Attribute balance thread

[Odd Hermit]

Everything I said about attributes in the beta I pretty still feel the same about, and my dislike is stronger due to companion builds being stuck with poor point distributions. Now that the game is release it just feels like the dead horse has been resurrected and it's a good time to talk(/rant) about them again. So...

 

PERCEPTION/RESOLVE

 

Interrupts are not a strong enough effect for damage-classes to invest in these. And a caster with a melee enemy on them is in trouble anyway, I rarely am trying to stand and cast while engaged and if I am it seems like Dexterity is a better bet to get casts off faster.

 

Deflection also doesn't help non-tank build's survivability enough either. Deflection has almost the opposite of diminish returns right now, it sucks at all lower values and gets increasingly good as you get higher and higher values up until you can barely be hit. So if you're not stacking it, it feels like a small-medium investment in it is just a waste. This is also why many +deflection wizard spells kinda suck(plus costing spell/rest being an issue when they last only one encounter).

 

CONSTITUTION

 

Durable/higher deflection class/builds have enough endurance/health without investing in it. Fragile ones have low base endurance/health which means they get less out of it. And for the most part, even a 15-30% boost in endurance won't stop a squishy from dropping like a stone.

 

MIGHT

 

High might just feels like it gives a lot more than just a 21-30% boost to damage for me. I assume it's because it can be the different between dealing little-to-no damage to higher DR enemies, and actually getting substantial damage through. And as get higher and higher damage as you get better gear/level up, it just keeps modifying bigger numbers. It's a very no-brainer attribute to dump points into currently.

 

 

The overall result right of these issues right now is that min/maxed builds are generally better at their jobs. There are almost no builds that can really benefit from more balanced attribute spreads.

 

Potential solutions:

 

Buffing interrupts. I don't really like how interrupts work/how random they are, but it would make perception and resolve potentially matter more.

 

Putting accuracy on perception again. Not sure it wouldn't become just another thing to dump con/resolve to stack another damage attribute but at least it'd make the NPCs stuck with it better.

 

Giving constitution a substantial flat /level endurance/health bonus. I'm not sure this alone would make it worth it, but it'd be better.

 

Adding concentration to constitution, maybe in addition to the above.

 

Tweaking deflection to have diminishing returns(or tweak the thresholds for miss/graze/hit/crit?), so smaller investments are worth more.

 

Putting +duration on Resolve. Int is a bit of a no brainer caster attribute currently. However, with only AoE it becomes bad for many other classes so I'm not sure it'd hold up with only +AoE.

 

Class-specific benefits for different attributes. This is the easiest/safest way to tweak things, but it probably conflicts with their goal of a simple/understandable system(though I'm not sure we have that currently anyway...).

 

 

 

 

 

 

 

 

[illathid]

Everything I said about attributes in the beta I pretty still feel the same about, and my dislike is stronger due to companion builds being stuck with poor point distributions. Now that the game is release it just feels like the dead horse has been resurrected and it's a good time to talk(/rant) about them again. So...

 

PERCEPTION/RESOLVE

 

Interrupts are not a strong enough effect for damage-classes to invest in these. And a caster with a melee enemy on them is in trouble anyway, I rarely am trying to stand and cast while engaged and if I am it seems like Dexterity is a better bet to get casts off faster.

 

Deflection also doesn't help non-tank build's survivability enough either. Deflection has almost the opposite of diminish returns right now, it sucks at all lower values and gets increasingly good as you get higher and higher values up until you can barely be hit. So if you're not stacking it, it feels like a small-medium investment in it is just a waste. This is also why many +deflection wizard spells kinda suck(plus costing spell/rest being an issue when they last only one encounter).

 

CONSTITUTION

 

Durable/higher deflection class/builds have enough endurance/health without investing in it. Fragile ones have low base endurance/health which means they get less out of it. And for the most part, even a 15-30% boost in endurance won't stop a squishy from dropping like a stone.

 

MIGHT

 

High might just feels like it gives a lot more than just a 21-30% boost to damage for me. I assume it's because it can be the different between dealing little-to-no damage to higher DR enemies, and actually getting substantial damage through. And as get higher and higher damage as you get better gear/level up, it just keeps modifying bigger numbers. It's a very no-brainer attribute to dump points into currently.

 

 

The overall result right of these issues right now is that min/maxed builds are generally better at their jobs. There are almost no builds that can really benefit from more balanced attribute spreads.

 

Potential solutions:

 

 

Buffing interrupts. I don't really like how interrupts work/how random they are, but it would make perception and resolve potentially matter more.

 

Putting accuracy on perception again. Not sure it wouldn't become just another thing to dump con/resolve to stack another damage attribute but at least it'd make the NPCs stuck with it better.

 

Giving constitution a substantial flat /level endurance/health bonus. I'm not sure this alone would make it worth it, but it'd be better.

 

Adding concentration to constitution, maybe in addition to the above.

 

Tweaking deflection to have diminishing returns(or tweak the thresholds for miss/graze/hit/crit?), so smaller investments are worth more.

 

Putting +duration on Resolve. Int is a bit of a no brainer caster attribute currently. However, with only AoE it becomes bad for many other classes so I'm not sure it'd hold up with only +AoE.

i

 

Class-specific benefits for different attributes. This is the easiest/safest way to tweak things, but it probably conflicts with their goal of a simple/understandable system(though I'm not sure we have that currently anyway...).

I agree with most of this, except putting accuracy back into the stat picture. Even with relatively low gains, accuracy is just too important.

[Odd Hermit]

 

I agree with most of this, except putting accuracy back into the stat picture. Even with relatively low gains, accuracy is just too important.

 

 

I agree, just figured I'd add it to the list as it's been suggested by others.

 

I do wish we had more ways to gain accuracy for spells - there're lots of +accuracy sources for weapons.

 

But of course with the AoE control spells available it could cause balance issues.

[HozzM]

CON for me is the one that really stands out, as in, no builds really need it and for most builds you can take points from it and not really notice.  To me it is clearly the weakest Attribute.

 

I don't think making it better at what it does is the answer though...I think maybe they should take deflection from PER and give it to CON and put Accuracy back to PER?  Or something else.

Agree with you also that Might feels like the strongest attribute, at least the most useful across all classes and builds.

[illathid]

CON for me is the one that really stands out, as in, no builds really need it and for most builds you can take points from it and not really notice.  To me it is clearly the weakest Attribute.

 

I don't think making it better at what it does is the answer though...I think maybe they should take deflection from PER and give it to CON and put Accuracy back to PER?  Or something else.

 

Agree with you also that Might feels like the strongest attribute, at least the most useful across all classes and builds.

I like the idea I saw in another thread: each point of Con lets you ignore a percentage of armor action penalty. So characters with really high con would be able to wear armor with little or no penalty.

[termokanden]

Agree for the most part, except about accuracy. I hate having low-accuracy characters, so for me it would just end up being sort of a tax.

 

As you say, int also needs to do more than just AoE. For example, if duration is out of the picture, rogues for example would now have just about zero use for int ever. Now it's a dump stat for single-target builds.

 

This attribute system was always what worried me most about PoE, but now that I've actually tried the game, I feel that even though it's not perfect, it's still a lot better than AD&D was.

Edited March 29, 2015 by termokanden

[Voss]

I still advocate the idea of having con reduce armor recovery times, as it gives both armor and con a reason to exist.  

 

I do think a (small) accuracy buff needs to go on perception.  The companions and monsters seem quite obviously built with this in mind, as the up to 20 point gap in accuracy in places shows.  It could even be 1 acc per point above 10, and it would be something.

 

Resolve... eh.  Resolve (Deflection2) might as well not exist.  I have no idea what design decision prompted this stat, other than 'we gots to have 6' 

[termokanden]

I like the con idea a lot. As it is now, people are playing characters with 3 con and it doesn't really make that much of a difference.

[sleepofthejust]

The attempt to streamline the attributes have only ended up making things more opaque. We all undersand what the developers intend for each statistic just based on the tooltip. However, the relative importance of different attributes can only be determined by having played the game. You can't know at character creation how important your endurance will be compared to your damage or deflection.

They intended to get rid of dump stats, but instead what happened is that some stats (might) are so good that everything else may as well be dump stats. It wouldn't be that hard to fix; maybe critical chance is based on dex and crit damage on might. Maybe Constitution gives a % boost to armor DR (cause stronger people can wear bigger armor). Maybe one of the stats could add blanket status effect chance (resolve? constitution makes sense for knockdowns). Maybe perception increases your armor penetration (cause you can see the weak points).

I agree that any stat increasing accuracy is probably too good unless you redesigned all the enemies and encounters. Also might's damage increase needs to get nerfed...sorry guys.

Just to be clear, these are minor gripes, the writing is great.

Edited March 29, 2015 by sleepofthejust

[Odd Hermit]

 

CON for me is the one that really stands out, as in, no builds really need it and for most builds you can take points from it and not really notice.  To me it is clearly the weakest Attribute.

 

I don't think making it better at what it does is the answer though...I think maybe they should take deflection from PER and give it to CON and put Accuracy back to PER?  Or something else.

 

Agree with you also that Might feels like the strongest attribute, at least the most useful across all classes and builds.

I like the idea I saw in another thread: each point of Con lets you ignore a percentage of armor action penalty. So characters with really high con would be able to wear armor with little or no penalty.

 

 

Yeah I like this idea. I suggested something similar awhile back too:

 

Toughness: Requires 12 Constitution: +3 stam, health, and fortitude/level, -10% Recovery Penalty from armor, -30% damage taken from criticals

 

http://forums.obsidian.net/topic/71046-attribute-based-talent-suggestions/

[Sensuki]

I'd just like to say "I told you so" regarding Perception and Resolve to anyone involved in making these attributes reading this.

[Mungri]

Perception and resolve are my first dump stats except on front line tanks. Though I even dumped it on my barbarian to max his might / con / int for carnage.

 

I tried putting points into con on chanters and priests but it does too little compared to boosting might, dex and int on them. So just dump perception / resolve, leave con at 10, and max might, dex and int is how I'm playing any AoE class.

[DragonDai]

From what I've seen/experienced, it seems like there are exactly two builds in this game, statwise, unless you are purposefully gimping your character/choosing stats for roleplaying reasons. You either max Resolve and Perception and put the rest into Intelligence, or you max Strength and Intelligence and put the rest into Dexterity. I know there are some niche builds that don't fit those two cases, but I think that about 90% of characters could have 18+/8/18/8/18/8 for non-tanks or 8/8/8/18+/18/18+ for tanks. For the tank, you could theoretically make the case that spreading the 18 in Intelligence over to Might or Constitution or Dexterity might not be a bad idea to keep their defences up, but even then, 10/10/10/18/12/18 isn't REALLY that much different, and I think the bonuses to durations of effects (like knockdown on a fighter tank) is probably superior to a couple extra points in Fortitude or Reflex.

And if you're about to say, "Well, no one is forcing you to min/max." that's true. But why make a suboptimal character? I already put an exemption for roleplaying at the very beginning. Yeah, if you REALLY want to RP a super resolute Barbarian, go for it. The game really excels at making your choices matter in non-combat settings, so it's 100% viable to make non-min/maxed characters for roleplaying purposes. But there are a ton of people who play these sorts of games and couldn't give two craps about RPing a super perceptive wizard. They just want to have fun blowing stuff up with a fireball. And your super perceptive wizard is going to be WAY less good at that than your extremely mighty wizard is. 

And beyond that, even if you're stating a character for RP reasons, if you don't invest points into Might as a rogue, you're gana have a bad time on anything but Easy. Oh, and if you don't invest points into Might as a Barbarian, you're gana have a bad time on anything but Easy. Also, if you don't invest points into Might as a Cipher...well, you get the idea...

I don't know. Maybe it's bias, maybe it's me being stubborn, but I spent a LOT of time since release just playing with the character creation portion of the game, and every character I make comes out with almost identical stats. It's disheartening. Something needs to be done. And, TBH, the fact that the stat system is in the shape it's in at release is, IMO, a pretty poor reflection on Early Access/Kickstarter titles, Obsidian, or possibly both. The game shouldn't have been released with this kind of glaring flaw attached to it. And that's a real shame.

[Bryy]

I like the con idea a lot. As it is now, people are playing characters with 3 con and it doesn't really make that much of a difference.

Wait really?

[Sensuki]

The attributes were better before accuracy was removed. Now it basically just ends up being

 

Tank: RES/PER/CON
DPS: MIG/DEX/INT

 

Caster: INT/MIG/DEX

 

boring.

[TheUsernamelessOne]

I don't know what was discussed during the beta period, but I do know that playing at launch I felt like the game was lying to me about what stats I needed (not literally but still). It also doesn't rely on your attributes enough for roleplaying purposes; most of the time the special options unlocked by high stats don't do anything particularly useful, even for roleplay purposes. It's not like you get any big reward from combat anyway, why not include more ways to shortcut encounters through optional dialogue or those nifty, criminally underutilized scripted encounters? Sort of like how Wisdom and Charisma were the most important stats in Planescape: Torment even though they contributed practically nothing to combat.

 

I guess that's getting away from a simple rebalancing patch. Still, though. You're telling me Ciphers rely on Might and not Perception? The class who are stereotyped as scientists and detectives? Tanks get more out of spending points on Resolve and Perception than they do on Constitution? It's absurd. It's not even balanced.

[RunningWolf]

What i dont like is how unflexible current builds are. I feel like game is pushing me towars strict tank/glass cannon builds with really bad characters inbetween, on Hard that is. Nor BG, nor nwn2 had this problem. And i dont see HOW can you roleplay on hard with ironman.

[Gromnir]

I'd just like to say "I told you so" regarding Perception and Resolve to anyone involved in making these attributes reading this.

funny, we were gonna same... though am expecting we got complete different pov on the issue.

 

oh, and am complete disagreeing that the attributes were better before accuracy were removed.  accuracy was/is too important to link it so heavily to a single attribute.  have one attribute that were effectively essential were not only running counter to the design philosophy o' poe, but it were stoopid.  

 

the poe attributes ain't near as significant to overall character success as folks are imagining.  whether that is a good or bad thing is debatable, but we believe it were intended.  

 

"What i dont like is how unflexible current builds are. I feel like game is pushing me towars strict tank/glass cannon builds with really bad characters inbetween, on Hard that is. Nor BG, nor nwn2 had this problem. And i dont see HOW can you roleplay on hard with ironman."

 

this is just not true.  

 

our first character (hard mode) is proving to be a very impressive melee combatant, as well as giving us many dialogue options that would otherwise not have been available to us.

 

m 10

c 10

d 10

p 18

i 14

r 16

 

for a dual-wielding rogue, the above is horrible attributes as far as any mechanics monkey is concerned, and yet we ain't gnashing teeth and pulling hair in frustration.  our character is effective in and outta combat.  

 

from a pure mechanics perspective, talents and abilities quickly become far more important that starting attributes.  from a rp perspective... well, as we said, we ain't feeling handicapped.  quite contrary.  

 

 

 

HA! Good Fun!

Edited March 29, 2015 by Gromnir

[Cantousent]

I've only recently been able to play the game. I will say, from my perspective, that PotD on Iron Man mode is essentially reducing the game to the mechanical aspects of the CRPG. The point of iron man is to prove that you can create the most efficient 'build,' not enjoy the writing the most. I might or might not do one of the PotD iron man runs, but I, like everyone else should, take for granted it's not the mode meant for the inner RPG child. It's a purely bragging rights endeavor meant to prove that you've got some game theory chops.

 

I don't get a chance to read the posts as thoroughly as I like, so where I've repeated someone else's view, take it as an affirmation.

[gkathellar]

I'd actually much rather put +AoE on Resolve than Deflection, if only because that gives Paladins and Barbarians a reason to invest in Resolve over Intelligence.

 

Really, though, the relationship between attribute mechanics and their fluff is kinda problematically tenuous right now.

[Tequilaflavor]

CON for me is the one that really stands out, as in, no builds really need it and for most builds you can take points from it and not really notice.  To me it is clearly the weakest Attribute.

 

I don't think making it better at what it does is the answer though...I think maybe they should take deflection from PER and give it to CON and put Accuracy back to PER?  Or something else.

Agree with you also that Might feels like the strongest attribute, at least the most useful across all classes and builds.

 

I actually find CON fairly useful (in moderation) on my "frontline offtank" PC. It's a stat that should only be taken by "frontline grunts" which is why it is %-based (which isn't neccessarily a poor design choice, imho).

 

If CON got changed to a flat endurance/health bonus, it should be balanced around how much Fighters and Barbarians currently get out of it - which might potentially make it too good on Wizards and the like, depending on how the encounters get balanced. Seems unlikely, but it's something to keep in mind.

 

RES is, in my opinion, a decent secondary tank stat. It's not super-amazing, but it's in line with the majority of the stats in usefulness and I would not want my frontline fighters without at least some RES for that reason.

 

Now PER . . . really stinks, as far as I can tell right now. Whenever I spend points in it, it just feels bad. Interrupts aren't reliable or good enough to invest in, the high deflection bonus is nice (deflection IS useful in moderation because it works on a "1% to not get hit" system) but PER feels like a tertiary tank stat at best and you usually don't have the points to invest in it.

 

Not sure what to change about it . . . maybe make interrupts better ?

[gkathellar]

If you actually examine how much Endurance a point of Con grants you, you will be somewhat less impressed with it.

 

Resolve and Perception are much more important to a tank - it's just that they have little or no importance to everybody else.

[Endrosz]

If you dump both Mig and Con on a tank (like I did), your Fort save will be abysmal and you'll be owned by any mob that has a special ability targeting Fort. Like, wolf packs with Knockdown, to give an early example. And even if the difference isn't much in Endurance, in PotD, it still matters quite a bit. You need to survive the initial spike in dmg. Neither Knockdown nor Yell is worth a high Int (and I'm not familiar with monks).

 

Having said that, I'll just add my vote to the piles:

 

-- The armor penalty decrease idea is very good. Unique, fitting, strong enough that is matters, helps the weakest attrib. *applauds*

-- Do not add back accuracy, it's far too important. There is more than enough spread with Talents and enchants and accurate/inaccurate weapons and weapon style.

[DragonDai]

If you dump both Mig and Con on a tank (like I did), your Fort save will be abysmal and you'll be owned by any mob that has a special ability targeting Fort. Like, wolf packs with Knockdown, to give an early example. And even if the difference isn't much in Endurance, in PotD, it still matters quite a bit. You need to survive the initial spike in dmg. Neither Knockdown nor Yell is worth a high Int (and I'm not familiar with monks).

 

Here's the thing. Sure and 8/8/8/18+/18/18+ tank is going to suck vs fort and be not so great vs reflex. But all you have to do is go 10/10/8/18+/14/18+ and suddenly, you're tank is 100% fine. Get a few buff spells going if there is a particular combat that is troublesome, but most of the time, you'll be 100% fine.

 

The issue here isn't that super min-maxing will leave **** in your armor (literally and figuratively in this case lol), but that there is basically no real diversity in builds. Is the character a tank? Yes? Okay, max out P and R and get a healthy dose of the rest, based on your preferences, it really doesn't matter. Is your character not a tank? Okay max out I and M for casters or M and D for not casters and get a healthy dose of the rest, based on your preferences, it really doesn't matter. So doing it "right" (in terms of combat, not RPing/out of combat stuff), means having two stats as high as they can go and the rest are basically meaningless.

 

On the other hand, if you do it "wrong" (again, based solely on combat, totally ignoring the OOC/RPing stuff), you are making a blatantly suboptimal character who is just straight up worse at his job with no upside (outside of RPing/OOC uses for stats) that a "correctly" built character. In other words, there aren't any meaningful choices when it comes to attributes in combat. Either you spend your attributes well (and by "well" I mean 2 stats maxed and the rest irrelevant) or you do less good in combat than you could (likely WAY less good). There are no meaningful choices there.

 

Now, I keep placing a caveat for Roleplaying/Out Of Combat. It's totally reasonable to say "I'm going to give my rogue an 18 perception because I want to RP a very perceptive rogue and I want to see the out of combat options that come with having a high perception." That's 100% legit. But sadly, it comes at a cost. Your rogue WILL be worse (and likely a LOT worse) than a rogue that doesn't do this and instead maxes Might and Dex and has a good number of points in Int. Now that's a trade off I'm sure a lot of players will be happy to make. But the thing is, it isn't a tradeoff the players should have to make in the first place. It's a bad choice. It means that if you want to RP your character via their attributes, you might have to make a suboptimal character for the combat portions of the game. That's super lame, and it's certainly not interesting.

 

Sadly, I am not sure there is a good fix beyond a pretty serious rework of the attributes. I think there are a lot of good ideas in this thread (and others like it). I like the idea of Con doing more so that it's actually useful. I like the idea of Perception and Resolve having some more/better offensive effects than they have now. And I'd like to see damage matter for tanking (so that a tank can't just dump Might and be a meat shield). But I'm not sure I'm best qualified to make suggestions towards improving that. Hopefully stuff will change, but, at best, I am skeptical it will.

[Tequilaflavor]

If you actually examine how much Endurance a point of Con grants you, you will be somewhat less impressed with it.

 

Resolve and Perception are much more important to a tank - it's just that they have little or no importance to everybody else.

 

Well, I just imagined it would work as the tooltip suggests, giving me a couple of % bonus on my max endurance and health per point, which is pretty useful for a fighter who has a very high endurance to begin with (thus benefiting more from %-based boni).

 

I think, considering that it gives endurance and health, it's pretty much in line with most other stats (my fighter sometimes has to worry more about health than endurance).