[MECHANICS] Stats, weapons, and rules of thumb

[Prince of Lies]

Hmm, I have question about a Two Handed Aumaua Fighter. Is it fair to say that starting base of 20 Might, 14 Dex, 14 Per isn't optimal on difficulty less than PoTD? Would it make sense to lower Might to 16 to have Dex and Per at 16 as well?

 

Or keep the 20 base Might but have equipment/other variables to boost Dex/Per in your opinion?

 

When regarding non-POTD, Suboptimal does not equal Unusable. You can definitely get around fine with 20 Might, 14 Dexterity, 14 Perception. 

 

That being said, I would probably pump perception up a couple of points. Reduce constitution by a couple points.

 

If your finding you're dying more at 8 constitution (-10% HP at 8 Con, 37.8 Base (versus 42) + 10.8 per level (versus 12 per level) ), take the tough talent or boost your constitution with gear.

[XEternalXDreamsX]

 

Hmm, I have question about a Two Handed Aumaua Fighter. Is it fair to say that starting base of 20 Might, 14 Dex, 14 Per isn't optimal on difficulty less than PoTD? Would it make sense to lower Might to 16 to have Dex and Per at 16 as well?

 

Or keep the 20 base Might but have equipment/other variables to boost Dex/Per in your opinion?

When regarding non-POTD, Suboptimal does not equal Unusable. You can definitely get around fine with 20 Might, 14 Dexterity, 14 Perception.

 

That being said, I would probably pump perception up a couple of points. Reduce constitution by a couple points.

 

If your finding you're dying more at 8 constitution (-10% HP at 8 Con, 37.8 Base (versus 42) + 10.8 per level (versus 12 per level) ), take the tough talent or boost your constitution with gear.

Eder will be my primary Tank so I could see lowering Constitution to give my PC more of an edge with Perception. Like you said, I could always take Tough or boost health through gear if I see myself not performing great on my formation. Thanks, man. I'm also going to have another frontliner so I don't get focused down.

[Prince of Lies]

No problem

 

If you use Abraham as a pet, you get reduced recovery penalty (varies depending on armor used, you attack faster) and + 10 HP heal on kill. 

[Lampros]

 

So what's the bottom line? It seems like that the OP is saying that Perception, Might, and Dexterity are all roughly similar from a DPS perspective?

 

bottom line is Perception > Might (but it's close). similarly, +1 acc > +3% damage (but it's close). (though this ranking breaks down at very high accuracies, though that is rare for much of potd)

 

for martial classes, Dexterity > Perception or Might. For casters, it might be closer to Perception > Might > Dexterity depending on your build/playstyle (basically when you have a finite number of spells, getting the most out of each of your spells may be more important than simply emptying your spellbook faster)

 

 

That's quite counter-intuitive? I thought the conventional thought has been that casters need Dexterity most because of the need to prevent being interrupted and thus cast fast.

[Boeroer]

I don't want to put anybody's nose out of joint, but what Kaylon was saying (besides other stuff) and what was interesting to me:

 

---

Your example of "-30% recovery time equals to 42.8% action speed" is only true (numerically) if you leave out the attack animation phase (which is not affected by the recovery bonus). So basically you have +42.8% recovery speed, not overall action speed (which would include animation and recovery phase).

 

While an action speed bonus of 30% would affect animation phase and recovery phase alike.

 

If you then take your +42.8% recovery bonus and add the unaltered attack animation phase you'll end up with +34% overall attack speed which is not that much of a difference to 30%.

---

 

That's what I understood Kaylon was saying and what he claimed you didn't address in your response.

 

While he can sometimes show the posting style of an asperger-shaped dead blow hammer (don't ask me to draw that shape) - I think I understand what he wanted to say there and I think he has a point? And I think you didn't address that particular point.

 

The discussion was very enlightening and informative so far but would have profited from less "sorry to burst your bubble" or "go and look up this and that because you didn't get it" and a bit more benignity. It's easy to say that because I'm not involved.

[thelee]

I don't want to put anybody's nose out of joint, but what Kaylon was saying (besides other stuff) and what was interesting to me:

 

---

Your example of "-30% recovery time equals to 42.8% action speed" is only true (numerically) if you leave out the attack animation phase (which is not affected by the recovery bonus). So basically you have +42.8% recovery speed, not overall action speed (which would include animation and recovery phase).

 

While an action speed bonus of 30% would affect animation phase and recovery phase alike.

 

If you then take your +42.8% recovery bonus and add the unaltered attack animation phase you'll end up with +34% overall attack speed which is not that much of a difference to 30%.

---

 

That's what I understood Kaylon was saying and what he claimed you didn't address in your response.

 

While he can sometimes show the posting style of an asperger-shaped dead blow hammer (don't ask me to draw that shape) - I think I understand what he wanted to say there and I think he has a point? And I think you didn't address that particular point.

 

The discussion was very enlightening and informative so far but would have profited from less "sorry to burst your bubble" or "go and look up this and that because you didn't get it" and a bit more benignity. It's easy to say that because I'm not involved.

 

i get that that's what he was saying, but what was really frustrating me was that i kept on trying to say that that was not what i was saying was the net effect... though kaylon kept pushing it as if that's what i was saying and that this was undermining my conclusions, which it wasn't (and in fact he was selectively quoting me out of context to push his point), and also that's not what i was saying and in fact even if i had said something ridiculous like "the action speed reduction is superliminal and makes you travel back in time" it still wouldn't have affected the results from the simulations.

 

grr anyway don't want to relitigate it.

 

EDIT: in case it isn't clear, what i've been saying repeatedly is that it's a +42.8% through the recovery. maybe my phrasing is just unintuitive or confusing. i've even also said "ignoring the attack" to clarify that i'm talking about the recovery time only (and part of my aggravation earlier was kaylon selectively quoting the part RIGHT before my very very important qualifiers). i'm trying to illustrate how a single, increasing recovery time bonus has outsized importance (increasing returns) because of the inversion to get to the net action speed, which is a more comparable number to damage boosts and such and that failure to account for this on the part of Obsidian's designers is what probably led to the outcome where 2w was just basically the definitive choice for weapons for so long. i edited my OP earlier today to try to make this more clear.

Edited October 20, 2018 by thelee

[thelee]

 

 

So what's the bottom line? It seems like that the OP is saying that Perception, Might, and Dexterity are all roughly similar from a DPS perspective?

 

bottom line is Perception > Might (but it's close). similarly, +1 acc > +3% damage (but it's close). (though this ranking breaks down at very high accuracies, though that is rare for much of potd)

 

for martial classes, Dexterity > Perception or Might. For casters, it might be closer to Perception > Might > Dexterity depending on your build/playstyle (basically when you have a finite number of spells, getting the most out of each of your spells may be more important than simply emptying your spellbook faster)

 

 

That's quite counter-intuitive? I thought the conventional thought has been that casters need Dexterity most because of the need to prevent being interrupted and thus cast fast.

 

 

Like I said, it depends on play style and build. For example: enemy AI is relatively predictable and gameable. If you have suitable engagement and stats/distance, interruption may be less of a concern because you're only taking incidental damage at best. You can also cancel spellcasts before they get interrupted.

 

Another example: if i have a martial/caster multiclass, i might privilege dexterity alot to spend less time casting (and risking interrupts) and more time attacking. but if i'm a safe single-class debuffer, all dexterity lets me do is empty my spellbook faster and then i'm stuck with a relatively crappy autoattack. in such a case it might be more useful to make each spell I do cast much more impactful by increasing perception. (this also depends on difficulty because on PotD with upscaling running out of spells and still having a decent amount of combat left is going to be more common than on, say, story mode)

Edited October 20, 2018 by thelee

[Boeroer]

 

[blah...]

 

i get that that's what he was saying, but what was really frustrating me was that i kept on trying to say that that was not what i was saying was the net effect... though kaylon kept pushing it as if that's what i was saying and that this was undermining my conclusions, which it wasn't (and in fact he was selectively quoting me out of context to push his point), and also that's not what i was saying and in fact even if i had said something ridiculous like "the action speed reduction is superliminal and makes you travel back in time" it still wouldn't have affected the results from the simulations.

 

grr anyway don't want to relitigate it.

 

EDIT: in case it isn't clear, what i've been saying repeatedly is that it's a +42.8% through the recovery. maybe my phrasing is just unintuitive or confusing. i've even also said "ignoring the attack" to clarify that i'm talking about the recovery time only (and part of my aggravation earlier was kaylon selectively quoting the part RIGHT before my very very important qualifiers). i'm trying to illustrate how a single, increasing recovery time bonus has outsized importance (increasing returns) because of the inversion to get to the net action speed, which is a more comparable number to damage boosts and such and that failure to account for this on the part of Obsidian's designers is what probably led to the outcome where 2w was just basically the definitive choice for weapons for so long. i edited my OP earlier today to try to make this more clear.

 

Yes, good explanation. I think now most people can really understand what you meant and how it's different from what Kaylon was saying.

 

By the way I don't think that Kaylon was being disingenuous but that you just talked past each other a bit.

Sometimes blunt hammer style - but I never saw insidious dagger style.

Edited October 20, 2018 by Boeroer

[Torm51]

Man this is depressing.  What an RPG!!! Every martial character should max DEX, wow thats boring and bad design on obsidians part, but the math guy is right 100%

Edited October 20, 2018 by Torm51

[Boeroer]

The differences between DEX/PER/MIG are not that big. And here it's only about dps. MIG is also useful for healing, PER is useful for finding traps and secrets and so on. I think the usefulness of stats is pretty balance except RES which is often still a dumpstat.

Edited October 20, 2018 by Boeroer

[Torm51]

The differences between DEX/PER/MIG are not that big. And here it's only about dps. MIG is also useful for healing, PER is useful for finding traps and secrets and so on. I think the usefulness of stats is pretty balance except RES which is often still a dumpstat.

I will be honest Boer I have been less active in the game, it just seems worse.  Talk me off the ledge man.  Res sucks again not even for lessening debuffs??!?!

[thelee]

 

The differences between DEX/PER/MIG are not that big. And here it's only about dps. MIG is also useful for healing, PER is useful for finding traps and secrets and so on. I think the usefulness of stats is pretty balance except RES which is often still a dumpstat.

I will be honest Boer I have been less active in the game, it just seems worse.  Talk me off the ledge man.  Res sucks again not even for lessening debuffs??!?!

 

 

I think Prince of Lies put it best earlier upthread "suboptimal does not mean unusable." This isn't BG or BG 2 where a bad stat allocation makes for a literally unplayable character.

 

I have made at least two decent characters that valued res and they came out pretty well. It's not useless (*cough* BG/BG2/IWD charisma *cough*), it's just niche. (And those niches can be very, very powerful, since Res gives you deflection which has increasing returns)

 

(one of those characters was a support tank that could get up to 30+ res with very high uptime ~100%, which is some pretty serious debuff reduction; basically meant every hit was turned into a graze, and anything that grazed barely even registered)

Edited October 20, 2018 by thelee

[Boeroer]

With good RES, Lone Wolf and Clarity of Agony you can pretty much remove every afflicton (including self damage of Berserker Frenzy, Sacred Immolation or Alacrity) - so RES can have its uses. But it's still the stat that gets dumped most - followed by CON. That is because most people favor offense over defense I think.

[Hulk'O'Saurus]

Perception would be my dump stat of choice... 

 

But I'll leave it at that. 

[AndreaColombo]

^ No you won’t—can’t throw the bone and leave us wanting for the meat

[Boeroer]

Because ACC of scrolls is not determined by the character's ACC?

[Hulk'O'Saurus]

If you're talking about that particular character, he was at a solid 50% crit rate before he started using more scrolls.

 

There are just so many ways to raise Accuracy and you can add conversions as well. We shouldn't forget debuffs.

 

At worst I've stopped taking it over 10, and it's fine. I really do value Dex the most, closely followed by Int.  

[thelee]

If you're talking about that particular character, he was at a solid 50% crit rate before he started using more scrolls.

 

There are just so many ways to raise Accuracy and you can add conversions as well. We shouldn't forget debuffs.

 

At worst I've stopped taking it over 10, and it's fine. I really do value Dex the most, closely followed by Int.  

 

In terms of dumpability, you're also not generally mathematically wrong. For anyone else just listening in: my simulations only took into account stats of 10+ (my omission). I haven't updated for the full range of stats, but because of double-inversion nonsense when you bring might and dex down you need a much larger amount of equivalent damage or action speed boosts to counter that penalty, whereas with perception--even though though the penalties have increasing returns--you can counter accuracy penalties with accuracy bonuses 1:1. Perception had such a minor edge over might on net effects for stats >= 10 that I feel comfortable saying without running more simulations that the double-inversion means that might becomes a more precious stat than perception when deciding what to take below 10.

 

So when it comes to which of might, dex, or perception you can dump, perception is the most dumpable. Though personally to me it doesn't make much sense to dump perception below 10 just to boost might some more (dex though).

 

Double-inversions, boy... nice way to come with situations where 1 + 1 = 3

Edited October 21, 2018 by thelee

[swapoer]

Very informative and accurate according to my own research.

I did something like your post for POE 1 and I come out with some Rule of Thumb for POE 1 too. My ration and thinking are much the same as yours.

 

https://forums.obsidian.net/topic/92293-a-dps-spreadsheet-calculatorwip/

 

 

Sabres +10% damage => ~3might or ~6% net damage.

 

What a blast. You are absolutely right and the result is just so easy to miss. I take it that you take the chance to crit, hit, graze and miss into consideration when calculating net damage. Normally, only 60% damage will go through the deflection.

BTW: I use D*CTH instead of “net damage”. D*CTH means Damage×Chance To Crit, hit, graze and miss. 60% is a good rough estimate of CTH.