Concerns with the proposed Perception change

[HoopleDoople]

We've known for a while now that the devs are likely to change Perception so that it is linked to Accuracy in place of Deflection. There is some promise behind the idea but I'm concerned that it may be outweighed by the negative consequences of such a change. If the devs are indeed committed to such a change I would like them to understand the problems that would likely arise so they can take steps to eliminate or mitigate them.

 

Before I start criticizing I'd like to highlight some of the benefits that we could expect from changing Perception:

 

The Good

• Perception will now be a valuable attribute for everyone instead of just tanks.
• Starting with higher Accuracy will make the first act less difficult, bringing it more in line with the difficulty of the rest of the game
• Tanks will no longer be able to stack Deflection to absurd levels that render them effectively immune to most enemies.
• Constitution will no longer be a nearly universal dump stat as tanks will have to be able to survive additional damage.
• Melee DPS characters will now have a use for each and every attribute. There will still be more optimal choices but at least there will be real consequences for forgoing any particular attribute.

The Bad

• Perception will be a nearly mandatory stat. Even tanks will find Perception fairly desirable. Tanks need Reflex somewhere (and Dexterity probably isn't worth it until you've covered Damage and Accuracy) and the Accuracy is very useful for the initial attack, offensive abilities, and scroll usage.
• Higher maximum accuracy will make the later game even easier than it is now.
• Ranged characters will receive what amounts to almost a pure buff. Their damage and disabling capabilities will be enhanced with the only real cost being lack of Concentration.
• CC spells/abilities will be boosted to potentially game breaking levels, and CC is already the optimal answer in almost every battle or situation. When CRPGs are at their best every battle has its own unique optimal strategy, a locked door requiring a specific key. CC is currently a skeleton key and we risk making it a battering ram.
• Similarly weapons that inflict Stun or Prone on crits will become significantly (more?) overpowered.
• Companions will fall even further behind hired Adventurers. Companion attributes are already sub-optimal without bringing Accuracy into the discussion.
• The Wood Elf and Hearth Orlan are currently top tier choices for Ranged and Melee DPS, respectively. As both these races have bonuses to max Perception, their status will only be enhanced making other choices less competitive.

There are definitely some real benefits to changing Perception but at least as many pitfalls we need to avoid. Quite simply Pillars of Eternity is in need of many balance changes to reach its fullest potential, and tweaking the Attributes only makes this more clear. Let me detail some specific ideas:

 

Mitigating the downsides

• Reduce or remove accuracy bonuses from spells and abilities with CC effects.
• Enhance enemy AI behavior and unit composition to make defensive stats not nearly so irrelevant on Ranged characters.
• Give specific bosses or enemy types immunity to hard CC.
• Apply an internal cooldown to weapons that can inflict stun or prone on crit.
• The option to have enemies scale to level is a good first step to making the game less easy near the end, but specific battles may also need to be enhanced.
• Implement a custom difficulty option (separate sliders for enemy quantity, attributes, and AI intelligence) that will allow players to more accurately obtain the challenge they desire. This should allow for the game to be played at a difficulty beyond PoTD.
• Players who have completed the game should gain the option to build companions upon hiring as they would when hiring an adventurer. Only appearance, sound, and (probably) class would be set in stone.
• Buff the least popular races to give players multiple viable options for different build types. In particular the Nature Godlike and all Dwarves have awful bonus abilities. Humans are also rather lacking in my opinion.

I am quite confident there's more potential upsides, downsides, and fixes I've left out. Please feel free to share your own thoughts on this matter or respond to my particular points.

[cavemandiary]

I think that these are all valid points, but ultimately it boils down to how much accuracy is given per point of perception before we cry wolf.

[Fardragon]

If it's enough to notice, it's enough to cause the problems listed above.

 

If it's not enough to notice, then it's a straight up nerf to the stat.

 

 

Perception used to boost accuracy in the beta. Has the reason that was removed gone away?

Edited June 30, 2015 by Fardragon

[PrimeJunta]

IMO PER isn't the problem stat ATM. It's highly useful for pretty much everyone as interrupting is always desirable. Try a high-PER barb for example: with that and Carnage he really makes life miserable for the enemy frontline.

 

CON is a problem though. It's dumpable for most builds; only monks really need it because they have to take damage to use their specials, and even they can get by with above-average.

[Luckmann]

If it's enough to notice, it's enough to cause the problems listed above.

 

If it's not enough to notice, then it's a straight up nerf to the stat.

 

 

Perception used to boost accuracy in the beta. Has the reason that was removed gone away?

 

Actually, yes. There was a calculation error for combat resolutions in regards to additive/multiplicative stacking, although the exact specifics is something you'd have to ask someone more knowledgeable than me about.

 

Perception used to give +2 Accuracy per point, but it was wildly overpowered because the calculations were off, and when they fixed the math, they also changed the Attribute Bonuses at the same time, meaning that Perception was "double-fixed" as an issue.

 

If Perception gets +Accuracy, it's getting +1 Accuracy per point, and it should be fine, and I've been asking for that since pretty much forever. At the end of the day, we're talking a maximum of a +11 Accuracy, and even so it requires you to sink all your points into it at the expense of other things, and be of a specific race and culture. Accuracy isn't the be-all/end-all that it's sometimes made out to be; it's important, yes, but not to the point of making everything else moot.

 

IMO PER isn't the problem stat ATM. It's highly useful for pretty much everyone as interrupting is always desirable. Try a high-PER barb for example: with that and Carnage he really makes life miserable for the enemy frontline.

 

CON is a problem though. It's dumpable for most builds; only monks really need it because they have to take damage to use their specials, and even they can get by with above-average.

I agree insofar that Perception isn't really a problem, but I think that Perception suffers from a thematic/conceptual issue. I don't think that Interrupt is worthwhile, though, although I'll be honest and say that I haven't tried an interrupt Barbarian build - but if it works (and I have no reason to doubt your word) then it's likely because of how Barbarians work with Carnage - it's hardly something that's really applicable to most other classes or builds, and does little to ultimately save Perception.

 

The overall issue that I see is that the system is lopsided and too easy to min-max, because most attributes are only useful if you are either a Tank or DPS. Tank? Perception/Resolve. DPS? Might/Intellect/Dexterity. Constitution? Never. But that's partly an issue with the armour dichotomy, and won't be resolved with attribute tweaks alone.

 

That's the kind of "balance" issue I see. But from a power-per-points point of view, I agree that Constitution is by far the worst offender, followed by Intellect. Constitution because it's largely useless for everyone - the ones that needs it doesn't benefit from it, and those that benefit, doesn't need it. Intellect because it's a consolidated caster attribute and a no-brainer for all spellcasters (and Barbarians).

Edited June 30, 2015 by Luckmann

[Fardragon]

IMO PER isn't the problem stat ATM. It's highly useful for pretty much everyone as interrupting is always desirable. Try a high-PER barb for example: with that and Carnage he really makes life miserable for the enemy frontline.

 

CON is a problem though. It's dumpable for most builds; only monks really need it because they have to take damage to use their specials, and even they can get by with above-average.

Agreed, Con is the biggest problem, but they should never just look at one stat in isolation - if they are going to change one, they should re-evaluate them all.

 

How about, instead of accuracy, Perception boosting critical hit chance or multiplier? How about Con giving a small DR multiplier, or reducing the speed penalty for armour?

[PrimeJunta]

I really like the idea of CON reducing the armor speed penalty. It'd be thematically appropriate and make it interesting for a larger variety of builds. Even the fabled frontline muscle wizard.

[Luckmann]

I really like the idea of CON reducing the armor speed penalty. It'd be thematically appropriate and make it interesting for a larger variety of builds. Even the fabled frontline muscle wizard.

 

I like the idea too, but if memory serves, I liked the idea of the engagement system, too. I'm just not sure. That being said, my standing suggestion actually is CON: ±1 Endurance, ±3% Endurance, ±2% Armour Recovery Penalty, ±2 Concentration, ±2 Fortitude. It's just that I'm not entirely sure.

 

 

 

IMO PER isn't the problem stat ATM. It's highly useful for pretty much everyone as interrupting is always desirable. Try a high-PER barb for example: with that and Carnage he really makes life miserable for the enemy frontline.

 

CON is a problem though. It's dumpable for most builds; only monks really need it because they have to take damage to use their specials, and even they can get by with above-average.

Agreed, Con is the biggest problem, but they should never just look at one stat in isolation - if they are going to change one, they should re-evaluate them all.

 

How about, instead of accuracy, Perception boosting critical hit chance or multiplier? How about Con giving a small DR multiplier, or reducing the speed penalty for armour?

 

 

I don't see the need to change Perception from +1 Accuracy, really. Adding Critical Chance is the same as adding +Accuracy, and adding Critical Multiplier is too narrow, only benefiting specific builds. The beauty of +1 Accuracy is that it benefits everyone, whether it's the Tank that tries to compensate for his loss of Accuracy as a meatshield, the rogue trying to get more consistent critical hits, or the wizard that wants to bump up his aim (which is hard for wizards, because they don't get the same bonuses other combatants do) at the tradeoff of maybe not being able to boost both +AoE and +Duration (assuming that it's finally split up between Intellect and Resolve again).

 

Nevermind an option to being a Muscle-wizard.

Edited June 30, 2015 by Luckmann

[Sabotin]

My concern with it just that it will make combat faster. I'm enjoying PoTD more than Hard, just because it's slower due to higher enemy defenses.

[Tennisgolfboll]

Great post topic creator.

 

Yes cc is allready overpowered, this change will indeed make it just insane.

 

If all spells get +11 accuracy (white that wends+orlan) then the developers have no clue about balance.

 

However accuracy isnt to high for melee so they can do the PE buff if they also remove 10 accuracy from spells

Edited June 30, 2015 by Tennisgolfboll

[Tennisgolfboll]

But is changes to basic attributes this late really good for a game. Im not so sure.

 

Imagine if bg ee.had strength changed for example, not sure it would be good in the long run.

 

At first is fun and all but after one wonders what is next? Am i playing beta? An mmo?

[player1]

My concern is that instead of 3 tank and 3 DPS stats, we will now have 2 tank and 4 DPS stats.

 

With Con not being that impactful for tank, that leaves tank with lots of excess points to invest in DPS stats.

 

On the other hand, DPS combatants have one extra stat to add to the mix which is useful for all DPS classes.

Edited June 30, 2015 by player1

[Torm51]

My concern is that instead of 3 tank and 3 DPS stats, we will now have 2 tank and 4 DPS stats.

 

With Con not being that impactful for tank, that leaves tank with lots of excess points to invest in DPS stats.

 

On the other hand, DPS combatants have one extra stat to add to the mix which is useful for all DPS classes.

Extra stats in this game will not make you a dpser.  You must spend talent points to make yourself into a competitive dpser.  If a tank invests into dps talents he will not be a very good tank.  Maybe on offtank with ok damage. 

[Luckmann]

Great post topic creator.

 

Yes cc is allready overpowered, this change will indeed make it just insane.

 

If all spells get +11 accuracy (white that wends+orlan) then the developers have no clue about balance.

 

However accuracy isnt to high for melee so they can do the PE buff if they also remove 10 accuracy from spells

That would be absolutely crazy. The overwhelmingly vast majority of spellcasters are not going to be Orlans from The White That Wends. The game needs to get less focused on min-maxing, not more. Even an Orlan spellcaster from The White That Wends (this is the part where we lament the lack of white fur for Orlans and no portraits with white fur) that maxes out Perception to get that +11 will be passing on other things, such as Might, which would've increased the damage of the spells used, Intellect or Resolve (assuming Intellect is no longer a consolidated caster stat), Dexterity (which would be useful if Robes counter as Clothing, instead of an arbitrary form of armour, or if the spellcaster used Clothing), or even Constitution (assuming it's buffed).

 

+10 Accuracy is good, yes, but you don't have infinite points. What're you going to dump?

[Fardragon]

I would prefer to see accuracy with spells, ranged, and melee weapons separated out.

[Luckmann]

I would prefer to see accuracy with spells, ranged, and melee weapons separated out.

 

Technically, they are already separate (at least Melee and Ranged are), just not when it comes to the Attribute Bonuses, and it's (unfortunately) not clarified on the character sheet It's theoretically possible for items (or whatever) to give bonuses only to ranged, or only to melee, or (probably) only to spells. In regards to the Attribute Bonuses, though, I think that it'd be a mistake to split them up. It would only really serve to pigeon-hole the bonuses and add too much weight to individual Attributes depending on whether you're ranged, melee or a spellcaster, since the attributes would still be able to stand on their own in regards to non-melee, non-ranged, or non-spellcasters.

 

Basically, it'd be virtually impossible to balance.

[Fardragon]

I think it's a good thing for your choice of attributes to reflect how you want to play the character.

 

I think the big problem with the PoE atributes is it's adoption of a 3-18 scale, which was originally intended to be rolled randomly on 3d6 to simulate a normal distribution.

 

Now, if there was an option for randomly rolled attributes (no re-rolling or reallocation of points)  they would work better.

Edited July 1, 2015 by Fardragon

[Tennisgolfboll]

 

Great post topic creator.

 

Yes cc is allready overpowered, this change will indeed make it just insane.

 

If all spells get +11 accuracy (white that wends+orlan) then the developers have no clue about balance.

 

However accuracy isnt to high for melee so they can do the PE buff if they also remove 10 accuracy from spells

That would be absolutely crazy. The overwhelmingly vast majority of spellcasters are not going to be Orlans from The White That Wends. The game needs to get less focused on min-maxing, not more. Even an Orlan spellcaster from The White That Wends (this is the part where we lament the lack of white fur for Orlans and no portraits with white fur) that maxes out Perception to get that +11 will be passing on other things, such as Might, which would've increased the damage of the spells used, Intellect or Resolve (assuming Intellect is no longer a consolidated caster stat), Dexterity (which would be useful if Robes counter as Clothing, instead of an arbitrary form of armour, or if the spellcaster used Clothing), or even Constitution (assuming it's buffed).

 

+10 Accuracy is good, yes, but you don't have infinite points. What're you going to dump?

Nah

 

Cc is the most overpowered thing in the game.

 

It does not need a buff. If pe adds to cc then it needs to have something to balance it out.

 

Otherwise they buff the most overpowered thing there is, which makes no sense. That would be crazy

[MadDemiurg]

Removing accuracy from CC spells would be a straight up nerf though  (which I don't think is necessary). This is a tricky situation tbh. Accuracy is very important, so it's good to keep in in check with predictable accuracy for each character at any given level. Which is why I don't like perception change at all - it makes balancing enemy defences more difficult if not impossible.  A character that didn't max Perception should not be a whimp. Likewise, a character that did should not be overpowered. I don't think there's a good solution for this.

[Tennisgolfboll]

I agree with you Mad. But since PE accuracy is coming something needs to be done with cc or it will go from being overpowered to really broken.

[Uni]

I think perception giving +1 accuracy is a good step. That combined with making constitution good via the often suggested armor penalty removing(i like the suggestion of Luckmann in this thread) would make every stat valuable and somewhat break the boring pure tank/pure dps min max -dichotomy that the current attribute system promotes. Ofc constitution could be buffed in other ways but the armor one just feels good, logical and gives constitution its cool niche.

Only having one stat give deflection helps tanks not getting so ridiculous and they get enough reflex from weapon and shield style anyway. I dont think the added accuracy is problematic outside of CC, but some of those spells are already overpowered and should be toned down anyway. We could use some powerfull cc spells having no or +5 accuracy instead of +10 and so on, they are simple to fix numerically.

Internal spell balance as a whole is in a poor state and should be fixed. Casters would be way more interesting without current op no brainers and the never used bad spells. Many interesting spells are underpowered and the game sufffers from that by making encounters more repetitive.

If improving attributes makes those imbalances even more glaring then we might even get the op spells nerfed and worse ones buffed in a more hasty manner. Making good systematic changes will sometimes make current balance problems worse but it is definitely worth such hassle in the long run.

[HoopleDoople]

 

Great post topic creator.

 

Yes cc is allready overpowered, this change will indeed make it just insane.

 

If all spells get +11 accuracy (white that wends+orlan) then the developers have no clue about balance.

 

However accuracy isnt to high for melee so they can do the PE buff if they also remove 10 accuracy from spells

That would be absolutely crazy. The overwhelmingly vast majority of spellcasters are not going to be Orlans from The White That Wends. The game needs to get less focused on min-maxing, not more. Even an Orlan spellcaster from The White That Wends (this is the part where we lament the lack of white fur for Orlans and no portraits with white fur) that maxes out Perception to get that +11 will be passing on other things, such as Might, which would've increased the damage of the spells used, Intellect or Resolve (assuming Intellect is no longer a consolidated caster stat), Dexterity (which would be useful if Robes counter as Clothing, instead of an arbitrary form of armour, or if the spellcaster used Clothing), or even Constitution (assuming it's buffed).

 

+10 Accuracy is good, yes, but you don't have infinite points. What're you going to dump?

 

 

You may not have infinite points but as a long range caster you don't have to make much of a sacrifice to achieve high Perception either.

 

Any ranged character is going to want to be a Wood Elf and probably take The White That Wends as a culture. This gives you Distant Advantage and allows for 20 Perception. Next you dump Resolve and Constitution as they are purely defensive stats. Compared to currently optimized ranged characters you'll be squishier due to a lower deflection score but stronger CC should more than make up for it.

 

This leaves you 52 points to distribute between Might, Dexterity, and Intellect. This leaves you only 2 points short of reaching 18 with each of these attributes. A Ranger or ranged Rogue will probably take the hit on Intellect while casters would probably prefer to sacrifice some Dexterity and/or Might.

 

Removing accuracy from CC spells would be a straight up nerf though  (which I don't think is necessary). This is a tricky situation tbh. Accuracy is very important, so it's good to keep in in check with predictable accuracy for each character at any given level. Which is why I don't like perception change at all - it makes balancing enemy defences more difficult if not impossible.  A character that didn't max Perception should not be a whimp. Likewise, a character that did should not be overpowered. I don't think there's a good solution for this.

 

 

You definitely have a point here. If the Perception change goes through an alternate solution may be to give all enemies a passive +10 defense against CC effects on PoTD only. This would allow casual players to build sub-optimal characters without being punished while min/maxers wouldn't be able to break the game by pumping Perception and spamming CC.

[MadDemiurg]

I'd say ability accuracy overall needs some tweaking for more fun gameplay, like encouraging combo moves.

 

I think accuracy values (from highest to lowest) should be:

• damage spells - since damage spells are limited (unlike basic attacks) they should be less rng dependant
• defence debuffs (-def/ref/fort/will) - many spells currently do smth like lower fortitude when also targeting fortitude. This is flawed, and can be compensated by giving these spells higher accuracy. This would encourage spell combos more as well.
• basic attacks & other debuffs
• hard cc (paralyse, petrify, dominate etc)

(Any accuracy buffs should affect all abilities equally.) However this is a huge change which requires tuning most of the existing abilities. But just slapping a +10 CC resist on everything feels like a crutch.

Edited July 1, 2015 by MadDemiurg

[Tennisgolfboll]

Lots of spells have +5 modifers etc. Atleast the cc ones could be have that removed

Edited July 1, 2015 by Tennisgolfboll

[HoopleDoople]

I'd say ability accuracy overall needs some tweaking for more fun gameplay, like encouraging combo moves.

 

I think accuracy values (from highest to lowest) should be:

• damage spells - since damage spells are limited (unlike basic attacks) they should be less rng dependant
• defence debuffs (-def/ref/fort/will) - many spells currently do smth like lower fortitude when also targeting fortitude. This is flawed, and can be compensated by giving these spells higher accuracy. This would encourage spell combos more as well.
• basic attacks & other debuffs
• hard cc (paralyse, petrify, dominate etc)

(Any accuracy buffs should affect all abilities equally.) However this is a huge change which requires tuning most of the existing abilities. But just slapping a +10 CC resist on everything feels like a crutch.

 

A complete ability/spell tuneup to account for the Perception change and existing imbalances would be wonderful. I just don't think this is a realistic expectation to have, which is a big part of the reason I'm so concerned. The amount of rebalancing needed to properly account for a change to a core mechanic is much higher than we could reasonably expect, even in an expansion. Thus my goal is to brainstorm some simple changes that can minimize the potential damage.