Rogue advice

[Zenicetus]

Pike/Quarterstaff are indeed one of the best melee options for rogues. Everyone who is obsessed with dw is just following the stereotypes I think. DW can deal more damage in certain circumstances but is very risky. I wouldn't run a dw rogue. Also I wouldn't run ranged rogue cause they don't get much outside of SA tbh. So 2h reach rogue is probably the only way I'd use rogue in a party.

 

For dual wielding, isn't there at least some advantage in alternating between two weapons with different damage enchantments and debuff effects? Or is that just armchair theory? I haven't tested it.

 

Aside from that, I admit that I'm using a dw Rogue just because it fits my preconceived idea of the class, compared to running around with a pike or other 2h weapon. Works fine so far, but I'm only playing on Hard.

[dam]

Which weapon would be better fit ranged rogue : Persistence A hunting bow from endless paths with huge damage multiplayer 1.45 or Borressaine War Bow with stun procs ?

 

That's situational.

 

Your Hunting Bow will let you get more sneak attacks off before your debuff expires on the target.

Your War Bow will punch through DR a little bit easier because it has a higher damage range, although do not hold your breath for sneak attacks from the stun proc, if it's anything like the Phantom stun proc, it's much too short.

 

Alternatively and if you're facing high DR opponents, the Hold Wall x-bow from the smith in Gilded Vale has a prone proc.

[dam]

 

Barbarian is incidentally better with super fast dw if you're into bug abuse. Pike rogue does more single target dmg than pike barb if you play your cards correctly,

 

Yeah, OSA is definitely not working as intended.  While rogue does more single target dps (especially at low levels when the 1.5 sneak attack bonus matters more), barb's carnage can't be ignored and is a more interesting mechanic than just more single target dps -- but that's subjective I suppose.  Even the cipher can get a 1.4 multiplier at the cost of a talent.  The way those damage modifiers work, they have less total effect as you stack them higher (they all work off base damage, as they should).

 

If rogue's sneak attack was its own separate multiplier, the class would have something there.

 

 

The class already has something there, you're forgetting the synergy between its talents, and possibly even race if you went that way :

- hearth orlan : +10% crit

- dirty fighting : +10% crit

- vicious fighting : +10% crit

- sneak attack : +50% damage

- deathblows : +50% damage again, seeing it doubles your SA damage when attacking a target w/ 2 afflictions

- house doemenel talent : +30% crit mult

- reckless assault : +20% melee damage

- savage attack : +20% melee damage

 

 

Let's work under the assumption that you're wielding a standard 2-handers with a range of 14-20 damage, hereafter written "BD" for Base Damage.

 

Unless my math is wrong :

- sneak attack : BD + (BD * (0.5 + 0.5 + 0.2 + 0.2)) = BD + (BD * 1.4) = 34-48 damage , on a regular hit

- crit SA : BD + (BD * (0.5 + 0.5 + 0.2 + 0.2 + 0.3 + 0.5)) = BD + (BD * 2.2) = 45-64 damage , on a crit , discounting other crit multiplier bonii

Not sure crit modifiers are applied this way, they might be applied to  your end damage after all the other modifiers are factored in.

 

 

Make the same crit with 20 might (+30% damage), which is achievable mid-game with +3 might gear :

- crit SA : (BD * 1.3) + ((BD * 1.3) * 2.2) = 58-83

 

And that, is using a regular weapon, you can get 25% more from a Lash, and 45% more from Superb.

You can also get +0.5 crit mult by using battle axes (that, or just +15% damage from 2-handed spec).

You can get an additional 0.1 crit mult if you side with the Looters, at some point during Act 3.

You can, if you want to push it, factor in Kana's fire chant +25 flat burn damage.

You can also get +8 might from a priest (I believe that's not cumulative with the +3 from gear though) for an extra ((8-3) * 3)%

Oh and if you really, really are going for style points, you can even factor in +25% from a Slaying enchant.

I think there's even a +0.5 crit mult 2-handers in Dyrford Village, so you could get +15% from 2-handed spec and the +50% crit mult.

 

 

While they're additive, that's still a lot of modifiers, and on a class/race combo with +30% natural crit chance, and a very high accuracy.

 

At this point, even the Master Below's DR won't last very long (although I still have to meet him on PotD).

 

 

 

As for barb AOE vs rogue single target damage, that is a matter of viewpoint.

Keep in mind that 5 monsters at 20% hp still hit like they're at 100% hp.

I for one would much rather have 2-3 full hp monsters remaining, than 5 at 23% and still mauling the tank.

Edited April 11, 2015 by dam

[dam]

Ok so I've taken the time to do a bit more theorycrafting and this is what I've come up with.

 

If someone thinks my math is in error, kindly correct.

If someone knows how the crit damage modifiers are applied (base damage, or end damage after other mods ?), kindly correct.

 

 

Tall grass has 18-26 base damage (and costs 12k, your best investment in the whole game for a 2H rogue).

It's exceptional, crits inflict prone, and converts 10% hits into crits.

 

A hit benefits from :

MODAL: 20% modifier from Reckless Assault

MODAL: 20% modifier from Savage Attack

ENCHANT: 30% modifier from Exceptional

ENCHANT: 25% additional damage from a Lash

STATS: 30% from 20 might (you can get 21 from +3 might gloves but let's keep things easy)

PASSIVE: 50% damage from being a Sneak Attack, or you're doing it wrong

PASSIVE: 50% extra damage on your SA because you took Deathblows like a good rogue

PASSIVE: 15% modifier from 2handed spec

Total: 20 + 20 + 30 + 25 + 30 + 50 + 50 + 15 = 240%

 

A hit deals BD + (BD * modifiers) damage, or BD * (1 + (modifiers))

In our case the modifiers total 240% or an additional 2.40 BD damage, a regular hit therefore deals 340% of the weapon's base value.

Thus discounting DR, a regular hit with Tall Grass deals : 61-88 damage

 

 

A crit, which a Hearth Orlan has (10% natural + 20% vicious fighting + 10% Tall Grass) 40% chance of landing, also gains :

PASSIVE: 50% crit modifier just for being a crit

PASSIVE: 30% crit modifier from house doemenel , after all you made a rogue , that's the only faction that's worth

 

 

Now that is the part where I'm unsure how the game computes the final numbers.

Either these modifiers are added to the other list of modifiers, or our previous end damage gets multiplied by our crit mods.

Case 1 : 240% + 50% + 30% = 320% , so a crit will deal BD + 320%BD or 420% BD damage

Case 2 : (240% * 180%) = 430% , so a crit will deal BD + 430%BD or 530% BD damage

 

Case 1 damage is : 18-26 => 75-109

Case 2 damage is : 18-26 => 95-138

 

Edit:  the ingame Cyclopedia states "a crit increases the total damage done by 50%", so I'd be inclined to go with case 2.

 

 

 

Lategame, add an extra 15% when your weapon becomes Superb, and 25 flat damage from Kana's Flaming Weapons.

Here as well, I'm unsure how the Chanter's flaming weapons damage is computed:

+25 flat ?

+25%, like a Lash ?

 

 

These findings look consistent with my current L7 rogue, who landed a 84 damage crit with :

- 18 might

- no Deathblows

- no 2H spec

 

 

Thoughts ?

 

 

On a side note, Tall Grass is borderline game breaking with the Prone proc, I literally prone-locked a Drake on the Warchief bounty in Valewood.

Edited April 12, 2015 by dam

[Stun]

^Yep, Tall Grass Kicks ass.

 

I'd be interested to see if there's any meaningful changes in the damage totals if, instead of a hearth orlan, you went with a Living Lands Coastal Aumaua, since they can start the game with 21 might, 24 when they get the +3 might gloves.

 

(although, on my current rogue run, I found the Rabbit Fur Gloves, which give a +10% crit multiplier bonus. That'd probably be a better choice for the glove slot on a rogue.)

Edited April 12, 2015 by Stun

[dam]

^Yep, Tall Grass Kicks ass.

 

I'd be interested to see if there's any meaningful changes in the damage totals if, instead of a hearth orlan, you went with a Living Lands Coastal Aumaua, since they can start the game with 21 might, 24 when they get the +3 might gloves.

 

(although, on my current rogue run, I found the Rabbit Fur Gloves, which give a +10% crit multiplier bonus. That'd probably be a better choice for the glove slot on a rogue.)

 

Well it's very easy to compute isn't it ?

 

+10% crit multiplier, assuming a 50% hit-crit rate, means 5% additional damage.

+3 might is 9% additional damage.

So that's your answer regarding rabbit gloves vs might gloves.

 

 

Now on to the race, that depends on the rogue you're building.

If you're building around the Tall Grass, as I currently am, you want to get more crits to :

- proc Prone

- set up your own SAs

- set up your own Deathblows if the target is flanked

Which translates into easier fights because of the CC, and in more damage obviously.

 

24 vs 21 might is a 9% difference in damage dealt.

I prefer the 10% more crits plus CC and more deathblows.

[dam]

Ok so, remember how I suggested we move away from our DnD habits of very fast weapons inflicting flat damage bonii for sneak attacks, in favour of slower, 2-handed weapons that vanquished foes' DR ?

 

Well here you go, Adra dragon down in 3 seconds flat, on Hard, with a level 11 party :

http://forums.obsidian.net/topic/77486-spoiler-master-below-down-in-3-seconds-no-cheese/

[MadDemiurg]

Ok so, remember how I suggested we move away from our DnD habits of very fast weapons inflicting flat damage bonii for sneak attacks, in favour of slower, 2-handed weapons that vanquished foes' DR ?

 

Well here you go, Adra dragon down in 3 seconds flat, on Hard, with a level 11 party :

http://forums.obsidian.net/topic/77486-spoiler-master-below-down-in-3-seconds-no-cheese/

That doesn't show that dw is good, only that petrify is good. You don't need dw to make it good.

[Daemonjax]

 

Well it's very easy to compute isn't it ?

 

...

 

24 vs 21 might is a 9% difference in damage dealt.

 

 

 

Math can be more complicated than it seems at first glance.

 

+3 might is 9% additional damage only for characters that have a natural 10 strength and  have no other damage modifiers (sneak attacks, weapon masteries, weapon enchantments, etc.).

 

I'd put those +3 might gloves on a spellcaster with the lowest might (or maybe a moon godlike), and the +10% crit multiplier (.5 crit damage modifier becomes a .6 which is pretty insignificant considering every other damage modifier that character has) on whoever has the highest accuracy and no other gloves to wear.

Edited April 14, 2015 by Daemonjax

[dam]

 

Ok so, remember how I suggested we move away from our DnD habits of very fast weapons inflicting flat damage bonii for sneak attacks, in favour of slower, 2-handed weapons that vanquished foes' DR ?

 

Well here you go, Adra dragon down in 3 seconds flat, on Hard, with a level 11 party :

http://forums.obsidian.net/topic/77486-spoiler-master-below-down-in-3-seconds-no-cheese/

That doesn't show that dw is good, only that petrify is good. You don't need dw to make it good.

 

 

I should hope it doesn't show that DW is good, this is with 2H...

[dam]

It's been pointed out to me that Petrify spells (and by extension traps) seem to apply a 400% damage taken to the enemy, so, sadly, this kill isn't as impressive as I had hoped it would be

[BramSd]

So, for a 2H pike-wielding Rogue, how would you distribute the attributes?

 

I get you want to max DEX and MIG (thinking 19/17 respectively) and you can bring CON down a bit as you woun't need much of it, but then what? Are you gong to increase INT for the added duration, or dump it for RES? And PER? Leave it around 10-12 or put more into it?

 

Would like to know your thoughts on it/

[Stun]

It's generally a good idea to max out Might and Intelligence first. In fact, I'd pump intelligence all the way to 18 even if it means you have to drop Dex a few points, actually. As much as it may seem that the added few % points to action speed will make a difference, it doesn't. Although it's probably a good idea to at least keep your dex in the teens.

 

But after that, none of the other attributes matter and you can drop them to single digits without taking a combat performance hit. The game recommends perception on Rogues, but I still haven't figured out why that would be. For the Interrupt? Well, you're using a pike. You'll get interrupts all the time with it anyway (it's a high interrupt weapon). And eventually you'll find a Pike that inflicts Prone (knockdown - a forced interrupt. Can't beat that)

Edited April 21, 2015 by Stun

[SilchasRuin]

So, for a 2H pike-wielding Rogue, how would you distribute the attributes?

 

I get you want to max DEX and MIG (thinking 19/17 respectively) and you can bring CON down a bit as you woun't need much of it, but then what? Are you gong to increase INT for the added duration, or dump it for RES? And PER? Leave it around 10-12 or put more into it?

 

Would like to know your thoughts on it/

Max Might, Int, and Dex, dump (go to lowest value) Con and Perception and put anything remaining in resolve.

 

Resolve is slightly better than perception because will save is usually what is targeted for charm/confuse/dominate. Constitution is functionally worthless on a rogue except for fortitude saves as Rogues have low base health and constitution is a percent modifier so the actual amount of Health/Endurance lost from minimum constitution compared to average constitution is negligible. You want max int because the best weapon for Pike rogue is Tall Grass which gives prone on Crit which will have its duration buffed or reduced by your int. Dexterity is arguably the most important stat for damage on a Rogue as it's multiplicative with all other damage modifiers, outside of possibly speed and recovery modifiers (though Im unsure about those), which means compared to might, by mid game dexterity will dramatically outscale might as far as value for dps goes.

 

If for some reason you absolutely must have non-minimized perception and constitution you'd be better off dropping might than dex or int (maybe, depending on rate of prone application if your rogue stays focused on a single target average int may work if used in conjunction with high crit).

Edited April 21, 2015 by SilchasRuin

[SilchasRuin]

It's generally a good idea to max out Might and Intelligence first. In fact, I'd pump intelligence all the way to 18 even if it means you have to drop Dex a few points, actually. As much as it may seem that the added few % points to action speed will make a difference, it doesn't. Although it's probably a good idea to at least keep your dex in the teens.

 

This is flat out false, dexterity is far stronger than might as a dps increase on a rogue, and is far better at reapplying prone as well. Dexterity is a multiplicative modifier with all damage boosting stats, Might is additive with the majority of damage boosting stats, including reckless assault, sneak attack, deathblows, vicious weapon modifier, qualitiy modifier, and I believe critical modifier. 30% bonus damage from might to 300% damage is only 330% total dps whereas 30% attack speed is something like 20% dps increase, so would increase dps to 360%.

[rheingold]

The pike build sounds interesting, I'll give it a go with my next rogue character, currently running a orlan rogue dw, non optimized on hard and oh my is she the original glass cannon. Does insane damage with 2 stilettos, but if anything looks at her she dies. Fun times. BTW the stiletto you can buy in the beginning which grants jolting touch is an absolute beast even at the end game - you just have to enchant it.

[Stun]

This is flat out false, dexterity is far stronger than might as a dps increase on a rogue, and is far better at reapplying prone as well. Dexterity is a multiplicative modifier with all damage boosting stats, Might is additive with the majority of damage boosting stats, including reckless assault, sneak attack, deathblows, vicious weapon modifier, qualitiy modifier, and I believe critical modifier. 30% bonus damage from might to 300% damage is only 330% total dps whereas 30% attack speed is something like 20% dps increase, so would increase dps to 360%.

Not really. Do you know what the difference between 18 Dex and 10 dex is? Yeah. +24% action speed. This means that a pike-wielding rogue with 18 Dex will be able to hit something with his pike 4 times in the time it takes a pike-wielding rogue with 10 Dex to hit it 3 times.

 

Sounds positively game changing, right? Except that it isn't. Because fights don't last that long in PoE, unless you're a really Sh*tty player, Or you're soloing. And we're also dealing with Pikes. They're slow weapons. And no amount of Dex will make them fast weapons.

 

Maybe if we were talking about Stiletto dual-wielding it'd be a different matter, but again...we aren't.

Edited April 21, 2015 by Stun

[KDubya]

 

This is flat out false, dexterity is far stronger than might as a dps increase on a rogue, and is far better at reapplying prone as well. Dexterity is a multiplicative modifier with all damage boosting stats, Might is additive with the majority of damage boosting stats, including reckless assault, sneak attack, deathblows, vicious weapon modifier, qualitiy modifier, and I believe critical modifier. 30% bonus damage from might to 300% damage is only 330% total dps whereas 30% attack speed is something like 20% dps increase, so would increase dps to 360%.

Not really. Do you know what the difference between 18 Dex and 10 dex is? Yeah. +24% action speed. This means that a pike-wielding rogue with 18 Dex will be able to hit something with his pike 4 times in the time it takes a pike-wielding rogue with 10 Dex to hit it 3 times.

 

Sounds positively game changing, right? Except that it isn't. Because fights don't last that long in PoE, unless you're a really Sh*tty player, Or you're soloing. And we're also dealing with Pikes. They're slow weapons. And no amount of Dex will make them fast weapons.

 

Maybe if we were talking about Stiletto dual-wielding it'd be a different matter, but again...we aren't.

 

 

 

How can attacking four times with a pike instead of three not be a game changer? But it would if it was four stiletto strikes instead of three?

 

I am confused here. 24% more actions is 24% more actions regardless of what you are doing, it is always more and doing more is better than doing less.

 

Who would you rather be the rogue with 300% added base damage attacking at the base speed or be the rogue with the 300% added base damage attacking 24% faster?

 

Your battles finish before you get four attacks off? In that case I'd recommend upping the difficulty. Of course this is just the opinion of a ****ty player.

[SilchasRuin]

 

This is flat out false, dexterity is far stronger than might as a dps increase on a rogue, and is far better at reapplying prone as well. Dexterity is a multiplicative modifier with all damage boosting stats, Might is additive with the majority of damage boosting stats, including reckless assault, sneak attack, deathblows, vicious weapon modifier, qualitiy modifier, and I believe critical modifier. 30% bonus damage from might to 300% damage is only 330% total dps whereas 30% attack speed is something like 20% dps increase, so would increase dps to 360%.

Not really. Do you know what the difference between 18 Dex and 10 dex is? Yeah. +24% action speed. This means that a pike-wielding rogue with 18 Dex will be able to hit something with his pike 4 times in the time it takes a pike-wielding rogue with 10 Dex to hit it 3 times.

 

Sounds positively game changing, right? Except that it isn't. Because fights don't last that long in PoE, unless you're a really Sh*tty player, Or you're soloing. And we're also dealing with Pikes. They're slow weapons. And no amount of Dex will make them fast weapons.

 

Maybe if we were talking about Stiletto dual-wielding it'd be a different matter, but again...we aren't.

 

Alright so a few problems: the time length fights last is only relevant insofar as there is potential for attacks to be wasted. Damage, by the same token is also wasted whenever an attack deals more damage than an enemy has health. The only point relevantly in favor of might as opposed to dex is if on an arbitrarily small scale, additional might results in an enemy dying in 2 hits rather than 3, or the best case in which an enemy dies in one hit due to the additional damage from might. However, it is just as possible that an enemy should die in 2 hits no matter the might, making the second attack faster is thus a simple speed increase, However this is all functionally irrelevant, trash encounters that are as fast as you describe are not necessary to make decisions for unless your sole goal is to speed to the end of the game on easy as fast as possible. For PotD, the interesting fights that require min-maxing for greatest efficiency gains are boss battles which do not end in an attack or 2.

 

Now a simple comparison of dex to might for the best case for might on a rogue. I will treat each point of dexterity as worth a 2% dps increase, based on some information I've heard regarding dex, if dex functioned as, as written, a straight 3% increase in the number of attacks per second it would be worth 3% dps, which would exacerbate might's relative inefficiency even more. A level 1 rogue with sneak attack and a normal weapon with 10 might 18 dexterity does 174% of base dps. A level 1 rogue with sneak attack and a normal weapon with 18 might and 10 dexterity does 174% of base dps. So the best case for might on rogue is precisely equal to dexterity. Every additional damage modifier will cause dexterity's effective dps contribution to remain constant while might's will decrease as a percent gain. Example: a level 3 rogue with 10 might 18 dexterity with Reckless Assault, Two-Handed Style and a Fine weapon will deal 232% of base dps. A level 3 rogue with 18 might 10 dexterity and the same talents will deal 224% of base dps. As the number of increase damage modifiers grows, the contribution of might as a % of dps steadily declines while dexterity remains constant.

 

Now of course, for a rogue this is all relatively pointless from a practical perspective, excepting for the decision of which land, as there's no reason to not max both dexterity AND might, still from a theoretical perspective knowing that dexterity out-values might is useful,

Edited April 21, 2015 by SilchasRuin

[rheingold]

 

 

This is flat out false, dexterity is far stronger than might as a dps increase on a rogue, and is far better at reapplying prone as well. Dexterity is a multiplicative modifier with all damage boosting stats, Might is additive with the majority of damage boosting stats, including reckless assault, sneak attack, deathblows, vicious weapon modifier, qualitiy modifier, and I believe critical modifier. 30% bonus damage from might to 300% damage is only 330% total dps whereas 30% attack speed is something like 20% dps increase, so would increase dps to 360%.

 

Not really. Do you know what the difference between 18 Dex and 10 dex is? Yeah. +24% action speed. This means that a pike-wielding rogue with 18 Dex will be able to hit something with his pike 4 times in the time it takes a pike-wielding rogue with 10 Dex to hit it 3 times.

Sounds positively game changing, right? Except that it isn't. Because fights don't last that long in PoE, unless you're a really Sh*tty player, Or you're soloing. And we're also dealing with Pikes. They're slow weapons. And no amount of Dex will make them fast weapons.

Maybe if we were talking about Stiletto dual-wielding it'd be a different matter, but again...we aren't.

Alright so a few problems: the time length fights last is only relevant insofar as there is potential for attacks to be wasted. Damage, by the same token is also wasted whenever an attack deals more damage than an enemy has health. The only point relevantly in favor of might as opposed to dex is if on an arbitrarily small scale, additional might results in an enemy dying in 2 hits rather than 3, or the best case in which an enemy dies in one hit due to the additional damage from might. However, it is just as possible that an enemy should die in 2 hits no matter the might, making the second attack faster is thus a simple speed increase, However this is all functionally irrelevant, trash encounters that are as fast as you describe are not necessary to make decisions for unless your sole goal is to speed to the end of the game on easy as fast as possible. For PotD, the interesting fights that require min-maxing for greatest efficiency gains are boss battles which do not end in an attack or 2.

 

Now a simple comparison of dex to might for the best case for might on a rogue. I will treat each point of dexterity as worth a 2% dps increase, based on some information I've heard regarding dex, if dex functioned as, as written, a straight 3% increase in the number of attacks per second it would be worth 3% dps, which would exacerbate might's relative inefficiency even more. A level 1 rogue with sneak attack and a normal weapon with 10 might 18 dexterity does 174% of base dps. A level 1 rogue with sneak attack and a normal weapon with 18 might and 10 dexterity does 174% of base dps. So the best case for might on rogue is precisely equal to dexterity. Every additional damage modifier will cause dexterity's effective dps contribution to remain constant while might's will decrease as a percent gain. Example: a level 3 rogue with 10 might 18 dexterity with Reckless Assault, Two-Handed Style and a Fine weapon will deal 232% of base dps. A level 3 rogue with 18 might 10 dexterity and the same talents will deal 224% of base dps. As the number of increase damage modifiers grows, the contribution of might as a % of dps steadily declines while dexterity remains constant.

 

Now of course, for a rogue this is all relatively pointless from a practical perspective, excepting for the decision of which land, as there's no reason to not max both dexterity AND might, still from a theoretical perspective knowing that dexterity out-values might is useful,

For me it is good to know. Ironically enough if you are a min/maxer then it's probably irrelevant, but if you want a bit of room for role playing and dialogue it's very useful. I could potentially max dex and put strength around 14 which would leave points to put elsewhere, and frankly it probably wouldn't make much of a difference to the combat effectiveness which is great.

[SilchasRuin]

An interesting thing to note: 18 might 20 dexterity will outscale 20 might 19 dexterity when bonus damage is +235% or greater without including might. This means eventually Elf/Godlike will become a stronger racial choice for Rogue (melee) than Aumana, assuming that each point of dexterity is worth at least 2% dps increase. This is without taking into account racial abilities, but based on this, taking them into account, my suspicion is that Death Godlike may be the best end game Rogue race. The only question is how strong the 10% crit conversion on Hearth orlan is in the context of the high accuracy and the 20-30% hit to crit conversion rogue can get naturally, in conjunction with the numerous +crit damage modifiers that can be stacked.

Edited April 21, 2015 by SilchasRuin

[Stun]

Now a simple comparison of dex to might for the best case for might on a rogue. I will treat each point of dexterity as worth a 2% dps increase, based on some information I've heard regarding dex, if dex functioned as, as written, a straight 3% increase in the number of attacks per second it would be worth 3% dps, which would exacerbate might's relative inefficiency even more. A level 1 rogue with sneak attack and a normal weapon with 10 might 18 dexterity does 174% of base dps. A level 1 rogue with sneak attack and a normal weapon with 18 might and 10 dexterity does 174% of base dps. So the best case for might on rogue is precisely equal to dexterity. Every additional damage modifier will cause dexterity's effective dps contribution to remain constant while might's will decrease as a percent gain. Example: a level 3 rogue with 10 might 18 dexterity with Reckless Assault, Two-Handed Style and a Fine weapon will deal 232% of base dps. A level 3 rogue with 18 might 10 dexterity and the same talents will deal 224% of base dps. As the number of increase damage modifiers grows, the contribution of might as a % of dps steadily declines while dexterity remains constant.

Where are you getting your math from? This is not World of Warcraft. In PoE, a +24% action speed increase is NOT an increase in weapon swing speed (per second or otherwise). It's a decrease in recovery time between weapon swings. And that timer is paused when you make any movement (like moving from one opponent to the next). And by the way, Dexterity + The weapon is not the only factor here. The type of armor you're wearing also determines your action speed (in fact it's a bigger factor than dexterity)

 

But the bottom line is that you can't brush away the nature of the encounters in this game just because your math demands it (like you're doing here). Things simply cease being a matter of DPS when most fights see your rogue only needing to hit with his pike 2 or 3 times.

Edited April 21, 2015 by Stun

[SilchasRuin]

 

Alright so a few problems: the time length fights last is only relevant insofar as there is potential for attacks to be wasted. Damage, by the same token is also wasted whenever an attack deals more damage than an enemy has health. The only point relevantly in favor of might as opposed to dex is if on an arbitrarily small scale, additional might results in an enemy dying in 2 hits rather than 3, or the best case in which an enemy dies in one hit due to the additional damage from might. However, it is just as possible that an enemy should die in 2 hits no matter the might, making the second attack faster is thus a simple speed increase, However this is all functionally irrelevant, trash encounters that are as fast as you describe are not necessary to make decisions for unless your sole goal is to speed to the end of the game on easy as fast as possible. For PotD, the interesting fights that require min-maxing for greatest efficiency gains are boss battles which do not end in an attack or 2.

 

Now a simple comparison of dex to might for the best case for might on a rogue. I will treat each point of dexterity as worth a 2% dps increase, based on some information I've heard regarding dex, if dex functioned as, as written, a straight 3% increase in the number of attacks per second it would be worth 3% dps, which would exacerbate might's relative inefficiency even more. A level 1 rogue with sneak attack and a normal weapon with 10 might 18 dexterity does 174% of base dps. A level 1 rogue with sneak attack and a normal weapon with 18 might and 10 dexterity does 174% of base dps. So the best case for might on rogue is precisely equal to dexterity. Every additional damage modifier will cause dexterity's effective dps contribution to remain constant while might's will decrease as a percent gain. Example: a level 3 rogue with 10 might 18 dexterity with Reckless Assault, Two-Handed Style and a Fine weapon will deal 232% of base dps. A level 3 rogue with 18 might 10 dexterity and the same talents will deal 224% of base dps. As the number of increase damage modifiers grows, the contribution of might as a % of dps steadily declines while dexterity remains constant.

Were are you getting your math from? This is not World of Warcraft. In PoE, a +24% action speed increase is NOT an increase in weapon swing speed. It's a decrease in recovery time between weapon swings. And that timer is paused when you make any movement. And by the way, Dexterity + The weapon is not the only factor here. The type of armor you're wearing also determines your action speed (in fact it's a bigger factor than dexterity)

 

But the bottom line is that you can't brush away the nature of the encounters in this game (like you're doing here). Things simply cease being a matter of DPS when most fights see your rogue only needing to hit with his weapon 2 or 3 times

Dexterity is the only stat that actually affects action speed and recovery time, initially it only affected one of these, but currently it effects both of these. You should not be wearing armor on a midline rogue so armor is completely irrelevant. The timer also is not paused, but rather slows down (by a multiplier is my understanding, so higher dexterity will reduce the time it takes even while moving). As to where my math is from, it's from this thread: http://forums.obsidian.net/topic/72272-combat-mechanics-attack-speed-recovery/ although I rounded (in favor of might, I might add) for simplicity of calculations. Otherwise it's based purely on the fact that attack speed is a multiplier with all damage modifiers, while might is additive with most modifiers.

 

You completely ignored my first point. If the rogue hits 3 times and kills the mob in 3 seconds, that's markedly better than hitting the mob 3 times and killing it in 4 seconds. Further, because dexterity affects the time between the rogue's next attack on another mob, you can calculate the entire set of enemies the rogue fights as if they are single enemy with one large health bar, with the exception that over damaging will reduce the efficiency of any attack, and if the rogue has already finished recovery time during the movement between targets, which would actually make a paused timer better for dexterity, not worse. If every fight you're going against your rogue is barely attacking you're playing on too low of difficulty and overleveled for encounters. Also again, while marginal increases in efficiency are wonderful for small scale trash fights that might actually end fast, there are larger boss and trash battles where fights do not end very quickly, and where the dps gains from dexterity are very very apparent.

 

Finally you haven't addressed at all just how little might does on a rogue compared to dex. The chance of might causing your character to actually fulfill the conditions in which it would actually be better than dex (short encounter where might causes the amount of damage needed to go from 4 swings to kill to 3) is laughably low, as especially at higher levels might's actual contribution to damage becomes on the order of 9%(at 235% bonus damage, the addition of 24% damage from 18 might is ~9%) rather than the 24% shown.

Edited April 21, 2015 by SilchasRuin