Jump to content

Welcome to Obsidian Forum Community
Register now to gain access to all of our features. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more. If you already have an account, login here - otherwise create an account for free today!
Photo

Paladin Leader in 2.0

leader paladin build 2.0

  • Please log in to reply
49 replies to this topic

#41
Ymarsakar

Ymarsakar

    (5) Thaumaturgist

  • Members
  • 550 posts

Retaliate does benefit from the DR bypass talent, although I'm not sure if it is just the melee one.

 

Attack speed is basically the same as -50 recovery = 30-33% attack speed, it delays the next attack. Dexterity affects the actual speed at which attacks are made. And I think that's the only one that does it. It may affect recovery as well.

 

Two handed swords don't benefit as much from DR penetration as faster weapons do.

 

Here's a quick test, doing some more. Identical paladins level 1, except one has a dex of 4 vs a dex of 18 using -50 recovery brigandine and maces vanilla. By the time the seventh attack of the dex 18 paladin had landed on the other, the dex 4 paladin had made 5 attacks. No interrupts were counted. 18+24=42% difference in action speed, for a difference of 7/5 attacks with only a few frames of difference in the last attack landing.

 

For morningstar two handed it's about the same 7.5/5 split in attack numbers. Now it should get interesting with vulnerable attack and various other stuff on...

 

15 attacks vs 10 attacks morning stars, 18 dex vs 4 dex + vulnerable on.

 

I got something like 7.5/5 the first time but it was a few frames off so I counted it up to 10 on the right side. 4 dex side interrupted 3 times, so that might be an issue as well affecting overall hits. Now we'll do the reverse with vulnerable attack on the 18 dex side. I'll bump up resolve on both and see if it'll stop the interrupts.

 

6.5/5

13/10

 

6.5/5 for mace/shield+vulnerable attack vs mace/shield.

 

Taking a different methodology single weapon attacks are 2.8 seconds vs 5 seconds.

 

Vulnerable on, it's about 3 seconds per attack vs 5 seconds.

 

Tentative conclusion is that -20% attack speed hurts people with higher dex or rate of fire builds. Now moving unto armor differences.

 

Naked 18 dex 4 attacks in 10 seconds (using a better method here by casting a spell on somebody and using that as a time stop watch, my old technique with cipher) using morningstar +vulnerable -20% attack speed. Almost exactly 12 attacks in 30 seconds.

 

Morningstar in 30 seconds, we got 13 attacks with 18 dex and naked. 8.33% difference in total attacks after 30s between using vulnerable and no vulnerable naked with two handed morning star.

 

4 dex 9.3 attacks every 30 seconds. With vulnerable on, it's about 8.4 That's a 8.7% difference.

 

-20% attack speed will probably cripple anybody using fast weapons though, or if their attack speed is buffed to 50%+ like the wizard's alacrity or via potions, while naked.

 

I changed the methodology on the last few experiments, so will have to redo the brigandine armor sets, although my tentative conclusion is that attack speed debuff/buff matters less if you are already at -50 recovery.

 

I also tested vulnerable attack with the fire retalation at 50% endurance, and it does penetrate through to people around you if you turn on that modal.

 

The dexterity or action speed, makes a much more accurate representation of how many attacks you can get per x time. Of course this doesn't factor in the arquebus and arbalest reload times, but dexterity affects reload times as well.


Edited by Ymarsakar, 14 September 2015 - 08:47 PM.


#42
KDubya

KDubya

    (10) Necromancer

  • Members
  • 1543 posts
  • Deadfire Backer

I've always actually wondered about Vulnrable Attack, and that -20% speed is factored in. If it's just as it says, a straight you'll be hitting the mob 20% slower, it's a pretty poor talent in a lot of situations. From a pure damage perspective, you average hit will need to do less than 20 damage (including enchant buffs such as lash) to make it even break even. I'm not sure exactly how much lower your average hit would need to be justify the opportunity cost of losing an alternate Talent choice, as well. If that's how it operates, it's certainly not useless, but really only for for fast weapons, early game, or classes with very few multipliers.

 

Early game 2handers, for example, you're just doing 1.0, maybe 1.3 with a fine weapon+2hd style, and few accuracy buffs, so breaking that 20 damage threshold average will be rough. Late game, you'll have that number closer to 2.15, as well as buffed accuracy and often a level advantage over mobs, making Vulnrable attack actually lower your damage. That's also not factoring in the lower chance of per-hit things, like a rangers stunning blow, Tall Grass knockdown (with carnage fun times) or the frisky things the Soulbound weapons.

 

On the other hand, it does just say Attack Speed, and if it doesn't also apply the penalty to recovery frames, it'd change things greatly, making the talent weirdly better suited to a plate mail wearing 2 hander, whose got a massive recovery penalty compared to a fast weapon dual wielder in light armor. For both though, the talent would quickly move into must have range, as the recovery penalty is a much bigger part of the attack speed equation than the actual attack itself. I think it'd be really weird if it works this way, but it's the only way I see Vulnrable attack being a good talent choice for more than a few specific builds, or something taken early and later respec'd out of. (Though treating stuff that way always feels a little cheezy to me.)

 

Also on the note of not having a clue how Vuln. Attack works, what does it apply that -DR too? Would it also apply to say, a Fire Godlike Barbarian tank with Retaliatex2 on? If so, that'd be a lot of extra value from it.

 

 

For things like this I use my Fight Club to test out different builds.

 

Works best with a high level so you can higher higher level mercs. Also not recommended for Trial of Iron, best to save before hand and then revert back when you go back to your real game.

 

The combat is all conducted out of combat so there are some quirks. Your endurance will regen constantly so the only thing that matters is accumulated health loss. This has the side benefit of extending combat, a few crits or misses would make a difference when you are dealing with only 200 endurance but with 1500 or so health it tends to average out pretty well. Activated abilities won't be able to be used. All attacks from a Rogue will get sneak attack bonuses.

 

1.) Free up four character slots for mercs

 

2.) Hire two identical "Targets" these will stand there and get attacked. High con will give a larger health pool to test against. Monks might be better than Barbarians as their base deflection is higher so you can simulate more challenging enemies. By taking deflection talents, shield use and armor worn you can simulate various DR and deflection values.

 

3.) Hire your two test subjects. I like to keep everything the same except for what I want to test. If I want to test dual wield vs two hander I'd make identical guys that took only those different talents. If the two builds are wildly different you won't know where the difference came from.

 

4.) Have your test subjects attack their own target. It helps to set the speed to fast for this.

 

5.) Review the results, change another variable and go again.

 

Based on tests like this I've come to really like the DR penetration talent. Even with Estocs with their natural DR penetration the talent is better to have once the DR goes past 10 or so.



#43
Ymarsakar

Ymarsakar

    (5) Thaumaturgist

  • Members
  • 550 posts

The only feasible situation I've heard where the DR penetration talent is bad is when the extra attack or so is worth whatever damage the talent would have penetrated. Usually that only applies to things like sabres, because their base attack is higher or exceptional/superb weapons. It's definitely more reliable and up front damage, as even a graze might do more damage depending on enemy dr and base attack dmg. Whereas an extra attack might just miss or graze. So it would be that extra graze plus whatever faster attacks for interrupts you needed earlier, against the extra damage from DR bypass.

 

With faster weapons and higher stacked attack speed buffs, it'll be more than 1 extra attack of course.

 

New stuff

 

4 DEX 8.5 over 30s vs 10 over 30s No ARMOR
 
18 DEX 12 over 30s vs 14 over 30s No Armor
 
It's Vulnerable attack on vs off.
 
15% difference in attack speed.
 
14.28% difference in attack speed for 18 dex no armor.
 
 
In this test with estocs, there wasn't much difference between high dex or low dex build. Still got a 14-15% difference in attack speed, as in difference of number of attacks over x time. Some of the numbers are a little bit fuzzy because they aren't exact and the timer sometimes ended after or just before an attack. And there's always that .1 or .2 of time difference for when the clock starts, I can't usually time it exactly to 30s
 
Second trial with no armor vulnerable attack off was 9 attacks over 30s no armor 4 dex vs 13 attacks over 30s no armor 18 dex.
 
I run the tests through a lot of variables just to see if there's some extraneous factors I'm not aware of affecting things, like the speed of estocs vs morningstars. These ones are probably more accurate because I've taken the interrupt issue out, by having a punching target up that doesn't hit back.
 
Overall, accounting for some percentage errors due to the number of attacks I might have miscounted, we had 8-9% difference and 14-15% difference, so the actual difference for 20% attack speed is probably closer to 10%, modified by the recovery of the armor a little, or maybe that's just statistical error. Now the extra fun begins for when stack attack speed up past 50% and more.
 
A great place to test things in combat is at an inn inside (so you can hire and rest) and then find a chokepoint, then aggro everything. That'll start the combat timer but if you have a bottleneck, you can just tank them while your other characters do tests on each other. Or as I prefer, on the inn residents.
 
Remember the opening sequence of Baldur's Gate 2 the Shadow of Amn? I was a little bit disappointed the Caed Nua dungeon didn't have something like that, except we would do the testing on the "subjects". Then again, there are alternatives.

Edited by Ymarsakar, 15 September 2015 - 08:05 AM.


#44
Tennisgolfboll

Tennisgolfboll

    (4) Theurgist

  • Members
  • 305 posts
Can you sum up your findings in some short cliff notes

#45
Ymarsakar

Ymarsakar

    (5) Thaumaturgist

  • Members
  • 550 posts

Which ones are you interested in Tennis?



#46
Tennisgolfboll

Tennisgolfboll

    (4) Theurgist

  • Members
  • 305 posts
The basics about how dex in armor effects attacks.

3 dex vs 18 dex with 50% recovery
And 3.dex vs 10 dex? 50% recovery armor again
What is the difference?

Edited by Tennisgolfboll, 15 September 2015 - 10:07 AM.


#47
Ymarsakar

Ymarsakar

    (5) Thaumaturgist

  • Members
  • 550 posts

For the attributes, they seem very close to the listed stats.

 

4 dex vs 18 dex is a difference of 42% action speed. And the number of attacks in armor closely relates to that, as in you actually do get 40% more attacks. I think this is more noticeable with fast weapons vs slow two handed weapons because the animation is just faster and you just notice yourself getting more attacks in 30s. Whereas the same works with two handed weapons too, it's just sometimes hard to notice that 10%-20% difference most people get with balanced stat builds. Since most people usually don't run builds with min dexterity and then another guy with max dexterity, they're usually running 10 dex or 15 dex, something like that.

 

From previous tests using older builds like 1.06 and 1.4, action speed also affects the reload time on blunderbusses and such weapons, so it's like total frames over the sequence is reduced by the action speed.

 

I haven't tested with 10 dex substantially. 42% is the difference between 18 and 4, so the difference between 4 and 10 would be 6. About 18% action speed difference if I have that right. Or 10 vs 18 would be 8x3 or 24% action speed difference.

 

Generally might, perception, and dexterity all together multiply their effects and bonuses. So based on previous combat experience, if I had +30% attack damage multiplied by 10 attacks, it might be 13 dmg x 10=130 points of damage. Whereas if dex boosts the attacks to 12 over same time, that's 13x12=156 dmg.

 

But these are just tentative estimates, for in game results I think more testing with different builds is needed to see exactly how much difference it would make for various different offensive or defensive builds. Certain spells have very low recovery time frames, so people have noticed that wizards/druids/priests with high attack speed buffs can start spamming a lot of spells, which aren't even fast cast.

 

So it is as they listed it, 3 points of action speed per dexterity, up or down. There doesn't seem to be any exponential curves going on at higher or lower levels. I think that would normally depend on the weapon used, since the weapon speed may have a high limit and a low limit for attack rates. Speculating at the moment, but I think 18 dex probably almost entirely eliminates the debuff from the -50 recovery armor. Dexterity action speed affects all the frames of every action, including recovery it seems. Whereas the recovery just affects the attack speed, the delay between attacks. So dexterity affects spells, dual wielding, attacks more as it reduces the frames on each of those animations. Whereas for two handed swords or slower weapons, dexterity doesn't reduce the frames as much as the recovery portion is being reduced. That's because the slower weapons attack less, so they have less frames to be reduced. Well that's my explanation for why a high armor low dex two handed warrior doesn't get much dps malus.


Edited by Ymarsakar, 15 September 2015 - 10:55 AM.


#48
Tennisgolfboll

Tennisgolfboll

    (4) Theurgist

  • Members
  • 305 posts
Ok good. Thanks for the clarification i misunderstood your findings before.

#49
Ymarsakar

Ymarsakar

    (5) Thaumaturgist

  • Members
  • 550 posts

They were some raw data I put down as I was going to sleep and needed a record of it. If anyone needs to know something specific, they can always send me a message, running experiments can be fun. Assuming they don't want to know the methodology in order to duplicate the experiment (science).



#50
Teioh_White

Teioh_White

    (4) Theurgist

  • Validating
  • 372 posts

I think this is a great example of the tool tips being grossly insuffcient to give the player an accurate idea of what is actually happening with the numbers.







Also tagged with one or more of these keywords: leader, paladin, build, 2.0

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users