How to Fix the Attribute Design in Pillars of Eternity

[Namutree]

Is it just me but I can't seem to find the urge to playtest the beta until these changes are implemented...

I feel the same way.

[Sensuki]

Yeah I think decoupling AoE and Duration is not good because yeah sure okay, you're making it harder to choose for some classes, but I think that will have a negative effect on the balance.

 

even with AoE scaling and AoE increase enemy only, I think Durations are way more valuable for pretty much every class except *maybe* the Wizard as they could then dump fireballs ontop of their party. But the thing is if you smartly position your party for buffs, and smartly aim offensive AoEs and debuffs, you can get away without much AoE increase at all (if any) from the lowest value, at least that's what I've found anyway. I don't think the double the square area is that big.

 

You can make up for that deficit with smart play. But you can't make up for a duration deficit.

Edited September 13, 2014 by Sensuki

[Shallow]

For AoE, what about giving every AoE spell a base AoE, then allowing you to scale the AoE up or down in size based on your INT, this would fix the "only occasionally useful"ness of the AoE bonus, without going down the road Sawyer suggested, which to me just feels sort of off.

Personally I'm I'd like the decoupling duration and AoE, I like me some customization, however it wouldn't be an easy change to implement, as you'd have to retain/recreate balance whilst doing so, perhaps if duration was moved into resolve spell range could be added to INT.

Edited September 13, 2014 by Shallow

[Matt516]

Yeah I think decoupling AoE and Duration is not good because yeah sure okay, you're making it harder to choose for some classes, but I think that will have a negative effect on the balance.

 

even with AoE scaling and AoE increase enemy only, I think Durations are way more valuable for pretty much every class except *maybe* the Wizard as they could then dump fireballs ontop of their party. But the thing is if you smartly position your party for buffs, and smartly aim offensive AoEs and debuffs, you can get away without much AoE increase at all (if any) from the lowest value, at least that's what I've found anyway. I don't think the double the square area is that big.

 

You can make up for that deficit with smart play. But you can't make up for a duration deficit.

 

Agreed. Even with Josh's selective fringe idea (which would buff AoE considerably), AoE is still much, much weaker than Duration simply by virtue of the fact that benefiting from increased AoE is very situational, whereas you always benefit from increased Duration.

 

Not that AoE needs to be buffed to be on a level with Duration (if that's even possible) - it's fine how it is or slightly buffed by Josh's idea. It's power level should just be taken into consideration when pairing it with other attributes. It needs to be paired with a "strong" attribute (like Accuracy, Deflection, Duration) rather than a "weak" attribute (like Interrupt or Concentration).

[Utukka]

Why not play around with allowing for a small increase to single target spell range? I imagine that could make it easily compete with duration and could open up even more ways to play.

[Seari]

Well I just tried making a backline wizard with the attribute changes proposed by Sawyer( INT=def+AoE, RES=duration+conc), and I just can't see this working. The stats are too spread out for a caster, as Sensuki and Matt said.

 

I'll try to experiment a bit more.

 

edit: If interrupt is removed from attributes, then there really is something missing.

edit2: Unless there's some other mechanic added, Sensuki's and Matt's changes are the best as far as I can see.

Edited September 13, 2014 by Seari

[Matt516]

Well I just tried making a backline wizard with the attribute changes proposed by Sawyer( INT=def+AoE, RES=duration+conc), and I just can't see this working. The stats are too spread out for a caster, as Sensuki and Matt said.

 

I'll try to experiment a bit more.

 

edit: If interrupt is removed from attributes, then there really is something missing.

edit2: Unless there's some other mechanic added, Sensuki's and Matt's changes are the best as far as I can see.

 

Yeah - as I've said, splitting Duration and AoE wouldn't be the worst thing ever, but it's certainly not optimal IMO. I think there have been a lot of good suggestions in the thread so far, and I'm loving the quality of discussion and variety of good, constrictive feedback and debate going on. Josh, if you see this, you should know that anytime you drop into a forum discussion, even for an hour or so, it provides a huge morale boost to us.

 

That said, I still prefer the original design presented in our paper for the most part. Granted, I'm biased - but still. I haven't really seen any alternatives that do a better job of achieving those two design goals while still making sense from a RP perspective.

[Ganrich]

I think the main Idea for splitting duration and AoE was:

 

Might - damage and healing

Dexterity - Attack speed

Con - health/stamina

Perception - Accuracy and interrupt

Intellect - AoE and Deflection

Resolve- concentration and duration

 

This is how I understood Josh's statement on the subject.

 

So in this example AoE would indeed be combined with a strong effect (Deflection). While Duration which is pretty powerful makes up for Concentration being weaker.

 

Many classes that have limited AoE can still use Intellect for the Deflection: Fighters, Rogues, Rangers, ect. Where Rangers can make use of the AoE if they have the appropriate animal companion. The Archer Rogue may not need Intellect here, but a melee rogue definitely could. This coupling of effects on intellect also, as has been pointed out, a very Combat Expertise vibe.

 

Where Resolve governing concentration and duration adds something similar to the above. Any class with any duration abilities benefit: Casters, rogues, fighters, barbarians, rangers, etc. Where the added concentration gives the casters the options of stepping into melee range more.

 

The other benefit to this is that you get more variety in builds with casters. Using the Wizard:

 

Wizard 1 - high might and high intellect makes a good ranged nuker. This needs to be coupled with Josh's AoE improvements or the mouse wheel suggestion. While you could be safer if you get attention from enemies with the higher Deflection.

Wizard 2 - high might and high resolve would make a solid muscle wizard. The AoE from intellect would be cumbersome at times in melee engagement with coned attacks since increased AOE could lead to a liability here. This makes you hit hard, your buffs last longer, and you have good concentration.

 

Replace might with perception and you have an interrupter build in either of these scenarios. Replace might with Dexterity and you have a solid quick casting build. You could even have a high constitution, high intellect, and decent Resolve and make a very tanky Wizard.

 

I am assuming we can get 2 high attributes with averaging the other stats around 10ish from the point buy, but the current system allows for 1 or 2 high (16+), a moderately high (14ish), and the rest around average. I am sure this will be tweaked. Just using it since it is what I have seen in beta.

 

I think you are right, Matt. Decoupling duration and AoE has to leave AoE with a powerful sister effect. Which IMHO this does by pairing deflection and AoE. It spruces up Resolve as well. It keeps those 2 attributes useful to most classes in a plethora of builds and can be very problematic to dump them. The other 4 attributes should be fine. As long as the Attack Speed increase from Dexterity is enough to be useful.

[Ganrich]

Well I just tried making a backline wizard with the attribute changes proposed by Sawyer( INT=def+AoE, RES=duration+conc), and I just can't see this working. The stats are too spread out for a caster, as Sensuki and Matt said.

 

I'll try to experiment a bit more.

 

edit: If interrupt is removed from attributes, then there really is something missing.

edit2: Unless there's some other mechanic added, Sensuki's and Matt's changes are the best as far as I can see.

Really? Moderate to high Might, Moderate to high intellect, and moderate to high Resolve. What is high or moderate is up to you and dependent on flavor.

 

Constitution can be lower... as can Perception (since most spells get a bonus to accuracy) and Dexterity if you don't wear heavy armor.

[Seari]

Wizard 1 - high might and high intellect makes a good ranged nuker. This needs to be coupled with Josh's AoE improvements or the mouse wheel suggestion. While you could be safer if you get attention from enemies with the higher Deflection.

You have to sacrifice duration, action speed and accuracy then. Which is why I said that stats are too spread out. Muscle wizard works tho.

 

edit: I have no idea how essential these 3 are, so no idea how to choose between them.

Edited September 13, 2014 by Seari

[Doppelschwert]

Well I just tried making a backline wizard with the attribute changes proposed by Sawyer( INT=def+AoE, RES=duration+conc), and I just can't see this working. The stats are too spread out for a caster, as Sensuki and Matt said.

 

If you feel like you have to spread out your stats, then I'd say that's mission complete, considering the design intent. The more you feel you have to sacrifice something to gain something, the more useful is every stat overall.

 

By the way, regarding the debate concerning how AoE should scale (mousewheel / no friendly fire at the edges):

There is stuff you can do with the latter that you can not with the mousewheel variant, like correctly targeting the enemies (X) if they are positioned around one of your guys (O) like this:

 

X

OX

X

 

I think that situation should be fairly common if someone tanks and there are a lot of enemies. For the most part, I also agree with ganrich.

[Ganrich]

 

Wizard 1 - high might and high intellect makes a good ranged nuker. This needs to be coupled with Josh's AoE improvements or the mouse wheel suggestion. While you could be safer if you get attention from enemies with the higher Deflection.

 

You have to sacrifice duration, action speed and accuracy then. Which is why I said that stats are too spread out. Muscle wizard works tho.

 

edit: I have no idea how essential these 3 are, so no idea how to choose between them.

You want an all around caster. That is fine. You can build an AoE nuker or a more duration nuker here. I don't think that they should be given both, but that is my opinion. Most the AoEs are pretty generous as is anyway.

 

Action speed I can give you with little debate.

 

A bonus to Accuracy is given on many offensive spells. I think it is a +10 on most, and that makes it less valuable for the Wizard anyway. However, you could forgo Might for Perception, and gain a little less damage, increased accuracy, + interrupt capability.

Edited September 13, 2014 by Ganrich

[Marceror]

Well I just tried making a backline wizard with the attribute changes proposed by Sawyer( INT=def+AoE, RES=duration+conc), and I just can't see this working. The stats are too spread out for a caster, as Sensuki and Matt said.

 

I've got one word for you: unlimited attribute rolls

[Ganrich]

 

If you feel like you have to spread out your stats, then I'd say that's mission complete, considering the design intent. The more you feel you have to sacrifice something to gain something, the more useful is every stat overall.

 

This is how I feel. I think if you put too many "must haves" under any attribute for a single class then you have defeated this idea with the attribute system. Each attribute needs to be equally desirable for most classes, but not so desirable that you can easily see the benefit of pumping 2 stats vs spreading a few points around. I think this attribute layout has each attribute being desirable for most classes, has a lot of variety, and doesn't make cookie cutter casters to be too easy to obtain. If each attribute has value then each attribute will require choices when placing points into it. Where you can still have a nuker, but a couple different flavors of it. You can have the AoE machine, the Duration machine, the interrupter, a fast caster, etc. I don't believe that there should be a single way to build a ranged nuker, but multiple differently-abled nuker types.

[Sensuki]

The reason I don't like the Durations/Concentration and Deflection/AoE is because it makes the attribute system a lot more unfair on certain classes than others.

 

for DPSers I would say the best attributes would be Might, followed by Perception

for Tanks - Constitution and (whatever governs Deflection)

for Damage based casters - Might and either durations or AoE, depending on class

and for support casters - Durations and AoEs (if on same attribute) and Action Speed

 

now this makes Action Speed harder to work into any build at all by default and it makes support based casters focus on two attributes just to get the two things that they already got from one attribute in the original system. I think it creates less choice for them, because you'll want to get both the AoEs and the Durations, and have less attribute points to spend on other things.

 

It also makes other things easier to get for other classes. Hopefully it's not just me that cares about this balance either.

 

So that's what I don't like about it. It's not terrible, but I don't think it's as balanced as our version, YMMV.

 

OH, yeah and of course - it makes Accuracy and Interrupt automatically overpower Concentration on it's own because of the combined benefit they give to Interrupt. The ACC/INT and DEF/CON thing was what made me go to all the effort to make the paper in the first place. They could change the Interrupt mechanics or something, but I don't know if that would be a good thing.

Edited September 13, 2014 by Sensuki

[Doppelschwert]

I agree, ganrich. In particular, I feel like the attributes that josh described can be balanced against each other simply by tuning the numbers. Duration too strong compared to AoE? Tone it down. Every class needs either AoE or Deflection, except maybe ranged rogue, but then you pay for it with decreased ranged accuracy you have to take care off. I think it's fine if the numbers are tweaked enough.

[Sensuki]

Would be pretty dumpable on Ranger and possibly Rogue too.

Edited September 13, 2014 by Sensuki

[Doppelschwert]

 it makes support based casters focus on two attributes just to get the two things that they already got from one attribute in the original system. I think it creates less choice for them, because you'll want to get both the AoEs and the Durations, and have less attribute points to spend on other things.

 

You're right, if you want your ultimate buff machine now, you pay for it, but you are free to distribute your stats in a different way. Buffs are huge in the game, so I think your claim is more about your expectation regarding the availability of strong buffing. In the original build, buffs where cheap, now they are not. IMO, it's fine if it costs a lot to buff, so I think the change is a good idea. For example, the priest now has an actual choice if he wants to focus on healing or wants to focus on buffing, which both was a no brainer in the original build.

 

I don't understand where you see choices going away, when in fact you have more choices to make with joshs system.

[Sensuki]

Buffs need to be nerfed and so does healing spells, they are all ridiculously OP atm - mostly because the default durations are too long, and the heal/sec is too high.

 

Durations and Concentration would be dumpable on quite a few classes IMO. The classes with the least durations would quite literally laugh at this attribute and just grab Deflection/AoE instead. Barbarians spring to mind - Intellect was already good on Barbarians, but now it's like LOL THX GUISE free Deflection! They don't need extra durations or concentration. There's other classes like that as well.

 

If the goal of the system is not balance then fine, whatever. But I think that our solution is the more balanced of the two. The Durations/Concentration and Deflection/AoEs makes it harder for some classes to get the main benefits they want and much easier for other classes to get theirs.

 

I guess it depends on what classes you intend to play. I don't have a preference so I'd rather the system be balanced properly.

 

and yeah like I've been saying - Acc/Int together would smash Concentration on it's own. If you raised the Concentration value to compensate then that would trivialize the interrupt system in general.

Edited September 13, 2014 by Sensuki

[archangel979]

Attributes should not be balanced for all possible builds. It should be enough that all classes have some use from each attribute, but every class does not need to have same amount of use from them, we go into boring same classes zone there.

The goal of this design was already accomplished if all classes have some use from each attribute, because compared to IE games that is a big change where classes needed 2-3 attributes only and rest were completely useless.

[Sensuki]

I am talking about balanced choices, not balanced for all possible builds. But I'm not sure who you're talking to exactly, but probably me.

[Ganrich]

Eh, I don't think that casters "should" get both AoE and duration from a single attribute. Why should other classes get to pick and choose from differing benefits but casters should inherently get both AoE and Duration to their plethora of spells?

 

I don't disagree with action speed, but as I said many moons ago in the original debate that increasing attack rate is always tricky because if the benefit is too small it is gimped and if too large it is too potent. It is a fickle thing to balance, and that is exacerbated by the attack speed implementations within the armor system. However, I can still see ample use of it being possible, but you will have to make a sacrifice of AoE, Duration, and/or interrupt to get it. Choices are important here.

 

On the note of Perception vs Concentration I can see your point. You have done more homework than I, and I won't argue that.

 

I just don't like giving casters the benefit of larger AoE and duration where the other classes only see a single benefit from intellect because they don't have 8 spells per spell level to make use of both AoE and Duration.

 

It may not be balanced 100%. I haven't mathed it out, and I don't plan to. I am going by gut feeling. So, I won't continue to debate you on it, but I really feel with Resolve being entirely defensive (with concentration and deflection) and AoE and duration being under 1 attribute you make cookie cutter builds far too easy to obtain. Tank? Max Con and Resolve and toss other points into dex or perception (most likely perception. Nuker? Max Might and intellect left over points go toward dex or resolve. Etc. It is what it is.

[Doppelschwert]

These are fair points, although I think barbarians benefit from having a higher concentration, given that they should probably be surrounded by enemies and be hit a lot.

 

Regarding the durations, that's true. I think that'd be better if there were more abilities with durations available to the classes currently lacking them in general, regardless of whether one uses joshs or your system. However, that could be managed with talents.

 

Don't get me wrong, I like your system as well - I just think there is an even split of classes that are better off in your system and in joshs proposal.

[Sensuki]

Concentration reduces DPS loss, if you want more DPS it's always better to max Might and Accuracy instead and cop the interrupts.

 

They could put it to vote in a KS update or something tbh, or test both internally and see what feels better - or give us two diff versions one after the other and then have us vote on it. Would be nice to test the differences.

Edited September 13, 2014 by Sensuki

[Ganrich]

Would be pretty dumpable on Ranger and possibly Rogue too.

Deflection on intellect fixes it from being a dump on rangers as long as the animal companion sees a deflection bonus for it. They should since they share a health pool with the ranger. The stag companion also has an AoE component, and that makes intellect interesting in this scenario.

 

While a rogue archer may dump intellect to a degree a melee could use the bonus from deflection to decrease incoming damage.

 

These 2 scenarios are dependent on the build you want to have with these 2 classes. You cannot make an attribute valuable to every build, but you can make it valuable to every class in some way.

 

I agree that I am unsure that these benefits are large enough to make them not be dumps, but intellect in its present form isn't necessary for these 2 classes either way. The rogue can lean on another party member for status effects if they dissipate too quickly. While the ranger only benefits from the duration on an ability or 2 anyway and very little from AoE.