Balancing Light Weapons

[tdphys]

Since we have additive damage calculations now:

(I'm assuming might bonus is additive) 

 

you can look at damage calcs like this:

 

damage = base*( 1 + might + crit*critchance)  

 

Higher base damage means each increase in might gives more damage for heavy (high base) weapons, which makes them the most effective damage dealers.  

 

You could balance this by giving significantly higher accuracy bonuses to light weapons (that's what is happening already, I think) ,  but those are static, you can't pump light weapon damage by picking stats...

 

unless dexterity works,  then you could increase crit rate by increasing the number of attacks per second.

 

So maxing dex over might would be better for light weapons.

 

Maxing might would be better for heavy weapons.

 

Heavy weapons would still be able to increase damage by higher dex, but not as relatively much as the higher critting light weapons.

 

This leaves light weapons specialists extra stat points to target things like maximizing interrupt.

 

So don't write off light weapons until the dex attack speed actually gets implemented.

Edited March 6, 2015 by tdphys

[Sensuki]

Unless Dexterity gives more total attack speed in percentage than Might gives in flat damage, then Might is always going to be better for dealing mmore damage.

 

The faster you attack, the more often your damage is reduced by DR, effectively raising the DR of the unit you are fighting by the bonus of your attack speed.

 

With damage mulitpliers, it might take you several hits just to make the same damage as a single hit from an Estoc. This is a problem with using percentile increases to damage instead of integers and also not balancing weapons and weapon speeds against DR, but instead, balancing them against DR 0.

 

Josh said a while ago that they might tweak the damage ranges, but it never ended up happening (yet). There's still one more patch to go, and then a day 0 patch - so we'll see.

[tdphys]

Unless Dexterity gives more total attack speed in percentage than Might gives in flat damage, then Might is always going to be better for dealing mmore damage.

 

The faster you attack, the more often your damage is reduced by DR, effectively raising the DR of the unit you are fighting by the bonus of your attack speed.

 

With damage mulitpliers, it might take you several hits just to make the same damage as a single hit from an Estoc. This is a problem with using percentile increases to damage instead of integers and also not balancing weapons and weapon speeds against DR, but instead, balancing them against DR 0.

 

Josh said a while ago that they might tweak the damage ranges, but it never ended up happening (yet). There's still one more patch to go, and then a day 0 patch - so we'll see.

 

Well, if your accuracy is high enough that more of those hits are criticals that give  cumulatively higher or equivalent to  damage compared to an Estoc, you might be okay.  I guess what I'm saying, is given percentile base damage amplification, the only route I see mathematically to make light weapons comparable is to cause heavy weapons to crit way less and light ones to crit way more ( and maybe bump their crit modifier relative to heavy weapons).   If you are dependent on making crits the way to maximizing damage, then more hits = more crits =  dex being your main stat to amplify damage as a light weapon wielder.   

If you nerf heavy weapons accuracy a bit, then might becomes the preferred way to maximize that damage, since crits are going to happen less anyways, might as well get the guaranteed damage from buffing might.   

[Sensuki]

Actually that won't fix the problem, because a crit: damage * (1+0.5), is only one multiplier among many. As a Rogue with an enchanted weapon I could have many seven-eight damage multipliers. Less crit damage (which guns already have) does not make much of a difference because all of those multipliers scale with the base damage of the weapon.

 

Here is an example

 

Rogue with Sneak Attack (x1.25), Blinding Strike (x1.25), 21 Might (x1.33), Fine Weapon (x1.15), Crit (x1.5), Damaging 2 Enchantment (x1.3), Two Handed Style (x1.1)

 

Let's say I roll 18. That becomes 18 * (1+0.25+0.25+0.33+0.15+0.5+0.3+0.1) = 51.84 damage. Dropping the crit multiplier to 1.2 only changes that to 48.24 - which isn't much difference.

Edited March 6, 2015 by Sensuki

[tdphys]

Actually that won't fix the problem, because a crit: damage * (1+0.5), is only one multiplier among many. As a Rogue with an enchanted weapon I could have many seven-eight damage multipliers. Less crit damage (which guns already have) does not make much of a difference because all of those multipliers scale with the base damage of the weapon.

 

Here is an example

 

Rogue with Sneak Attack (x1.25), Blinding Strike (x1.25), 21 Might (x1.33), Fine Weapon (x1.15), Crit (x1.5), Damaging 2 Enchantment (x1.3), Two Handed Style (x1.1)

 

Let's say I roll 18. That becomes 18 * (1+0.25+0.25+0.33+0.15+0.5+0.3+0.1) = 51.84 damage. Dropping the crit multiplier to 1.2 only changes that to 48.24 - which isn't much difference.

 

 

Yep,  you're right in the sense that every multiplier gets affected by base damage, so I can't just balance light weapons by buffing one modifier out of a slew of them.

 

The only thing I can think of to rebalance light weapons then is to go back to make the crit modifier multiplicative, leave everything else additive, and really nerf the accuracy of heavy weapons and buff light ones as stated above, or just give light weapons higher multiplicative crit bonuses.

 

Otherwise the math doesn't lie,  base damage rules all and heavy weapons are the way to go.  

 

In other words, if you could do:

 

Base damage of heavy weapon > light weapon

Crit Damage of light weapon > Base damage of heavy weapon.

 

and seriously reduce critical chance of heavy weapons,  I think you could rebalance things, leaving most additive modifiers how they are now, and not completely redoing the system.

 

Maybe you could call it "base critical" , an unchanging critical blow modifier that is multiplicative and tied to weapon type, no talents, no feats etc.

 

So 

BD = base damage

BC = base critical

 

damage = (BD)*(BC)*( 1+ mod1 + mod2 + mod3 )

 

and base critical only applies on a critical hit

Edited March 6, 2015 by tdphys

[Sensuki]

The simplest way to do it is just to alter the base damage of the weapons with a formula so that 1H weapons will fall off after a certain DR. I don't think Josh likes to balance things that way though. Seems to prefer doing things by 'feel' instead of by formulas.

 

However, one thing to take into account is DR bypass. Vulnerable Attack gives you 5 DR bypass at the cost of some attack speed. The attack speed cost is not as much for dual wield as it is for 1H single or 2H style, therefore you can attack really fast and do close to full damage with the right weapons - making dual wielding 1H normal style weapons very strong. It's okay with 1H Fast weapons but it's much better for 1H normal IMO.

 

You can also take Two Weapon Style to completely offset that penalty. Whereas for 2H you have x1.1 damage for Two Handed Style.

 

1H style and fast weapons are still pretty bad, but dual wielding two handed normal weapons is awesome IMO.

Edited March 6, 2015 by Sensuki

[mrmonocle]

The simplest way to do it is just to alter the base damage of the weapons with a formula so that 1H weapons will fall off after a certain DR. I don't think Josh likes to balance things that way though. Seems to prefer doing things by 'feel' instead of by formulas.

 

However, one thing to take into account is DR bypass. Vulnerable Attack gives you 5 DR bypass at the cost of some attack speed. The attack speed cost is not as much for dual wield as it is for 1H single or 2H style, therefore you can attack really fast and do close to full damage with the right weapons - making dual wielding 1H normal style weapons very strong. It's okay with 1H Fast weapons but it's much better for 1H normal IMO.

+ DR description says that 20% minimum of the rolled damage is guaranteed regardless of DT.

[tdphys]

The simplest way to do it is just to alter the base damage of the weapons with a formula so that 1H weapons will fall off after a certain DR. I don't think Josh likes to balance things that way though. Seems to prefer doing things by 'feel' instead of by formulas.

 

However, one thing to take into account is DR bypass. Vulnerable Attack gives you 5 DR bypass at the cost of some attack speed. The attack speed cost is not as much for dual wield as it is for 1H single or 2H style, therefore you can attack really fast and do close to full damage with the right weapons - making dual wielding 1H normal style weapons very strong. It's okay with 1H Fast weapons but it's much better for 1H normal IMO.

 

You can also take Two Weapon Style to completely offset that penalty. Whereas for 2H you have x1.1 damage for Two Handed Style.

 

1H style and fast weapons are still pretty bad, but dual wielding two handed normal weapons is awesome IMO.

 

 

I'd personally like to see balance being an emergent part of damage calculations, but I suppose you could balance things through added talents.

[Bazy]

Keep in mind that with the vast number of unique weapons in the game it's going to be pretty rare for someone to carry a "normal" weapon for too long. It may be the case that the unique effects of faster weapons may make them better than a number of slow weapons. 

[BJMJD]

I think, you should not calculate the damages for 1 attack but for a delay of time.

 

(I don't have access to the demo, so everything I will say is based on what saw on videos or read on the forum. Correct me if anything is wrong.)

 

In a case when the accuracy is equal to the deflection, only the roll matter:

 

1-15-> Damage*0

16-50-> Damage*0.5

51-100-> Damage*1

Which give us an average of:

Domage*(0*15+0.5*35+1*50) = Damage*0.675

 

2H weapon do a damage between 14 and 20 so an average of 17. So with the roll chance it is 17*0.675 = 11.475 per attack

 

A fast weapon do a damage between 8 and 12 so an average of 10. So with the roll chance it is 10*0.675 = 6.75 per attack

 

(Here is the point where I am not certain)

I saw on a video where a character was dual wield two fast weapons (It has certainly "Two Weapon Style").

This character was able to attack 3 times while an attack of 2H weapon.

 

So for the time of 1 attack of 2H, we have 3 attacks of dual wield fast weapon.

 

So in term of damage it means:

 

2H weapon -> 1*11.475 = 11.475 (with x1.1 of Two Handed Style it is 12.6225 )

Dual wield two fast weapons -> 3*6.75 = 20.25

 

Even if you keep only 20% of the damage due to the DR, it is:

 

2H weapon -> 2.295 (with x1.1 of Two Handed Style it is 2.5245)

Dual wield two fast weapons -> 4.05

 

So apparently, unless I forgot something, dual wield two fast weapons makes more damage than a 2H weapon.

[Sensuki]

None of your calculations include recovery time though, just some math based off ACC-DEF = 0

 

Matt516 did some pretty accurate calculations in one of the earlier beta versions that accounted for recovery time at that stage of development - right now we're not sure how the global recovery factor is handled.

 

Dual wielded fast weapons have an animation of 20 frames (0.6667s) and a recovery of 28 frames (20*1.4) (0.9333333s)

 

I'm not sure what the calculation of recovery time is for Two Handed Weapons because I'm not sure if the global recovery mult is additive or multiplicative. It could be 30 * 1.4 * 1.5 or 30 *(1+0.4+0.5) ... I don't know.

 

Attack Time for Dual Wielded fast weapons in no armor with 10 Dexterity is 1.6 seconds

 

Attack Time for Two Handed Weapons in no armor with 10 Dexterity is either 2.9 seconds or 3.1 seconds

 

So Dual Wielding fast weapons is not 3 times faster than a Two Handed Weapon.

Edited March 7, 2015 by Sensuki

[Sensuki]

Just did some testing on this topic.

 

There appears to be no *1.4 recovery mult in v435 and dual wielding stilettos is in fact 20 frames + 20 frames = 1.33 sec attack time

 

The recovery for 1H and 2H weapons I was getting 54 frames from my video, but it's probably 60 frames to account for dropped frames. So the attack time is actually 3 seconds.

 

Wow 1H is so bad lol

[Sensuki]

edit: dw

Edited March 7, 2015 by Sensuki

[Sensuki]

Anyone interested in formulas might be interested in this spreadsheet: https://docs.google.com/file/d/0Bzc2Dec83wSnNENNWjNsemhDcUU/edit?usp=drive_web

Matt516's drive on Pillars of Eternity stuff, *some* of the calculations and/or values are out of date

Edited March 7, 2015 by Sensuki

[ISC]

The simplest way to do it is just to alter the base damage of the weapons with a formula so that 1H weapons will fall off after a certain DR. I don't think Josh likes to balance things that way though. Seems to prefer doing things by 'feel' instead of by formulas.

 

However, one thing to take into account is DR bypass. Vulnerable Attack gives you 5 DR bypass at the cost of some attack speed. The attack speed cost is not as much for dual wield as it is for 1H single or 2H style, therefore you can attack really fast and do close to full damage with the right weapons - making dual wielding 1H normal style weapons very strong. It's okay with 1H Fast weapons but it's much better for 1H normal IMO.

 

You can also take Two Weapon Style to completely offset that penalty. Whereas for 2H you have x1.1 damage for Two Handed Style.

 

1H style and fast weapons are still pretty bad, but dual wielding two handed normal weapons is awesome IMO.

 

So what you're saying is that dw 2 maces + Vulnerable Attack is viable and > stilettos? 

[Sensuki]

Dual Wielding any of the 1H normal weapons with Vulnerable Attack and Two Weapon Fighting is quite good. It just depends. Personally I like Spears, Maces and Sabres over the others.

[GreyFox]

I hope Dex gets a buff but I'm one of the people who thinks it should add more than just speed....otherwise just have a SPEED stat...

 

Even OE uses mostly 2H weapons in their playthroughs...

[Sensuki]

I don't think there's much of a problem with Dexterity, it's a useful attribute it's just not an attribute that you want to MAX without finding some specific niche for it.

 

For me, the largest problem is Perception, because Interrupt is just not a useful attribute for casters, really. Weapon users can enchant their weapons with accuracy bonuses and damage multipliers, and they can take talents to improve their weapon use.

 

Spell casters cannot do either. Their spell damage does not scale, and they are non-weapon attacks so they merely use base accuracy + level, which is always going to be worse than weapon accuracy.

 

Perception as I imagined it - Accuracy and Interrupt, would actually be great for casters because they *NEED* the accuracy, as that's the only thing that can improve their attacks with spells drastically other than Might or Intellect.

 

Accuracy in the current version isn't really even that important after a point, because the extra damage multipliers you can get for weapons mean A LOT more than a 1% chance to deal +50% damage. When weighing +8 Accuracy against x1.3 damage, that's a no brainer - take the damage.

[ISC]

Accuracy in the current version isn't really even that important after a point, because the extra damage multipliers you can get for weapons mean A LOT more than a 1% chance to deal +50% damage. When weighing +8 Accuracy against x1.3 damage, that's a no brainer - take the damage.

 

By the way, shouldnt tool tips and weapon descriptions etc say "x1.3 base damage"? or perhaps even "+30% base damage". The point being that it would make the damage system a fair bit more obvious, because +50% damage from a crit sounds like much more than it is.

[Sensuki]

It's actually +30% rolled damage. I submitted some new mechanic descriptions in the bug forums to better display damage calculation. In the item tooltip it should probably say +30% damage instead of x1.3 because the multipliers are no longer multiplicative.

Edited March 8, 2015 by Sensuki

[SergeantHans]

I don't think there's much of a problem with Dexterity, it's a useful attribute it's just not an attribute that you want to MAX without finding some specific niche for it.

 

For me, the largest problem is Perception, because Interrupt is just not a useful attribute for casters, really. Weapon users can enchant their weapons with accuracy bonuses and damage multipliers, and they can take talents to improve their weapon use.

 

Spell casters cannot do either. Their spell damage does not scale, and they are non-weapon attacks so they merely use base accuracy + level, which is always going to be worse than weapon accuracy.

 

Perception as I imagined it - Accuracy and Interrupt, would actually be great for casters because they *NEED* the accuracy, as that's the only thing that can improve their attacks with spells drastically other than Might or Intellect.

 

Accuracy in the current version isn't really even that important after a point, because the extra damage multipliers you can get for weapons mean A LOT more than a 1% chance to deal +50% damage. When weighing +8 Accuracy against x1.3 damage, that's a no brainer - take the damage.

Would be cool with an enhancement to weapons that gave accuracy to spells

[Dr. Hieronymous Alloy]

I don't think there's much of a problem with Dexterity, it's a useful attribute it's just not an attribute that you want to MAX without finding some specific niche for it.

 

For me, the largest problem is Perception, because Interrupt is just not a useful attribute for casters, really. Weapon users can enchant their weapons with accuracy bonuses and damage multipliers, and they can take talents to improve their weapon use.

 

Spell casters cannot do either. Their spell damage does not scale, and they are non-weapon attacks so they merely use base accuracy + level, which is always going to be worse than weapon accuracy.

 

Perception as I imagined it - Accuracy and Interrupt, would actually be great for casters because they *NEED* the accuracy, as that's the only thing that can improve their attacks with spells drastically other than Might or Intellect.

 

Accuracy in the current version isn't really even that important after a point, because the extra damage multipliers you can get for weapons mean A LOT more than a 1% chance to deal +50% damage. When weighing +8 Accuracy against x1.3 damage, that's a no brainer - take the damage.

 

 

Yeah I hate to constantly beat this drum but it really does feel like the one glaring problem in an otherwise very solid system. I even like the engagement mechanics and other things that seem to be hot-button issues around here, probably because I'm more of a fan of turn-based systems anyway and a lot of those issues are just structural conflicts caused by trying to reconcile RTWP to a tactical game. 

 

Overall the game has surprisingly few balance issues, at least as per the BB. 

 

Wizard self-buffs (esp. Arcane Shield) seem a little underpowered and should probably have more uses/be castable out of combat. An in-combat buff has to be worth the free hit you're giving the enemy by not attacking first, and it's really hard for a single-target buff to hit that threshold.

 

There are some UI issues and terminology that could be clearer (which you've done a good job of detailing).

 

One-Hand-Only combat approach seems a little underpowered.

 

Some of the more fragile classes could probably benefit from a *slightly* more generous Endurance curve. Wizards aren't going to be putting all their points in Constitution anyway, give 'em a little more reward if they do.

 

There are a few other problems that might show up once we get the full game -- for example, I could see it being a problem that Paladin modal buffs seem to get completely superseded by late-game cleric spell buffs,and especially once the addons and expansions hit, character advancement and the main game plot storyline could get trivialized. But those are for down the road.

 

 

Just about all of those aren't really that serious though because it's a single player game and within certain tolerances you can play as you want. If you're deliberately going one-handed ok that's a choice and you're playing on challenge mode, basically. 

 

The biggest issue I see right now is, as you say, Perception not doing jack for ranged characters and casters. I'm taking heart from the fact that the devs seem to not be commenting on that at all -- I'm crossing my fingers and telling myself that means they're aware of the issue and have some solutions in testing. Either that or they've decided we're full of it and that discussion is over  

Edited March 8, 2015 by Dr. Hieronymous Alloy