435 réponses à ce sujet

[GhoXen]

Introduction

 

Due to a severe lack of precise information on what the various attributes do in-game, I've resorted to testing and theory-crafting. The goal of this thread is to discover what each and every primary, offensive and defensive attribute does. A future goal is to also find out the soft & hard caps for each attribute, if one is present.

 

This guide will be updated overtime as more tests are done and more information is discovered. The first post will contain the specific functions of the attributes, while the second post will contain the methods, assumptions and results of various tests used to determine information provided in the first post.

 

The guide will be written to show what "1 attribute" provides. Further, primary attributes will only have an effect for points above the base value of 10. E.g. "Strength" would indicate what "1 Strength provides for each point above 10".

 

 

Primary Attributes

Spoiler

Strength

• 0.5% Attack (Warrior only)
• 1% Guard Damage Bonus

Dexterity

• 0.5% Attack (Rogue only)
• 1% Crit Damage Bonus

Magic

• 0.5% Attack (Mage only)
• 1% Barrier Damage Bonus

Cunning

• 0.5% Crit Chance
• 0.5% Range Defense

Willpower

• 0.5% Attack
• 0.5% Magic Defense

Constitution

• 5 Health
• 0.5% Melee Defense

 

Offensive Attributes

Spoiler

Attack %

• Increases damage of basic attacks and most* abilities by 1%.
• Damage bonus applied after deduction by enemy armor.
• (Weapon - Armor) * (1 + Attk%), NOT Weapon * (1 + Attk%) - Armor.

*Note: some abilities may calculate damage differently and fail to apply the bonus properly. So far, Longshot's bonus damage based on range is not affected by attack %. I'll test more abilities in the future, as a lower priority task. Please let me know if you find any ability that doesn't benefit or doesn't benefit fully from attack %.

 

Guard Damage Bonus %

• Increases damage of an attack against Guard by 1%.
• Applies before damage reduction by armor.
• Stacks multiplicatively with other damage bonus attributes.

Armor Penetration %

• Ignores 1% of target's armor rating when dealing physical damage. 
• Example. 20% AP would penetrate 10 armor, if the enemy has 50 armor in total.
• For enemies, 1 armor rating decreases their damage taken by 1 point. If 10 armor ratings are penetrated, then they'd take 10 extra damage when attacked.
• Abilities' damage multipliers apply, so an ability that deals 300% weapon damage would inflict 30 extra damage even though only 10 armor ratings are penetrated.
• Does not stack with bonus damage attributes (crit, flank, etc.). I.e. AP gives the same amount of extra damage on both a critical and a non-critical hit.
• Does not affect non-physical damages.
• Does not affect targets with 0 armor (i.e. mage) or barriers, and works poorly against low armor targets (i.e. archer, assassin, shade, terror, etc.).
• Has a fairly complex non-linear comparison to Attack %. May be superior than Attack % on dagger rogues, and to a lesser degree bow rogues, when fighting dragons. Full spreadsheet here: https://mega.co.nz/#...16cRqS4Gqaalg2g

Barrier Damage Bonus %

• Increases damage of an attack against Barrier by 1%.
• Stacks multiplicatively with other damage bonus attributes.

Critical Damage Bonus %

• Increases damage of critical attack by 1% more than non-critical attack.
• I.e. default bonus of 40% means critical attacks hit for 1.4 times the damage of non-critical attacks.
• Applies before damage reduction by armor.
• Stacks multiplicatively with other damage bonus attributes.

Critical Chance %

• Increases chance to inflict a critical attack on every target hit by 1%.
• When attacking multiple targets, each hit is independently capable of triggering critical hits.
• No internal cooldown.

Weapon Damage

• Base damage per hit.
• Every attribute applies to weapon damage, which serves as the base damage.
• Generally a more useful indicator of overall damage output than DPS. DPS shows only base damage output of basic attacks, while weapon damage affects ability damage output.
• Expressed as "Mainhand Damage" or "Offhand Damage".

Bleed on Hit %

• BUG: This stat does not work. Will test again when it's patched.

Stagger on Hit %

• Increases chance to stun an enemy for 3 seconds on hit by 1%.
• The stun allows triggering of combos.
• No internal cooldown.
• No "Stunned" floating text like stuns from abilities.

Heal on Kill %

• Increases healing received on landing a killing blow by 1% of max health.
• E.g. 25% Heal on Kill would heal your character for 25% of their max health on every kill.
• Must land the killing blow personally via abilities (including DoT) or otherwise.
• No internal cooldown.

Flanking Damage Bonus %

• Increases damage of flank attack by 1% more than non-flank attack.
• I.e. default bonus of 25% means flank attacks hit for 1.25 times the damage of non-flank attacks.
• Applies before damage reduction by armor.
• Stacks multiplicatively with other damage bonus attributes.

Cooldown Modifier % (WIP)

• Will test if it stacks multiplicatively with abilities.

 

Defensive Attributes

Spoiler

Magic Defense %

• Reduces final damage taken from magical attacks by 1%.

Melee Defense %

• Reduces final damage taken from melee attacks by 1%.
• Applies after damage reduction from armor.

Ranged Defense %

• Reduces final damage taken from melee attacks by 1%.
• Applies after damage reduction from armor.

Cold Resistance %

• Reduces final damage taken from cold attacks by 1%.

Electrical Resistance %

• Reduces final damage taken from electrical attacks by 1%.

Fire Resistance %

• Reduces final damage taken from fire attacks by 1%.

Spirit Resistance %

• Reduces final damage taken from spirit attacks by 1%.

Guard

• Functions as extra health. Enjoys damage mitigation, unlike Barrier.
• Max Guard = 25% of Max Health.
• BUG: Vanguard's "Untouchable Defense" does not increase max guard at all, contrary to the stated 25%.
• Certain abilities confer extra bonus, e.g. 20% armor, when Guard is present.
• Guide on how to generate Guard here: http://forum.bioware...generate-guard/

Armor Rating

• Reduces damage by X points when attacked by melee or ranged physical attacks.
• X = 1 (MP); 2 (Nightmare) (will test Casual, Normal and Hard soon)
• Applies before Melee/Ranged Defense %.
• Some enemies may possess armor penetration properties.
• Note: for enemies' armor rating, 1 armor only decreases their physical damage taken by 1 point even on Nightmare.

Armor Rating Front

• Functions just like regular Armor Rating, but only used if attacked in the front.
• Armor rating on a shield increases this stat. Ergo, "armor rating" and "armor rating front" have identical values for all except shield users.
• BUG: Does not benefit from abilities giving +% armor ability buffs, unlike regular armor rating. However, + armor potion applies the bonus properly.

Bleed on Being Hit %

• BUG: This stat does not work. Will test again when it's patched.

Stagger on Being Hit %

• Increases chance to stun an enemy for 3 seconds when being hit by them by 1%.
• The stun allows triggering of combos.
• No internal cooldown.
• No "Stunned" floating text like stuns from abilities.
• Can trigger on ranged or direct magical attacks (i.e. staff), if the attacking enemy is within melee range.

Healing Bonus % (WIP)

• Should be straight forward. Low priority.

 

Mastercraft/Special Item Effects

Spoiler

Enhanced Ability Ring

• Increases the prescribed ability's duration or damage by 30%.
• Two identical enhanced ability rings do not stack.

Mastercraft Effect Stacking

• Identical mastercrafts do not stack!
• Different tiers of the same type of mastercraft do stack!
• E.g. +15 stam x2 will give only +15 stam; +15 stam and +10 stam will give +25 stam.

2% Chance Shield Bash (FT Serpentstone, variant)

• Casts a full-spec shield bash dealing 300% weapon damage as physical; capable of knockdown; bonus against Guard.
• Triggers on any type of weapon attack, including bow and staff.
• Operates exactly the same as a warrior's Shield Bash, including the damage multiplier.
• Due to the drastic low proc chance of  2%, other mastercraft procs provide greater damage output.
• This effect only exists on  variant of FT Serpentstone. The other variant of FT Serpentstone has Berserk instead.
• Higher tiers increase proc %.

10% Chance to Cast Chain Lightning at 50% Weapon Damage (FT Onyx)

• Casts a chain lightning dealing 50% weapon damage as electrical; hits up to 4 enemies within 5m; inflicts Shock for 8s.
• Aside from the damage multiplier, the mastercraft operates exactly the same as a mage's Chain Lightning.
• Higher tiers increase damage multiplier, but not proc chance.

10% Chance to Cast Walking Bomb at 50% Weapon Damage (FT Cotton)

• Casts a walking bomb, damaging the enemy for 50% weapon damage every second for 5 seconds, and then explodes for 50% weapon damage in a 5m AoE; capable of knockdown.
• Aside from the shorter duration and lower damage multiplier, the mastercraft operates exactly the same as a mage's Walking Bomb.
• Higher tiers increase damage multiplier, but not proc chance.

Increase Max Stamina

• Increases max stamina and stamina regeneration by the prescribed amount.
• Works well together with reduced ability cost mastercrafts.

Heal X% Damage Over 10 Seconds

• Heals the prescribed % of the most recent health damage taken over 10 seconds.
• The heal triggers only on health damage, not damage to Barriers or Guard.
• Does not stack with itself, will immediately start healing the most recent damage and override healing from previous occurrences of damage.
• Example: If hit for 1000 dmg, will start healing 15 dmg/s. After only 3s, if then hit for another 200 dmg, will start healing for 3 dmg/s instead from that point onwards, until the 10s pass or until attacked again within 10s.
• The healing does not generate floating text.

 

The Damage Formula

Spoiler

Player Formula:

[WeaponDmg * (1 + CritDmg%) * (1 + FlankDmg%) * (1 + BarrierDmg%) * (1 + GuardDmg%) - Armor *(1 - ArmorPen%)] * (1 + Attk%) * AbilityMultiplier%

• CritDmg%, FlankDmg%, BarrierDmg%, GuardDmg% and AbilityMultiplier% only apply when applicable.
• Armor applies only to physical attacks.
• Damage Calculator: https://mega.co.nz/#...16cRqS4Gqaalg2g (contains attack % vs AP % spreadsheet)

Enemy Formula:

[BaseDmg * (1 + CritDmg%) * (1 + FlankDmg%) - Armor * 2 *(1 - ArmorPen%)] * AbilityMultiplier% * (1 - Defense%) * (1 - ElementalDef%)

• Defense% can be melee, ranged or magic, where applicable.
• ElementalDef% & CritDmg% apply only when applicable.
• Armor applies only to physical attacks.

 

DPS Attribute Ranking

Spoiler

Assumptions: 

• Weapon damage 150; enemy armor 50. Existing stats: 10% Attack, 0% AP, 40% Crit Dmg Bonus, 25% Flanking Dmg, 10% Crit Chance.
• A well-balanced party, with removals dedicated to Guard (Shield Bash, War Horn, etc.) and Barriers (Dispel).
• Ranking is for a pure DPS character. Not a tank/support who may want Stagger % to allow the party easy combos, or crit % to trigger party-wide effects.
• Willpower's auxiliary effect of magic defense is considered, since it's simply too vital on harder difficulties for all characters.
• The ranking applies to only very general situations. For specific party composition, enemies, locations, etc. It would be best to use the damage calculator (listed above) for more precise min-maxing.
• Flanking damage % is best stat If a DPS in a flanking position. Otherwise it's worthless.

Crafting Priority:

3% Flanking Dmg > 1.75% Attack > 3% Armor Pen > 1.75% Crit Chance = 1.75 Dexterity (rogue) >1.75 Willpower > 1.75 Magic (mage) = 1.75 Strength (warrior) > > 1.75 Cunning > 3% Crit Dmg > 3% Barrier Dmg > 3% Guard Dmg

 

Gear Equivalents:

• 0.67% Flanking Dmg
• 1% Attack (benchmark)
• 2% Armor Pen
• 1.6% Crit Chance
• 1.5 Dexterity
• 2 Willpower/Strength/Magic
• 3.2 Cunning
• 7% Crit Dmg

 

Base Attributes

Spoiler

The below stats remain fixed regardless of level. They will only be affected by equipment bonuses and ability passives. Health increases marginally with level, and in terms of base health warrior > rogue > mage.

• Primary Attributes: 10 in all attributes. Strength increases by 0.5/level for warrior, Dexterity increases by 0.5/level for rogue, Magic increases by 0.5/level for mage.
• Crit Chance: 5%
• Crit Damage Bonus: 40%
• Flank Damage Bonus: 25%
• Armor: 38 (WIP)

Note: since in this game it's not possible to fight with fists, if no weapons are equipped the character will fight using "default weapons" that don't take item slots. These default weapons distort base stats. Namely, rogue's default dual daggers confer 18% crit damage bonus, while warrior's default sword & shield confer 3% attack.

[GhoXen]

Attack Tests

Spoiler

Attack Test 1

Level 13 rogue, base stats, Longshot unlocked. Longshot is used at greater than 15m range for bonus damage.

Setup 1: Bow 70 DPS 62 dmg per hit, 3% attack (base stats).

Setup 2: Bow 70 DPS 62 dmg per hit, 13% attack.

Target: Level 7 Ram (0 armor)

Test: using 10 basic attacks and 10 max range Longshots on armorless enemy. Critical hits are ignored.

 

Results:

Equipping the bow did not change "Main-Hand Damage" on attributes screen.

Setup 1: Basic 64.1 dmg (average); Longshot 513.6 dmg (average)

Setup 2: Basic 69.4 dmg (average); Longshot 525.4 dmg (average)

 

Conclusion:

Attack % appears to provide a flat % increase on basic attacks. Effect on abilities is rather confusing.

 

Attack Test 2

Level 13 rogue, base stats, Longshot and Poisoned Weapon unlocked. Longshot is used at greater than 15m range for bonus damage

Setup 1: Bow 132 DPS 117 dmg per hit, 3% attack (base stats).

Setup 2: Bow 132 DPS 117 dmg per hit, 40% attack.

Test: using 10 basic attacks, 10 max range Longshots, 10 ticks of Poison DoT on armorless enemy. Critical hits are ignored.

 

Results:

Setup 1: Basic 118.4 (average); Longshot 956.6 (average); Poison/s 47.2 (average)

Setup 2: Basic 165.7 (average); Longshot 1018.6 (average); Poison/s 80.2 (average)

 

Conclusion:

It seems Attack % provides a flat increase in damage output in most cases. However, damage of Longshot appears to be calculated differently. It seems the formula for Longshot's damage is something along these lines: WeaponDmg*(1 + Attack%) + WeaponDmg*RangeBonus%.

 

Attack Test 3

Level 13 warrior, base stats, basic attack. A short test to determine if attack % applies before or after enemy armor.

Setup 1: Sword 56 DPS 50 dmg per hit, 0% AP, 3% base attack.

Setup 2: Sword 56 DPS 50 dmg per hit, 0% AP, 13% attack.

Target: Level 11 Bear with 45 armor.

 

Results (mean):

Setup 1: 5.1.

Setup 2: 9.2.

 

Conclusion:

In theory, the Bear's 45 armor would reduce the weapon's damage by 45, lowering it down to 5. This appears to be the case in Setup 1. However, Setup 2's result can only be explained if Attack % was applied before armor. In which case 13% attack would increase the weapon's damage from 50 to 56.5, which is then reduced by 45 down to around 11.5. If Attack % was applied after armor, then there shouldn't be any detectable difference between the two setups.

 

What this means, when taken together with the AP test down below, is that 1% attack can in fact provide more than 1% damage when armor is involved, and when the weapon damage is high enough.

 

Attack Test 4

After suspecting that my previous test may be erroneous due to RNG, a new test is conducted to test the relationship between attack & armor.
Setup: level 14 rogue, bow 51 dmg per hit, 0% AP, 37% attack.

Target: level 11 rage demon with 45 armor.

Test: attack the target both while it's barrier is up (should be full damage) and while its barrier is down. 10 hits each.

 

Result (mean):

Health: 6.5.

Barrier: 68.4.

 

Conclusion:

The old theory of damage calculation clearly fails. However, based on the new theory of attack applying after armor mitigation, (51-45)*1.37 = 5.48, things start to make sense. Surprisingly, or I suppose unsurprisingly, using the new formula does not distort the observations in other tests involving attack %, which is how I did not discover the error in the other tests indirectly using attack %.

 

Armor Penetration Test

Spoiler

AP Test 1

Simple test of attacking armorless enemy (level 7 Ram) with and without AP. Posting result is rather pointless, since it's as expected - AP doesn't do anything whatsoever when the enemy has 0 armor.

 

Enemy Armor Test

First, I want to ensure that enemy armor works the same as player armor.

Level 13 warrior, base stats, basic attack.

Setup: Axe 210 DPS 186 dmg per hit, 0% AP.

Target 1: Level 11 Archer with 18 armor.

Target 2: Level 11 Bear with 45 armor.

Target 3: Level 11 Knight with 54 armor.

 

Results (mean):

Target 1: 136.

Target 2: 143.1.

Target 3: 179.9.

 

Conclusion:

It appears that enemy armor is not as effective as player armor. Instead of negating 2 damage, every point of enemy armor seems to negate only around 1 damage.

 

AP Test 2

Level 13 rogue, base stats, Longshot and Poisoned Weapon unlocked. Longshot in theory does not go through armor, while Poison in theory does.

Setup 1: Bow 81 DPS 72 dmg per hit, 3% base attack.

Setup 2: Bow 81 DPS 72 dmg per hit, 3% base attack, 15% AP.

Target: Level 11 Bear with 45 armor.

Test: Using 10 basic attacks, 10 max range Longshots and 10 Poison ticks on enemy. Critical hits are ignored.

 

Results (mean):

Setup 1: Basic 29; Longshot 222.8; Poison/s 28.9

Setup 2: Basic 34.9; Longshot 279.2; Poison/s 29.7

 

Conclusion:

As expected, elemental attack passes through armor entirely and isn't affected by AP. Clearly, Longshot also benefited from AP, so the effect of AP must have come later in the equation than attack.

However, it does not appear that 15% AP simply removed 15 armor. Basic attacks received an increase of roughly 5.9 damage, which can be explained if 15% of the 45 armor, 6.75 armor, was removed. This suggests that AP simply ignores the stated percentage of target's armor, regardless of how high or how low it is  (as long as it's not 0).

Longshot received an increase of roughly 56.4 damage, indicating that the ability's damage multiplier was applied after armor penetration. However, Longshot's formula is somewhat unusual, so I'll use a different ability in the next test.

 

AP Test 3

Level 13 warrior, base stats, unlocked Block and Slash.

Setup 1: Axe 210 DPS 186 dmg per hit, 0% AP.

Setup 2: Axe 210 DPS 186 dmg per hit, 21% AP.

Setup 3: Axe 210 DPS 186 dmg per hit, 53% AP.

Target: Level 11 Bear with 45 armor.

Test: Using 10 basic attacks and 10 Block and Slash counters against enemy.

 

Results (mean):

Setup 1: Basic 143.1; Counter 293.8.

Setup 2: Basic 152.1; Counter 314.7.

Setup 3: Basic 167.2; Counter 345.5.

 

Conclusion:

The assumption from AP Test 2 appears to hold true. 21% AP would in theory ignore 9.45 armor from 45 armor, and indeed, Setup 2 received an increase of roughly 9 damage over Setup 1. 53% AP would in theory ignore 23.85 armor, and this once again appears to be true as Setup 3 inflicts around 24.1 more damage than Setup 1.

Once again, AP appears to have been included in the calculation of ability damage. Block and Slash deals 200% weapon damage, and here it seems the damage bonus from AP was also doubled compared to basic attack.

 

AP Test 4

This test aims to determine the effectiveness of AP between different ranges of armor ratings. Level 13 warrior, base stats, basic attack only.

Setup 1: Axe 210 DPS 186 dmg per hit, 0% AP.

Setup 2: Axe 210 DPS 186 dmg per hit, 21% AP.

Setup 3: Axe 210 DPS 186 dmg per hit, 53% AP.

Target 1: Level 11 Archer with 18 armor.

Target 2: Level 11 Bear with 45 armor (data from previous test).

Target 3: Level 11 Knight with 54 armor.

 

Results (mean):

Target 1: S1 179.9; S2 183.3; S3 190.9.

Target 2: S1 143.1; S2 152.1; S3 167.2 (data from previoius test).

Target 3: S1 136; S2 150.2; S3 166.9.

 

Conclusion:

The theory from the previous tests once again hold true. The AP % simply ignores the prescribed % of enemy's armor, whether it's 54 or 18. As expected, the higher the enemy armor the more effective AP is. Unlike other attack stats which scale with the player's weapon damage, and become more effective as the player gets increasingly stronger weapons, AP scales with the enemy's armor, and becomes more effective as enemies get increasingly higher armor as they level up.

 

Detailed analysis of Attk vs AP below.

 

Attk vs AP Theorycrafting

Spoiler

Comparison Spreadsheet: https://mega.co.nz/#...16cRqS4Gqaalg2g (contains damage calculator)

 

The relationship between attack & AP is far far away in a universe that's not linear. I highly recommend using the calculator and enter your own values to see whether you should gear for AP or not.

 

Keep in mind that AP doesn't interact well with crit/flank, doesn't affect elemental damage, doesn't affect barriers, and works poorly against mage (0 armor) or light enemies such as archer, assassin or terror demon (50% of regular armor).

 

When the enemy's armor is roughly 33% your weapon damage, AP % will start to become more effective than Attk % in terms of non-crit basic attacks. When it's around 50% your weapon damage, AP% will no longer suffer compared to attack% when it comes to base 40% crit attacks (excluding flank crits).

 

Based on those stats, a dagger rogue (lowest weapon damage) would in theory benefit the most from AP. However, that is only the case if the dagger rogue in question is not flanking most of the time, has low crit % and crit damage, and if the dagger rogue is not attacking mages and archers as a priority. And to be honest, there is only one situation where that will happen - fighting dragons. For general combat & exploration, and against regular humanoid bosses who enjoy summing adds, even a dagger rogue, who has the most potential to benefit from AP, will want an Attack % set instead of an AP set. However, for particularly difficult dragon encounters, an AP set may be worth investing.

 

A bow rogue or SnS warrior can also benefit from AP in the same way, but only if they don't focus on crits, flanking or non-physical damage like poison. As such, similar to dagger rogue it may be worthwhile to invest in a separate AP set for dragon-slaying, but the bonuses are fairly marginal compared to what dagger rogues can expect.

 

As for SnS warrior? Damage really shouldn't be a focus; stagger %, or crit % + abilities which proc on crit (i.e. Sunder) may be more useful for on a tank. However, during the crafting stage you should be picking recipes with more slots for utility on weapons and defensive on armor in the first place.

 

2H warriors will get about the same range of damage against heavy armor targets whether they gear for AP% or attack%, which makes AP% essentially a bad stat for them as they'd suffer heavy damage losses against low/no armor targets, while gaining nothing compared to attack% on heavy armor targets.

 

AP is completely useless for mages, if that isn't obvious.

 

Armor Tests

Spoiler

Armor Test 1

Level 13 warrior, base stats, 2H weapon equipped, so no default shield.

Setup 1: base armor 39.

Setup 2: weak armor 84.

Setup 3: strong armor 167.

Setup 4: buffed armor 190.

Target: Level 11 Bear.

Test: getting hit in the front by 10 bite attacks from  Bear. Crits, swipe (weaker) and slam (stronger) attacks are ignored.

 

Results (mean):

Setup 1: 501.

Setup 2: 406.

Setup 3: 214.5.

Setup 4: 162.8.

 

Conclusion:

Glorious and clear result. It appears each point of armor reduces physical melee damage by 2 points.

 

Armor Test 2

A simple test to determine if ranged physical attacks are affected in the same way.

Setup 1: base armor 39.

Setup 2: weak armor 84.

Setup 3: strong armor 167.

Target: Level 13 undead archer.

Testing: getting hit in the front by 10 arrows. Crits are ignored.

 

Results (mean):

Setup 1: 329.7.

Setup 2: 263.1.

Setup 3: 126.6.

 

Conclusion:

Surprisingly, each armor rating only reduced the archer's damage by 1.5 points instead of 2 points. This indicates an innate effectiveness against armor using ranged weapons. It's also possible that the enemy actually has some Armor Pen stat.

 

Armor Test 3

To be sure, I went to test against a different ranged enemy.

Setup 1: base armor 39.

Setup 2: strong armor 167.

Target: Level 11 Hurlock archer.

Testing: getting hit in the front by 10 arrows. Crits are ignored.

 

Results (mean):

Setup 1: 477.5.

Setup 2: 190.2.

 

Conclusion:

This time around, the original theory that 1 armor negates 2 damage holds. The results of tests 2 and 3 indicate that while ranged damage operates the same as melee damage against armor, some enemies may possess armor penetration effects.

 

Armor Test 4

A test to see if armor affects magical damage at all. It was very quickly apparent that it does not.

 

Melee Defense Tests

Spoiler

Melee Defense Test 1

Test to check when melee defense % applies.

Setup 1: 201 armor, 0% melee defense.

Setup 2: 201 armor, 31% melee defense.

Target: level 11 Bear.

Test: getting hit in the front by bear's bite attack 10 times.

 

Results (mean):

Setup 1: 121.2.

Setup 2: 83.3.

 

Conclusion:

A popular theory that's going around is that melee defense reduces the power of an attack before armor is applied. However, that does not appear to be t he case here. In theory, if a 31% reduction was first applied to the bear's unmitigated attack (roughly 523 dmg), then the resulting dmg of 361 would be fully abosrbed by 201 armor, and the bear would only ever inflict 1 dmg per hit.

 

The implications here are pretty broad. We've seen that similiar attributes work similarly, so it's likely that ranged defense, magic defense, various elemental defenses all work in a similar fashion - applying the damage reduction only on the final damage, and not on the initial attack.

 

Another implication here is that damage reduction % is not as useful as Weaken, which reduces an enemy's attack. 30% weaken is significantly more valuable than 30% defense. The former would reduce a Bear's damage down to 1, while the latter will still allow quite significant damages through (~80).

 

Melee Defense Test 2

Due to how controversial the result is, a second test is performed just so that I'm sure.

Setup 1: 201 armor, 0% melee defense.

Setup 2: 201 armor, 31% melee defense.

Target: level 11 Templar Knight.

Test: getting hit in the front by regular fist attacks 10 times.

 

Results (mean):

Setup 1: 43.4.

Setup 2: 30.7.

 

Conclusion:

This test confirms the previous assumption. It seems clear that Melee Defense %, and most likely other Defense %, only apply to the final damage, after armor has reduced the initial damage. 

 

Critical Damage Bonus Tests

Spoiler

Critical Damage Bonus Test 1

Testing how crit damage bonus works.

Setup 1: Axe 138 dmg per hit, base 40% crit dmg bonus.

Setup 2: Axe 138 dmg per hit, 75% crit dmg bonus.

Target: level 11 Templar with 36 armor.

Test: attack enemy for 10 regular attacks on setup 1. Trigger 10 crits on each setup.

 

Results (mean):

Normal: 109.7

Crit 1: 155

Crit 2: 191.2

 

Conclusion:

Crit damage bonus appears to merely increase the damage by the prescribed % compared to normal attacks.

 

Critical Damage Bonus Test 2

Testing how crit damage bonus interacts with AP. This ought determine whether crit damage bonus is applied before or after damage reduction by armor.

Setup 1: Axe 186 dmg per hit, base 40% crit dmg bonus, 0% AP.

Setup 2: Axe 186 dmg per hit, base 40% crit dmg bonus, 21% AP.

Target: level 11 bear with 45 armor.

Test: record 10 crits on each setup.

 

Results (mean):

Setup 1: 233.6.

Setup 2: 238.7.

 

Conclusion:

Judging by the minor difference between the two results, the damage bonus from AP appears to have only applied once. On both regular and crit attacks, the 21% AP merely increased the damage of those attacks by the same value. The implication here is that critical damage bonus is applied to the attack before damage reduction by armor takes place. Regretfully, this also means AP scales poorly with critical hits.

 

Critical Damage Bonus Test 3

Testing how crit damage bonus interacts with Attack %.

Setup 1: Axe 138 dmg per hit, 75% crit dmg bonus, 3% base attack.

Setup 2: Axe 138 dmg per hit, 75% crit dmg bonus, 13% attack.

Target: level 11 outlaw with 36 armor.

 

Results (mean):

Setup 1: 219.2.

Setup 2: 238.5.

 

Conclusion:

As it has been established before, Attack % applies before damage reduction by armor. Further, based on the results it appears Attack % stacks multiplicatively with crit dmg bonus, when a crit is triggered. It seems this is another win in Attack %'s case against AP.

 

Proc Attribute Tests

Spoiler

Heal on Kill Test

Testing the proc of heal on kill.

Setup: Max Health 859, 25% Heal on Kill.

Test: Kill three targets while health is below 50%.

 

Results:

All three kills healed my character for 214, or 25% of max health.

 

Conclusion:

Heal on Kill heals the prescribed % of maximum health on every killing blow.

 

Stagger Test

Testing the duration and effect (for combo) of Stagger on Hit/on being Hit.

Setup: Stagger on hit 17%, stagger on being hit 14%, Shield Bash unlocked to test combo.

Test: try to trigger stagger either via attack or being hit, time the stun, check if it combos.

 

Results:

1. Stun lasts 3 seconds.

2. The stun allows combos.

3. There is no floating text notifying "Stunned".

4. The stun allows combos.

5. There is no internal cooldown. An enemy can get stunned immediately after recovering.

6. Regardless of attack type, as long as an enemy is in melee range the enemy's attack will have a chance of triggering Stagger on being Hit. i.e. bow or staff direct attacks can trigger stagger if the enemy attacks player while player is in melee range.

 

Bleed Test 1

Testing the duration and damage of the Bleed proc.

Setup: Weapon damage 135, Bleed on Hit 31%, 0% AP.

Target 1: level 11 Horror with 45 armor.

Target 2: level 11 Archer with 18 armor.

Target 3: level 7 Ram with 0 armor.

Test: proc bleed on Bear, record the duration and damage of Bleed. Bleed in theory should pass through armor.

 

Results:

Bleed failed to trigger on all three targets. Initially I thought it may not trigger on certain enemy types, but right now it seems the stat is simply broken. I'll test Bleed on being Hit to see if that works.

 

Bleed Test 2

Checking if Bleed on being Hit works.

Setup: 23% Bleed on being Hit.

Test: Will carefully monitor enemy health, just in case Bleed doesn't give any floating combat text like Stagger.

 

Results:

After leaving my character alone for 5 mins while 2 weak enemies hit on him, the enemies remain at full health. The Bleed effect appears to be broken for Bleed on being Hit as well.

 

Mastercraft Tests

Spoiler

2% Chance Shield Bash (Mastercraft)

The mastercraft didn't state how much damage it'd deal, so I figured... I'd put it on a bow.

Setup: 132 DPS bow, two pieces of Shield Bash Mastercraft for 4%, 5% base crit chanace.

Test: Just wanted to see if it procs on ranged attack, and how much damage it'd deal.

 

Result:

I almost didn't think that it'd proc, since I've started getting plenty of crits, but no Shield Bash. Eventually, after many many arrows, suddenly the a Shield Bash effect triggered on the distant enemy for 434 damage, and the enemy was even knocked down.

 

It appears that the Shield Bash mastercraft allows a full-blown 300% weapon damage Shield Bash with all its original effects to trigger on even ranged attacks. Statistically speaking, it's still inferior than 10% chance to deal Chain Lightning/Walking Bomb for 50% weapon damage.

 

It feels as though the mastercrafts simply didn't stack. However, difference between 2% and 4% is somewhat hard to detect, so I shall test this issue using more reliable means.

 

10% Chance Chain Lightning (Mastercraft)

This test aims to see if this type of Mastercraft stack in proc chance or damage.

Setup: level 13 warrior, base stats, weapon damage 134, 3 pieces of gear with Mastercraft "10% chance to cast Chain Lightning for 50% of weapon damage".

Target: level 7 Outlaws with 27 armor.

Test: examine if actual proc chance is closer to 30% or 10%. Examine if actual damage is closer to 50% of weapon damage or 150%.

 

Result:

Proc Chance: 7% (7 procs in 100 hits).

Damage (mean): 73.3 (around 50% of weapon damage without armor mitigation)

 

Conclusion:

It seems the Mastercraft does not stack. Despite having 3 pieces of the same Mastercraft, each claiming 10% chance and 50% damage, it is as though only a single piece of the Mastercraft was worn. A brief and undocumented test where only a piece was worn revealed similar proc chance and damage.

 

I also noticed that the Chain Lightning applied "Shock", had the same 5m chaining range, and was able to crit. It has effectively the same effect as regular Chain Lightning, except it has a lower damage multiplier.

 

10% Chance Walking Bomb (Mastercraft)

A brief test to examine if it works as ingame text. Prior to the explosion, the Walking Bomb dealt 50% damage per second as a DoT. The final explosion also dealt only 50% damage, but the range of the AoE was also 5m and caused knockdown just like a regular, but weaker, Walking Bomb. Of course, duplicates of the Walking Bomb Mastercraft did not stack.

 

+15 Stamina (Mastercraft)

I was initially going to stack stamina and test regen, but a very obvious flaw was quickly revealed. This mastercraft also does not stack! Despite wearing 3 pieces of gear each claiming to give 15 stamina, my character ended up with 115 max stamina instead of 145 max stamina.

 

Regen of +15 Stamina (Mastercraft)

Whether Stamina regen is affected by a higher max stamina may be the deciding factor between higher max stamina, or reduced ability cost.

Setup 1: 100 stamina

Setup 2: 115 stamina

Test: deplete stamina and record the time taken to regen to full.

 

Results:

In both setups, it took 25 seconds to regen from 0% to 100% stamina.

 

Conclusion:

Stamina regen appears to be percentage based. As a result, +15 stamina is superior than 7.5% reduced cost. Of course, there is nothing stopping you from getting them both. Those two mastercrafts would "stack" multiplicatively and give you an even higher effective resource pool.

 

Heal 15% Damage over 10s (Mastercraft)

Also didn't stack. Test will instead focus on how it heals.

Results:

- The heal only triggers if damage is dealt directly to health, not Barrier or Guard.

- Does not stack with itself. Heals only the latest damage. E.g. if hit for 1000 dmg, will start healing 15 dmg/s. After only 3s, if then hit for 200 dmg, will start healing for 3 dmg/s from that point onwards.

- The healing does not generate floating text.

 

+5% Damage for Every Nearby Enemy (Mastercraft)

After a user reported that 7.5% damage for every nearby enemy supposedly stacks, I decided to retest some mastercrafts, including this one, to see if it's just a failure of UI, but not failure of functionality. Both +15 Stamina and +3 Guard/hit still do not stack at all.

Setup 1: Level 19 warrior, weapon damage 91, base 4% attack.

Setup 2: Level 19 warrior, weapon damage 91, base 4% attack, 5% Dmg/Nearby Mastercraft x1.

Setup 3: Level 19 warrior, weapon damage 91, base 4% attack, 5% Dmg/Nearby Mastercraft x3.

Target: Level 11 Bear

Test: test whether the mastercraft stacks against a single target when no other enemy can even be spotted nearby. 10 regular hits with each setup, crits are ignored.

 

Results (expected vs actual mean):

Setup 1: 47.84; 48.8

Setup 2: 52.57; 51.8

Setup 3: 62.04; 52.1

 

Conclusion:

The increase in damage between setup 1 and 2 fall almost perfectly within the expected values. However, Setup 3's result falls far too short of what can be expected from a 15% base damage increase. It seems that at least the 5% Damage for Nearby Enemy mastercraft does not stack.

 

+7.5% Damage for Every Nearby Enemy (Mastercraft)

To remove all doubt, I went and farmed some of FT Bloodstone for the test. I also opted to use a weapon with higher weapon damage, so that the results may be more clear.

Setup 1: Level 19 warrior, weapon damage 147, 43% attack.

Setup 2: Level 19 warrior, weapon damage 147, 43% attack, 7.5% Dmg/Nearby Mastercraft x1.

Setup 3: Level 19 warrior, weapon damage 147, 43% attack, 7.5% Dmg/Nearby Mastercraft x2.

 

Results (expected vs actual mean):

Setup 1: 145.86; 145.6

Setup 2: 161.63; 156.6

Setup 3: 193.16; 155.1

 

Conclusion:

It's clear that the same type of mastercraft would work the same, regardless of tier. While the increase between setup 1 and setup 2 is expected, there is no increase between setup 2 and setup 3.

 

Minor Tests

Spoiler

Enhanced Ability Ring (Duration)
Horn of Valor, tested using no ring, 1 enhanced ability ring, and 2 enhanced ability ring.
Results: no ring 10s; 1 ring 13s; 2 rings 13s.
Conclusion: Duration bonus from Enhanced Ability Rings do not stack.

 

Enhanced Ability Ring (Damage)

Test to see if multiple enhanced ability rings stack with each other when it comes to bonus damage. Ability tested is Longshot, and a pair of +30% dmg to Longshot rings were used.

Setup 1: 132 DPS bow, base stats, no ring.

Setup 2: 132 DPS bow, base stats, 1 ring (+30%).

Setup 3: 132 DPS bow, base stats, 2 rings (+60%).

 

Results (mean):

Setup 1: 944.2

Setup 2: 1260.5

Setup 3: 1257.2

 

Conclusion:

It appears that contrary to the rumor I've heard, enhanced ability rings simply do not stack, period. Setup 2 enjoyed a very straight forward ~30% bonus over Setup 1. Any difference between Setup 2 and 3 appears to be merely the result of RNG.

 

Critical Chance Test

Simple test on how crits are triggered against multiple targets. Results show that when multiple enemies are hit, each hit can trigger crits indepently from each other. I've seen many cases of one attack triggering crit, while others don't, and also cases where multiple hits trigger crit on one AoE attack. The simple implication is that abilities that proc on crits will benefit immensely from AoE attacks.

 

On a side note, DoTs do not seem to ever crit. I've tested Poisoned Wepaon, Walking Bomb (DoT) and Spirit Mark thus far.

[GhoXen]

Change Log

• 4/12/14: Re-tested mastercraft stacking more extensively, namely "+% dmg per nearby enemy". Results unchanged.
• 1/12/14: attack, AP, attack vs AP, damage formula updated. Created gear equivalent list for DPS ranking. Made the thread's format more user-friendly.
• 29/11/14: enhanced ability rings, various mastercrafts updated.
• 28/11/14: armor rating, melee defense %, other defense %, guard, crit %, crit damage bonus %, other damage bonus %, heal on kill %, stagger %, bleed % updated. Added damage formulas for player and enemy.
• 27/11/14: armor penetration %, armor rating, barrier damage bonus % updated. Attack vs AP analysis.
• 26/11/14: primary attributes, attack %, duration % (item) updated.

[Disco Overlord]

Thank you so much for doing this.  A lot of the stat and attribute descriptions are way too vague.  I've been avoiding crafting attack% items because I stupidly thought it was attack rating, guess I need to redo some things!

[SomeoneStoleMyName]

Will keep an eye on this. Gj on taking some initiative.

To avoid confusion I suggest clarifying if the attack % stat affects only auto-attacks, or also abilities and spells. (I think it covers both). 

 

[Sevitan7]

Thank you for your work.

[ironhorse384]

I'll be following this thread

[konfeta]

There's another question:

 

Are the % damage buffs additive, multiplicative, or are there "classes" of damage buffs within which they are additive? For example, there are Fade touched materials which give bonuses like "+20% damage +100% damage taken" and "+10% damage if not hit for the last 5 seconds;" and then there are bonuses like Rift Mage passive for +15% damage to Weakened Targets or that warrior party damage buff that adds 35% to the whole party.

 

The difference between 1.2x * 1.1x * 1.15x * 1.35x and 1.8x isn't trivial.

[TheInvoker]

Cunning gives only 0.5 critical chance i think

[GhoXen]

Thank you so much for doing this.  A lot of the stat and attribute descriptions are way too vague.  I've been avoiding crafting attack% items because I stupidly thought it was attack rating, guess I need to redo some things!

 

Yeah, after finding out more about stats, now I'm aiming primarily for Willpower and Attack % on non-rogues. Guard/Barrier damage bonuses are extremely unworthwhile stats, since warriors already have more than enough tools to destroy guard easily, while barriers in general melts very quickly.

 

There's another question:

 

Are the % damage buffs additive, multiplicative, or are there "classes" of damage buffs within which they are additive? For example, there are Fade touched materials which give bonuses like "+20% damage +100% damage taken" and "+10% damage if not hit for the last 5 seconds;" and then there are bonuses like Rift Mage passive for +15% damage to Weakened Targets or that warrior party damage buff that adds 35% to the whole party.

 

The difference between 1.2x * 1.1x * 1.15x * 1.35x and 1.8x isn't trivial.

 

Good point. So far I only know that +25% guard is additive. I'll put it on my to-do list as part of clarifying the stat mechanics in general. I just hope that it's consistent across the board, and not sometimes additive, sometimes multiplicative.

[cloudblade70]

Also, regarding critical hit damage; is the base critical hit damage double a normal attack (100%) and then the bonus critical hit damage is added onto this? For example, if the displayed bonus critical hit damage on your attributes screen is 50%, your base attack is 100, and you land a critical hit, will you do 250 damage?

[Anelyn77]

While the dmg vs barrier / guard isn't stellar, it's still good to have. 

 

You can't instantly remove guard from multiple opponents nor dispel all barriers at once (enemies spread out etc), it's just not a major stat.

[GhoXen]

While the dmg vs barrier / guard isn't stellar, it's still good to have. 

 

You can't instantly remove guard from multiple opponents nor dispel all barriers at once (enemies spread out etc), it's just not a major stat.

 

It's just not as useful as magical defense. Magic hurts like hell on Nightmare, and it's very difficult to avoid it completely.

[Magma_Axis]

Cunning is only 0,5 % per point in my experience

 

If i craft +6 cunning bow, i get +3 % crit chance

[GhoXen]

Indeed. It's one of the first things I discovered. It has been corrected.

[Anelyn77]

Anything hurts until you start developing gear / skills

 

Enemy mages don't insta cast powerful spells, you have a window of time to interrupt / kill them, precast barrier / build up guard, spread around etc. Not saying MD is bad, by all means, but I wouldn't go for a piece of gear with 5 willpower vs one with 9 magic (rest of stats being identical).

[brazen_nl]

BM'd. Fantastic, GhoXen!

[WillieStyle]

Excellent post. Thanks for the good work. I'm curious about how exactly armor works, and whether armor penetration is a good stat.

[SomeoneStoleMyName]

I find it kinda tragic and humorous both that stacking willpower is superior to stacking strength for warriors.

Seems that the wanted stats atm are:

(Hard and Nightmare taken into account, where resistance is superior to guard damage)

Mage: Willpower > Cunning=Magic 
Warrior DPS: Willpower > Cunning > Con=Dex
Warrior Tank: Willpower=Cunning > Con
Rogue: Dex=Cunning > Willpower

Feel free to disagree on this assessment.

 

Note: One crafted armor with 100 health gives the equalent of 20 Constitution (Health only)

[Jais9]

Okay, why is the Attack stat so borderline useless? As an example, assume we're playing a warrior and pick around 10 passives that each give +3 Strength. That's +30 in total, which translates into a meagre +15% damage. That's nothing. Why wouldn't I rather stack Willpower (to at least also get some magic defense) or Constitution (for HP and melee defense) instead?

[Magma_Axis]

Why cunning = Magic for mage ?

 

Beacuse of Flashpoint ?

[Hellnhavoc]

Has anyone noticed anything with will power and stamina/mana regen?  There has to be a way to regen faster beyond those few skills.

[WillieStyle]

I'm not sure what's supposed to be the problem with the attack stat. 15% additional damage seems like quite a lot to me.

[WillieStyle]

Has anyone noticed anything with will power and stamina/mana regen? There has to be a way to regen faster beyond those few skills.

Warriors and Rogues also build stamina via auto-attacks. The game seems to be designed with the expectation that a nontrivial portion of Warrior and Rogue attacks will be auto-attacks. Which is one reason why warrior and rogue weapons have much higher dps than staves.

[SomeoneStoleMyName]

Is magic defense stat basically "resist all"?

 

If a 200 damage fireballs hits your face... will there be a different outcome from having 35% fire resist (with no magic defense) or 35% magic defense (with no fire resist)?

Do they add additive? 17% fire resist + 11% magic defense = flat 28% damage mitigation vs fire?