Aller au contenu

Photo

Weapon damage formula (WIP)


  • Veuillez vous connecter pour répondre
6 réponses à ce sujet

#1
peddroelm

peddroelm
  • Members
  • 2 396 messages
 
Reverse engineered based on ingame combat log tests. Still work in progress, will update if/when I do more tests.   
 
 
 
AmmoPowerDamage*ScalingFactor + WeaponDamage*ScalingFactor [+ IncendiaryAmmoDOTvsHealth*ScalingFactor]
 
AmmoPowerDamage = ModifiedBaseWeaponDamage*Ammo%*(1+Sum(PowerDamageBonuses  - RangeBonusPenalty))*VsDefenceMod
 
WeaponDamage = ModifiedBaseWeaponDamage*(1+Sum(WeaponDamageBonuses))* ModifiedWpnVsDefenceMod
 
 
ModifiedBaseWeaponDamage: For the most part this is BaseWeaponDamage. Damage upgrades for Assault rifles apply to BaseWeaponDamage unlike upgrades for the other weapon classes which are additive. BUG ? WAD ?? SquadMates also have this multiplied by weapon_Squad parameter (specific for every weapon, between 0.4 to 0.65 )
 
PowerDamageBonuses: from passive, gear, squad power bonus ,etc ..
 
 
VsDefenceMod: damage modifier vs non-synthetic health, synthetic health, armor, shields, barriers 
 
disruptor     0,1,0,1,0 
warp          1,1,1,0,1  (warp ammo does *2 damage vs health lifted by biotics )
incendiary 1,0.25,1,0,0
AP            1,1,1,0,0
 
WeaponDamageBonuses: Passives, Gear, Adrenaline rush (100% or 140%), Cloak, WeaponResearch(non assault rifles), headshot (50% vs YMIR ; 75% vs humanoids, geth Prime), Miranda Squad(up to 15%), vsfrozen(100%), PistolDoubleDamage(100%), sniper rifle extra HS damage research (50%), Squad_Ordered_to Attack (50%), vsLifted (100%), vsFallenfromNeuralShock(100%), RangeBonusPenalty,  etc
 
RangeBonusPenalty: Weapons have a optimal range interval to target. (details + graph + videos)
-when distance to target is in the optimal interval - no bonus or penalty due to range is granted. Testing in this "optimal" weapon range is consistent/reversible for all weapon classes. 
-when distance to target is closer that optimal: a linear additive damage is granted to weapon damage (NOT AMMO) for all weapon classes  + Collector Particle Beam HW , except Sniper Rifles (the bigger the enemy's 3d model the lower this damage bonus can get to  (ex: ~30% vs Colossus, ~70% vs YMIR ; ~90% vs humanoids))
-when distance to target is outside optimal range, ALL weapon classes (including Sniper Rifles & Collector HW) get an additive penalty to damage that also affects ammo damage ( at Kal'Reegar vs Colossus range (Mantis scope measures that distance between 62 to 66 meters as the colossus dances on his spot  ) - ~50% Mantis , - ~82% Katana, - ~50% Collector Beam HW ) 
   
 
ModifiedWpnVsDefenceMod: weapons have different multipliers vs the defense types. Penetration type of research improves this. (very powerful upgrade - somewhat similar effect to the headshot multiplier in ME3 vs the "right" defense type) 
few examples: (vsarmor,vsshield,vsbarrier)
 
Tempest 1, 1.5, 1.5  becomes 1, 1.5*1.5, 1.5*1.5 
Vindicator 1.25, 1.25, 1.25 becomes 1.25 * 1.25, 1.25 * 1.25, 1.25 * 1.25 
Mantis 1.5, 1, 1  becomes 1.5 * 1.5, 1, 1
Viper 1.35, 1.15, 1.15 becomes 1.35 * 1.5, 1.15, 1.15 
*BUG alert shotgun penetration upgrade doesn't work vs barriers. 
 
DamageScalingFactor: enemy hitpoints (Health, Shield, Armor, Barrier) don't scale based on difficulty setting and SHEP level. The damage you do them is scaled based on difficulty setting and SHEP level.  (http://forum.bioware.../#entry20154463)
  
 
-------------------------
Quick look on power damage. Evolving powers in ME2 changes their base damage (unlike ME3), which makes stacking power bonuses stronger (bigger effect).
 
 
Theory-crafting Builds is severely hampered by the lack of enemy hit points value and precise formula for DamageScalingFactor :(
---------------------------
 
 
new info: 
 
-assault rifle damage upgrades apply to base damage, the other weapon classes get additive bonuses (major factor in making Mattock/RevenantVindi/ and avenger deadly in full auto in ME2 )
-followers get 50% additive damage bonus for a few seconds when ordered to attack a target. (similar to the 200% they get in ME3) 
-sniper rifles DO get damage penalty outside optimal range
-collector particle beam heavy weapon is affected by additive damage bonuses (passives, gear, Miranda squad bonus, adrenaline rush, rangebonuspenalty, headshot .. )
 
Link to 20.03.2016 test numbers https://docs.google....dit?usp=sharing
Link to 21.03.2016 test numbers https://docs.google....dit?usp=sharing

  • Tonymac, capn233, RedCaesar97 et 1 autre aiment ceci

#2
capn233

capn233
  • Members
  • 17 238 messages

Good stuff.

 

The defense multiplier table seems like the fields might be different than the labels.  For instance. 0.25x multiplier for IA in health field, I imagine that should be in the synthetic health field unless the game is completely broken, but that doesn't seem like it would jive with experience.  I also imagine the second armor field is synthetic armor, which would make sense given that DA listed as 1x multiplier to it.

 

Is there any distinction between "non-synthetic" and "organic" for this?  There are some oddities with the way Reave works when you consider how it CC's Geth, but doesn't give player the health bonus; and it will not CC husks, but will give the player the health bonus.

 

Info about SR's is interesting.  That would probably explain some headshot "misses" a few people have had over the years.

 

As far as the new info about base weapon damage modifiers, it would seem that the implication is mainly that research improves AR ammo damage but not the damage for any other weapon,  It is unclear to me that this was intended, but I am trying to go off recollection of dev comments specifically in regards to ammo damage scaling as weapon research is carried out.  Oh and a bigger interaction with a little old weapon damage bonus called Adrenaline Rush.  For classes other than Soldier, I don't think that the difference should matter much, since there are largely a dearth of weapon damage bonuses for them outside of research.



#3
peddroelm

peddroelm
  • Members
  • 2 396 messages

The defense multiplier table seems like the fields might be different than the labels.  For instance. 0.25x multiplier for IA in health field, I imagine that should be in the synthetic health field unless the game is completely broken, but that doesn't seem like it would jive with experience.  I also imagine the second armor field is synthetic armor, which would make sense given that DA listed as 1x multiplier to it.

 

 

 

corrected . 

 

updated OP based of a few more test results
 
 
-pistol "double damage" upgrade is 100% additive weapon damage bonus ~30% chance to trigger (will need way bigger data sample to get more confidence on that % chance to trigger). Works for squadmates (tested predator Mordin) as well. 
-disruptor ammo vs YMIR armor did 0 damage. Its description is thus misleading (should read "shields & synthetic [health]"). Have yet to test is vs organic health.
-incendiary ammo & incinerate apply their listed damage to armor instantly (*2.4 for incinerate as per DEV claim) similar to AP and warp ammo. No DOT vs armor. THERE IS NO DAMAGE LOSS VS SYNTHETIC ARMOR.
-Mordin's 20% omnitool research might give him 25% tech power bonus (more tests needed - got an unidentified 5% power bonus in his incinerate tests vs armor)
-distuptor ammo will do its listed damage vs synthetic health
-incinerate & incendiary ammo damage vs [tested synthetic] health is applied as DOT - applied in way to small time intervals (very CPU intensive way to do it with little apparent benefit - ME3 0.5 second DOT ticks - save tons of CPU cycles and easier to read in a combat log :)
-frozen targets get 100% additive weapon damage (tested with Mordin's cryo blast). Too bad you need to remove protections before you can freeze them. Still health is the hardest "defense" since no weapon gets any bonus against it - cryo ammo is effectively +100% (additive) weapon damage vs health. Too bad frozen targets will sometimes fall out of LOS, behind cover.
-frozen target take a small amount of fall damage when doping from their feet to the floor
 

  • RedCaesar97 et KrrKs aiment ceci

#4
KrrKs

KrrKs
  • Members
  • 860 messages
-disruptor ammo vs YMIR armor did 0 damage. Its description is thus misleading (should read "shields & synthetic [health]"). Have yet to test is vs organic health.
-incendiary ammo & incinerate apply their listed damage to armor instantly (*2.4 for incinerate as per DEV claim) similar to AP and warp ammo. No DOT vs armor. THERE IS NO DAMAGE LOSS VS SYNTHETIC ARMOR.

 

Very nice to get these two confirmed.

Thanks for testing all that!



#5
RedCaesar97

RedCaesar97
  • Members
  • 3 829 messages

Interesting that the Collector Particle Beam heavy weapon benefits from weapon damage bonuses.

 

Also interesting that the damage upgrades for assault rifles are calculated differently than the other weapons. Although saying that the Avenger is deadly on full-auto is still clearly a joke.



#6
peddroelm

peddroelm
  • Members
  • 2 396 messages

bits an pieces  - will update them in the OP after I do more tests

 

- Soldier's commando provides a stealth 15% melee damage bonus (which isn't much in the grand scheme of things but it did make me waste a lot of time since I'm counting on KNOWN melee damage to quickly find out the damage scaling for a particular area POWER DAMAGE APPLIES TO MELEE DAMAGE) . And  a stealth 25% weapon damage bonus for headshots. New theory: Humanoid enemies have 75% base HS bonus . (And the 15% power bonus DID not affect disruptor ammo damage - on a different test and it worked for disruptor ammo & concussie grrrr).

 

- Dantius towers first floor vanguard has 270+ health and 450 barriers.  Same place Eclipse trooper has 130+ shields. The damage scaling factor for L13 Shep was 

5713.6 / (125 * 1.15) =39.74678 %

 

Will need to take a look at the melee damage formula at some point. ( Adrenaline Rush bonus doesn't work for melee. The cloak probably does..)

 

 

Ymir soldier test (heavy muscle weave 25% MD, 25% MD strength boost pads, 15% PD commando, 3% PD N7 chest)

Melee damage
Timediff:   196.970 YMIR Shield Damage : 230.469  
Timediff:     0.898 YMIR Shield Damage : 230.469  WTF ?!?!?! pretty high number
125 * 1.25 * (1 + 0.15 + 0.325) = 230.46875 // close but ??!!?
125 * 1.25 * (1 + 0.25 + (0.15 * 1.5)) = 230.46875 ?!?

125 * 1.25 * 1.25 * (1 + 0.03 + 0.15) = 230.46875 // BINGO  power damage bonuses affect melee damage lol (15% PD commando and 3% PD gear )


  • RedCaesar97 et KrrKs aiment ceci

#7
capn233

capn233
  • Members
  • 17 238 messages

- Soldier's commando provides a stealth 15% melee damage bonus (which isn't much in the grand scheme of things but it did make me waste a lot of time since I'm counting on KNOWN melee damage to quickly find out the damage scaling for a particular area POWER DAMAGE APPLIES TO MELEE DAMAGE) . And  a stealth 25% weapon damage bonus for headshots. New theory: Humanoid enemies have 75% base HS bonus . (And the 15% power bonus DID not affect disruptor ammo damage - on a different test and it worked for disruptor ammo & concussie grrrr).

 

This isn't an exhaustive test, but while I was going around looking at HP values and trying to get a few damage multipliers from various parts of the game, I shot an Asari Vanguard (health) using IA.

 

Level 30 Soldier - Commando - Heavy Pistol 5 (also Critical and AP), No Armor Bonus - Inferno Ammo

 

The short of it was... 180 dot ticks (similar to Incinerate), but the DOT amount seems low unless both range bonus and Commando power damage are not applied.

 

Summary

Carnifex impact: 49.067 ->

 

85.4*(1+0.15+0.5+Range)*0.303798 = 49.067

 

Range = 0.24124

 

IA DOT total: 15.592

 

85.4*0.6*0.303798 = 15.567

 

Which is somewhat unsatisfying, partly because it isn't quite the same as the sum of the DOTs, but then again I never got Incinerate and Reave DOTs to be right on the money compared to calculation.  But if I add in the range modifier or power damage modifier, the expected is a good deal higher than observed.

 

Mentioning range modifier just because in the dev mechanics thread it was stated that ammo gets range modifier, but it didn't look like it here.


  • KrrKs aime ceci