill, if you want to look at it that way) to be used effectively.  As it is, the current mechanics encourage creating "melee only" and "ranged only" character concepts exclusively.

4) Add foes that are not vulnerable to conventional attacks and / or attack the party in unconventional ways.  The only example of an unconventional attack in PoE is foes with domination-type abilities (especially Frampiers, who additionally prefer to target back-line characters with their abilities).

 

All of the above was noted in the backer beta (by various folks, not just myself) for the past 9 months, so...  For backer beta folks, it seems fairly obvious that Obsidian believes that feedback like the above is not something that they are interested in implementing in the game.  That makes it difficult to make constructive posts...

 

 

You indeed read that wrong. What I had in mind was, that when we reduce the gap between tanks and non-tanks and allow classes that are equipped with appropriate gear (for a classy example: a mace & shield mail-equipped priest) to offtank properly (at certain obvious penalties like not being able to cast quite as good anymore), then we could effectively nerf tank defense (and on the same time buff their offense) without making the game neccesarily harder.

 

The idea is to get rid of the hardcore MMO roles and go back to the idea of defined classes not neccesarily being pigeonholed in one purpose. This could ideally allow a tank to hold 2-3 enemies at best and allows rogues and everyone with a plate or mail armor to offtank quite well without sacrificing too much damage.

 

Basicly, make the tankyness of a character tied to armor and equipment selection, not to attributes and an arbitrary base value on class choice.

A cleric in D&D was exactly as tanky as a warrior except for the max constitution bonus. Even a mage had the same defense scores as a warrior when equipped with armor. The difference is: a warrior can hit hard while taking hits, whereas the cleric basicly hits nothing and has to deal with spell interruption. This is imho how it should be and the reason why there even is a concentration score. Remove the unneccesary class-bound deflection bonuses and put everything on equipment (it's okay to keep some deflection on attributes, as long as there's some kind of diminishing returns active here).

 

This is why I don't regard PoE as broken beyond repair: the mechanics for all this to work are already there. We have concentration, we have interrupt, wie have splitted defenses. It's just the numbers that require tweaking.

Edited April 13, 2015 by Zwiebelchen

[El Zoido]

Some enemies seem to like avoiding your tanks (e.g. Fampyrs for some reason always run around my tanks and right at my squishies) and I think that those fights are generally pretty interesting. Suddenly you need to decide whether you a) suck it up and hope that your squishies survive long enough for your fighters to mop up the other enemies or your other characters to kill them or b) disengage your fighters to attack the fampyr, with the danger of making even more enemies slip past unless you hold them down somehow.

Fampyrs also charm your characters, which can pose some risk as well (although some other enemies are better at it).

 

I would generally reduce the efficiency of tanks a bit by reducing health and defenses, as well as giving more enemies the ability to break up the standard patterns a bit (as the flanking and charming fampyrs sometimes manage).

[Malekith]

PrimeJunta, on 13 Apr 2015 - 5:05 PM, said:

 

Malekith, on 13 Apr 2015 - 4:55 PM, said:

Well, if you crank up the resistances to the point of resempling hard counters then yes, you can call it a numbers issue, not a design one. But then it defeats the whole point of not having hard counters in the first place.

 

I think the deal with "no hard counters" wasn't really that. The impression I get is that Josh really dislikes win-or-die RNG effects. So it's more like the other way around. Umber Hulks are horrid until you discover Chaotic Commands; Kangaxx is incredibly scary until you find out about

Spell Immunity: Abjuration

after which he becomes trivial, that sort of thing. Having enemies with high (=near-hard) defenses is different. 

Nothing in Kangaxx's encounter is base on RNG though. The whole point of hard counters is forcing you to change strategy and MIX it up, as your usual tactic is 100% doomed to fail.

Fighting Kangaxx straight up was futile, no matter your level.

But if you put some thought to it, there were 10 different ways to win that battle. Same with Umber Hulks. That is good design in my book, and i would want PoE to apply more of it.

Arcane Vail for example should offer complete protection to physical damage (or 95% reduction if you prefer) in order for enemy mages not going down like flies like they do curently.

Then you as the Player would have more interesting choices. Waste your mages spells in order to counter the enemy mages? Shoot them with bullets (guns bypass AV as per lore) wasting your first voleys on the mages, allowing the melee opposition to close on your gunners?  Ignore the mages in order to deal with the rest, accepting the spells mages will dish on you?

The whole game system should be redesigned in such a way.

[Psychevore]

As an aside, in a real-time game I think it ought to be pretty trivial to devise an AI that beats a human every time in a roughly fair fight.

In the game Offworld Trading Company (really cool game btw, not an RPG, just some economic simulator.. or something ), they had to purposely dumb down the AI for humans to even stand a chance.

 

And for that reason, the hardest difficulty is called "AI test" or something, as you simply cannot beat the AI.

Edited April 13, 2015 by Psychevore

[Sensuki]

(The stealth system in P:E is this kind of unfair IMO. Other than scripted setpieces, you will always get the drop on the enemy. That's not fun.)

Agreed.

 

We all know you're not using it in this thread to denote a person using rest a lot in IE

Funny because that's the exact reason I'm using it, aside with semi-accusing people defending rest-spamming of rest spamming themselves. Personally I don't care that people do it, but I do care that people complain about the combat not being fun because they did it.

 

 

 

Rest spammer? I guess.

Responding to the argument by ignoring it and instead insult the player. Class A discussion method to win them. Except not.

 

And while you probably don't care since it's not about rest to you anyway, I don't rest much in either PoE nor the IE games. Yes, I've had dozens upon dozens of battles in PoE with some of my members having Major Fatigue.

You got a problem with that? Hmmm?

 

Nah, you fed me the line for that one. Just had to do it.

 

I've had a few battles in Pillars of Eternity with characters that have Major Fatigue as well, but I'm finding that I clear a dungeon (or whatever) and then travel back to the Stronghold/Defiance Bay/Town/whatever and everyone has Major Fatigue from travelling instead .. so I just rest.

 

The Temple of Eothas and the Megadungeon are about the only places in the game where you can get a little thin. Raedric's Hold too I suppose depending on how early in the game you go there.

 

In the end, really, people will always complain about AI. That's how it is, and always shall be until we eventually make actual Artificial Intelligence that can think and act like a human, complete with errors in judgement and acting of course. But I don't see that happening in my lifetime.

I'm not complaining about AI, people are complaining about me taking advantage of the targeting AI because I remember what it does. Personally I think the targeting AI in Icewind Dale: Heart of Winter is very robust.

 

One of the differences with the Pillars of Eternity AI and the Infinity Engine AI for ability use is that Pillars of Eternity uses semi-random ability use instead of a script, through the use of different weights for different actions and adjusting of weights when certain conditions are met. It's kind of a shame that this doesn't really mean much though because regardless of what abilities the enemies seem to be using / doing - I can use the same set of abilities and win, rarely ever having to react to the actual ability/spell the enemy is using.

Edited April 13, 2015 by Sensuki

[PrimeJunta]

@Malekith Agree about Arcane Veil. It's currently a waste. Short-duration near-immunity to damage would be much better. Moreso if they restored the original idea of having firearms (and only firearms) penetrate it, as you also point out. 

 

But... again, you could do all this just by adjusting the numbers. Whether Josh wants to is a different matter, and I don't see much point speculating about that.

[archangel979]

 

 

 

 

A simple fix for the "lone tank" syndrome, and one I've suggested before, is to add penalties for each enemy attacking a character that exceeds their engagement limit. So, if your tank has an engagement limit of 3, he'll get penalized if 4 enemies are attacking him, and it will increase exponentially as more enemies attack. This would be in addition to the current flanking system.

My Eder has 4 engagement limit and the secondary melee easily takes care of what is left. This is not a solution.

 

 

Why should your tanks not be able to deal with the amount of enemies that their engagement limit allows them? I mean, if your tank has an engagement limit of 5, then him being in combat with 5 enemies shouldn't be an issue. The issue with tanks is not them tanking a lot of enemies, it's them tanking all the enemies. It also makes mob enemies more dangerous.

 

I have yet to have a tough fight that needed more than Eder to tank 4 melee and secondary melee to tank up 2 more. Some enemies are also ranged while others teleport and ignore tanks anyways.

 

 

...And that's not a problem. Go back to page 11 or 12, where Hiro posted a video where his lone tank is taking on 6 enemies at once. Those are the problem areas. If your tank has an engagement limit of 4 and there are 4 enemies total, him standing alone and taking them all on is not an issue. He'll get the flanked penalty, but that's about it. Lone tanks usually just means the party has to rest more often, nonetheless.

 

Tanks tanking is not a problem.

 

Numbers are not the issue, the issue is that melee enemies are not a danger. In my case Eder tanks as many as he can and my monk waits to see who passes by and tanks those. It still makes the combat super easy because Eder can easily tank 3-4 and my Monk can tank 1-2 which is usually all the melee that is thrown at you.

[tinderbox]

 

Um, yeah.  I'm sorry, if you believe that DA2 encounter design is a model that should be followed (because this is exactly how DA2 encounters worked), then...  I have a hard time taking anything that you suggest seriously.

 

 

I've never played a Dragon Age game in my life.

[PrimeJunta]

Yeah it's surprising how Edér is as good as immune to status effects as well so much of the time. It ought to be trivial to add challenge to tanking simply by making enemies target the tank's weaker defenses with some nasty status effects. Stun, Paralyze, Petrify, and so on. Would immediately neutralize tank-and-spank in the encounters where it's used.

[MReed]

 

In the end, really, people will always complain about AI. That's how it is, and always shall be until we eventually make actual Artificial Intelligence that can think and act like a human, complete with errors in judgement and acting of course. But I don't see that happening in my lifetime.

 

As an aside, in a real-time game I think it ought to be pretty trivial to devise an AI that beats a human every time in a roughly fair fight. Simply because a computer can react and compute so much faster than a human, and would possess all the data about an encounter. So it'd always be able to hit you with a debilitating CC attack, always be able to move out of the AoE of the spell you're casting, always perfectly position its AoE's, and so on and so forth. That would not be fun at all.

 

In that kind of situation, you'd have to have all the fights unfairly stacked to your advantage. That would not be much fun.

 

(The stealth system in P:E is this kind of unfair IMO. Other than scripted setpieces, you will always get the drop on the enemy. That's not fun.)

 

 

Honestly, I don't think this is true in most games.  Literal shooters, yes, but that's because the dominate variable at play in who wins is "How fast / accurately can you point the gun at the target" -- other factors only come into play when the accuracy / ROF variables are mostly the same (e.g. PvP).

 

A game like PoE is an immense problem from an AI point of view, and making a fair AI -- meaning it plays by the same rules as the player -- is an extraordinarily difficult task.  Now, if you want to make an obviously cheating AI, that is indeed easily.  Just setup the game so all RNG rolls are either 100 or 0, whichever is more favorable for the AI.  But that, as you say, wouldn't be a very interesting game...

[archangel979]

Yeah it's surprising how Edér is as good as immune to status effects as well so much of the time. It ought to be trivial to add challenge to tanking simply by making enemies target the tank's weaker defenses with some nasty status effects. Stun, Paralyze, Petrify, and so on. Would immediately neutralize tank-and-spank in the encounters where it's used.

That is almost impossible as if I am facing casters my first spell from Durance is +15 to will and +30 to concentration on whole group

Other spells attack his deflection or fortitude that are already very good.

[Sensuki]

Basicly, make the tankyness of a character tied to armor and equipment selection, not to attributes and an arbitrary base value on class choice.

A cleric in D&D was exactly as tanky as a warrior except for the max constitution bonus. Even a mage had the same defense scores as a warrior when equipped with armor. The difference is: a warrior can hit hard while taking hits, whereas the cleric basicly hits nothing and has to deal with spell interruption. This is imho how it should be and the reason why there even is a concentration score. Remove the unneccesary class-bound deflection bonuses and put everything on equipment (it's okay to keep some deflection on attributes, as long as there's some kind of diminishing returns active here).

That's what I liked about the AD&D AC system - as long as you had a low AC (usually through armor) you could tank.

[Malekith]

PrimeJunta, on 13 Apr 2015 - 5:26 PM, said:PrimeJunta, on 13 Apr 2015 - 5:26 PM, said:

@Malekith Agree about Arcane Veil. It's currently a waste. Short-duration near-immunity to damage would be much better. Moreso if they restored the original idea of having firearms (and only firearms) penetrate it, as you also point out. 

 

But... again, you could do all this just by adjusting the numbers. Whether Josh wants to is a different matter, and I don't see much point speculating about that.

That's why i say that it's a design problem and Obsidian can't fix it.

You "numbers tweaking"  would require such amound of tweaking that would effectively mean abandoning Josh' ideas.

So the only hope for this game is the modding community.

Well, let's hope the next kickstarter to have a completely different designer team.

Edited April 13, 2015 by Malekith

[tinderbox]

Yeah it's surprising how Edér is as good as immune to status effects as well so much of the time. It ought to be trivial to add challenge to tanking simply by making enemies target the tank's weaker defenses with some nasty status effects. Stun, Paralyze, Petrify, and so on. Would immediately neutralize tank-and-spank in the encounters where it's used.

 

Poison. Why is the wood beetle poison now so ineffective? It's like a bee sting compared to how it used to be.

[Sensuki]

Yeah it is, haha. Poison DoT is still kinda annoying though but it's not the OH **** SUPPRESS it used to be.

[Luckmann]

@Malekith Agree about Arcane Veil. It's currently a waste. Short-duration near-immunity to damage would be much better. Moreso if they restored the original idea of having firearms (and only firearms) penetrate it, as you also point out. 

 

But... again, you could do all this just by adjusting the numbers. Whether Josh wants to is a different matter, and I don't see much point speculating about that.

I don't think so. For example, I don't think there's even a mechanic for specifically resisting... anything. So it's not just numbers adjustment (far from it, I believe).

 

Let's take the oft-cited problem of Slicken vs. drakes. Slicken targets Reflex, unless I'm completely off. Fail, you go Prone.

 

You can bump up the drake's Reflex, but you can't give him resistance specifically vs. Prone.

 

To further complicate matters, we'd probably be fine with making drakes go Prone with Knockdown, right? But not Slicken. So it's not just vs. Prone, but suddenly vs. a specific form of spell or ability.

 

The same goes for Malekith's example. I don't even think it is possible to make Firearms freely punch through Arcane Veil, but at the same time allow 95% physical resistance (I would go so far as to say 100%; magical effects on weapons would still allow you to do that super-tiny amount of damage we'd see with 5% anyway).

 

There's no numbers to adjust, the game needs new functions and it needs to use it.

 

On a side note, I can't express how much I like the Arcane Veil idea. It's the exact kind of thing I think is needed. In spades. Massive spades. All over the game.

[MReed]

 

 

Um, yeah.  I'm sorry, if you believe that DA2 encounter design is a model that should be followed (because this is exactly how DA2 encounters worked), then...  I have a hard time taking anything that you suggest seriously.

 

 

I've never played a Dragon Age game in my life.

 

 

Then you should, because DA2 (in particular) works exactly the way that you described -- you should love it.  Being ambushed in this game occurs so frequently that it even comes up in reviews (although almost always in an unfavorable light, for some reason).

[PrimeJunta]

Nothing in Kangaxx's encounter is base on RNG though. The whole point of hard counters is forcing you to change strategy and MIX it up, as your usual tactic is 100% doomed to fail.

Fighting Kangaxx straight up was futile, no matter your level.

 

That's true, and Josh has said somewhere that he explicitly dislikes it. "Combat as puzzle" I think was the term he used. I.e. a hard encounter that becomes trivial once you know the solution. I don't have that strong a dislike for it personally, but I kind of sympathize; IMO "the solution" should ideally be at least a little tricky to apply. Many of BG2's major set pieces really aren't, so the game kind of goes from punishingly hard to easy at any difficulty level.

 

What redeems it though is that the process of going from punishingly hard to easy is really rewarding, and after that point there are scads of ways of gimping yourself to keep things interesting -- and for the encounters that do become rote (Kangaxx for example) there are always more that retain their challenge, or where the challenge changes because your party doesn't have the capabilities it did last time. Playing with Keldorn in the party (=effectively at-will, powerful Dispel Magics) is quite different than playing without him in the party.

 

IMO it's this incredible variety and richness what makes BG2 so awesome, and makes up for the numerous holes and exploits in it.

[Hassat Hunter]

And for me itemization is worse than BG1, to the point i would prefer if PoE abandoned the whole "magic items" consept and just have the same mundane weapons through the whole game.

BG1 was more low level than PoE, but the few Special items felt more special than anything PoE has. Varsona +2, The spider bane, ring of wizardy, gauntlets of dexterity,the +3 gratshord in Durlang's tower.... Varscona is just a +2 sword with added cold damage. But when it's the ONLY +2 sword in the whole game, and the only weapon that has cold damage, it feels Special. in a way that PoE's diablo inspired loot doesn't

Nothing in PoE comes close to these items. And in BG1 you didn't fought dragons for them.

Dragon scales in BG2-> you craft a special set armor, unique to the whole game

Dragon scales in PoE-> enchanting adds +2 to your sh ity armor

I have to agree, the whole "enchantment" thing is IMO the 1st worst thing (2 being trap/lockpick XP) in PoE, totally ruining the joy of finding new weapons or armor.

The only items which can be fun to get are (*shockingly*) those who cannot be enchanted! That should already speak for itself I think.

 

Crafting doesn't really affect my game much (I ignored it) but enchanting actively dragged the game down even if you're not using it.

[Sensuki]

Imagine if the game didn't have DR. Every single balance problem related to damage would easily be able to be solved.