Author Message

trainer772

Rank 5
trainer772
Joined
15 Feb 2012
Posts
3316
Location
Bowlofrice Korea, South
PostedOct 17, 2012 4:29 pm   Last edited by trainer772 on Feb 26, 2013 8:51 pm. Edited 8 times in total

The complete damage mechanics guide

I was planning on releasing this project earlier but I was really busy with school and was unable to do necessary testing to verify data. I decided to release this anyways and see how the feedback goes.

Credits to the various people that suffered multiple deaths for the sake of my curiosity.

1Byte
822mL
daisy77
Dinodude
dNocturnal
eTx
gl0ckS2
herrokitty
iDeadlysushi
Kirsi
koguy
Raaa
Soogie
Whatacreatur
xZiIe
Ziggy


Basic rules to set down before I begin:

1. All damage stats lie. No exceptions. Refer to number 3. The only reason Redduck provided these numbers was for users to get an understanding of roughly how much damage it deals relative to other guns.
2. Body parts have flat multipliers. There are no gradients (For example, hitting the chin of the head deals no less damage on average than hitting the eye).
3. All guns have a center value and a variance value. It determines the range of numbers it will pick the damage from.
4. All guns have a range value. It determines how much damage is lost over distance.
5. All decimal numbers are rounded down. This is one of the fundamentals of the entire game engine. Although I am unable to verify that it applies to damage, I am pretty sure it does, considering all other values (repair costs, supply earnings, AP item bonuses, binocular range, etc.) follow this principle.
6. All bullets will eventually deal 0 damage, given enough range.
7. Headshots deal 400% damage. This number has been verified by Redduck itself and is the basis for the entire mechanics guide.
Advertisement

trainer772

Rank 5
trainer772
Joined
15 Feb 2012
Posts
3316
Location
Bowlofrice Korea, South
PostedOct 17, 2012 4:30 pm   Last edited by trainer772 on Mar 09, 2014 8:52 pm. Edited 11 times in total
Damage basics

Disclaimer: all usage of statistical words such as center and variance do not carry their actual statistics-related definitions.

How exactly is damage calculated? The primary factors that determine damage dealt (excluding penetration-related factors) are:
-Center damage (the middle of the distribution curve)
-Damage variance (the range of values and the frequencies of each value)
-Distance
-Body part multiplier

Penetration-related factors to be discussed later include:
-Armor worn
-Gun's penetration rate and caliber
-Map surfaces penetrated & distance penetrated

=====

When a bullet hits a damageable surface (glass pane, box, explosive barrel, or person), the engine picks a number from the possible sets given the center and variance assuming that all numbers within the set have an equal chance. However, it is known that numbers in each given set are weighted, and some are likelier to appear.

D = c + randInt(-v/2, v/2)

So if SG550 had 40 damage and 4 variance, it would deal damage between 38 and 42. If AKMK3 had 39 damage and 10 variance, it would hit between 34 and 44. Most rifles and submachineguns have low variances, while sniper rifles tend to have higher variances.

A strange phenomenon should be noted: headshot damage is extremely consistent variance-wise, and most bullets that hit the head fall within a tight range if everything is held constant. Headshot damage tends to incur a lower number in the damage set. For example, if the damage set at a certain range for a certain gun was 23-27, it would most likely deal 23, 24, or 25 damage before being affected by the 4x headshot multiplier.

=====

Simple enough, right? Then we have to include distance. Every gun has a range coefficient, or an unknown value that satisfies its equation (damage)(range)^(meters). Refer to this post for some sample range analysis.

D = [c + randInt(-v/2, v/2)](rangecoefficient)^(meters)

=====

Then we include the body part multiplier. The only known value is headshot = 4x. I have estimated the others as follows:
Stomach: 1.5x
Chest: 1.25x
Arms: 1x
Legs: .7x
Refer to: http://www.aeriagames.com/forums/en/viewtopic.php?t=1867243

So now we have:

D = (multiplier)[c + randInt(-v/2, v/2)](rangecoefficient)^(meters)

For later calculations, I will use 0 meters to hold the distance factor at 1 unless otherwise tested.

=====

Disclaimer: example numbers are hypothetical.

So basically, if an SG550 (which is known to have low variance and approximately 40 damage) hits a player in the guts at 0 meters, we may run into the following set of damage

D = 1.5x41 = 61.5

So 61 damage would be dealt, since all numbers are rounded down.

Example: Glock (with a known 35.5 center damage and 2.5 variance) at 0 meters

D = 1.5x36 = 54 damage, enough to cut down on half health

=====

Several exceptions to damage stats:
While most stat damages listed are close to the actual center of the gun, there are several key exceptions listed as follows. Disclaimer: all values were tested by me and may be inaccurate.
Pistols: http://www.aeriagames.com/forums/en/viewtopic.php?t=1880451
Snipers: http://www.aeriagames.com/forums/en/viewtopic.php?t=1898235

trainer772

Rank 5
trainer772
Joined
15 Feb 2012
Posts
3316
Location
Bowlofrice Korea, South
PostedOct 17, 2012 4:30 pm   Last edited by trainer772 on Feb 26, 2013 8:40 pm. Edited 5 times in total
--

trainer772

Rank 5
trainer772
Joined
15 Feb 2012
Posts
3316
Location
Bowlofrice Korea, South
PostedOct 17, 2012 4:30 pm   Last edited by trainer772 on Feb 26, 2013 8:47 pm. Edited 9 times in total
Shotgun/Sniper damage

Sniper rifles: Why the OHK??

This aspect of AVA has always eluded players for years. When making this post, I will start with the following irrefutable assumptions to back my theories:
Assumption one: the leg multiplier is 70%
Assumption two: Sniper rifles can OHK to the leg at 0 meters. Notable examples include TPG (with "110" damage), PGM (with "114" damage), Mosin (with "99" damage), or FRF2 (with "95" damage).

Anyone who puts these two facts together can only come to realize that sniper rifle damages are incredible lies. FRF2 with its 95 damage simply cannot multiply by 1x on the arm and effectively OHK.

Bolt-action sniper rifles have center damages anywhere from 110 to 130 and huge variances of up to 20 points.. While I do not have exact numbers due to the immense difficulty associated with testing numbers greater than 100, this thread shows some approximate values.

It is important to note that damage =/= OHK. For example, the DSR + RB has a higher central damage than the TPG. However, its range (since all reinforced barrels come with a severe range penalty) that the OHK quickly drops off past 20 meters.

Bolt action rifles tend to have high center damages and high variances, resulting in high OHKs with occasional tags. Semi-autos tend to have lower center damages but lower variances, resulting in consistent OHK limitations. The damage stat given for semi-autos more accurately reflect their true damage than those of bolt actions.
=====

Here is an approximate OHK ranking, with center damage, variance, range, multipliers, and penetration considered. All sniper rifles on this list are assumed to be modded for maximum OHK. They are clustered into groups of similar OHK rates. Keep in mind that the actual user's OHK rates may vary depending on the distance that the user prefers to engage the enemy at, the accuracy with which he hits the enemy's weakest spots (their stomach), and the armor intensiveness of the game mode being played.

95-99% OHK
ASW & TPG
PGM
TPG EPKO

90-95% OHK
DSR (Extremely high OHK rate even to arms when using LRBII. OHK drops to 60-80 with RB)
Mosin (tags frequently on armor.)
FRF2 (questionable OHKs at longer ranges to the limb, even with LRB)
Silver Arrow (sometimes tags to limbs)

80-90% OHKs
AWM (tags to limbs)
Blaser (frequently tags to limbs)

60-80% OHKs
SV98 (almost always tags at longer ranges except to guts)
Kar 98 (Similar OHK to SV98)
Zebra (frequently tags to limbs)
WA2K (tags to limbs)
SR25 (rarely OHKs limbs)
Borte (never OHKs limbs)

40-60% OHKs
VSS Katana (requires 2 hits when not hitting guts in CQC)
M24 (ocasionally OHKs on guts)
Dragunov (ocasionally OHKs on guts in CQC)
MSG90 (ocasionally OHKs on guts in CQC)


0-40% OHKs
Mad Galil (almost always requires 2-3 hits)
HK417 (almost always requires 2-3 hits)
IMI99 (almost always requires 2-4 hits)
Galil (almost always requires 2-5 hits)
VSS (almost always requires 3-5 hits)

The Shotgun paradox: OHKing with 70-80 damage?

I eventually came to conclude that though sniper rifles deal ~120 damage per shot, one thing is clear: shotguns deal much more damage. Of course, this is assuming that you're at 0 meters and all pellets hit. Most shotguns have center damages of 250-350 per 8 pellets and have surprisingly stable variances. Guns like Blue Skull and FN TPS have damages up to the 350 range, whereas weaker shotguns like Striker and Spas 15 unmodded have lower damages of ~250. Most of the random OHKs can be attributed to missing some pellets, which could mean as much as 45 less damage dealt at 0 meters. Shotguns like Blue Skull and Saiga have consistent damage due to their tight pellet groupings more so than their excellent damage and range combination, whereas TPS and BG Crusher (T3) tend to OHk less often due to their big crosshairs.

Here is a ranking of OHK rates with range, penetration, damage, variance, and accuracy considered.

Blue Skull

TPS
Saiga

Spas 15 (T3)
Benelli

Remmington
Striker 12
Spas 15

Spas 15 (T1)
Spas 12

trainer772

Rank 5
trainer772
Joined
15 Feb 2012
Posts
3316
Location
Bowlofrice Korea, South
PostedOct 17, 2012 4:30 pm   Last edited by trainer772 on Feb 26, 2013 8:50 pm. Edited 2 times in total
Health, Armor, and Penetration

Refer to this post for armor values.


A 34-damage pistol shot on an EA08 at 0 meters to the guts would deal (34-20)x1.5= 21, which is approximately 1/5 of his health. The next shot would hit the health right away since there is no armor, resulting in 34x1.5=51 damage.


Pointmen and snipers have 100 HP. Riflemen have 105 HP. Pointmen with ghostgear supply kits have negligible health boosts. However, all armor have the same absorption abilities on all classes with the exception of a lack of helmet on sniper class and increased rifle helmet defense.

It should be noted that high-penetration submachine guns, assault/battle rifles, and sniper rifles have difficulty penetrating LCA/EA08, making it the best armor to absorb high-penetration guns like the MP7, Veresk, Para, and various sniper rifles. SCA/AA07 can absorb shotgun pellets and low-penetration submachine gun bullets effectively, making them unpopular to wear. SA09 is slightly weak against high penetration guns like Para, Veresk, and MP7; however, it has much greater durability and can survive additional blows.

AH07 and smallcal helmets tend to stay on the head longer but are worse at absorbing blows. However, LCH and EH08s can absorb incredible amounts of damage but tend to fall off easily, especially from grenade blasts. SH09s can stay on for multiple hits and absorb many rifle bullets.

Refer to this post for a grenade-armor interaction analysis.

trainer772

Rank 5
trainer772
Joined
15 Feb 2012
Posts
3316
Location
Bowlofrice Korea, South
PostedOct 17, 2012 7:03 pm
I think I'll grab another post.

irobbing

Rank 5
irobbing
Joined
02 Jul 2012
Posts
4235
Location
Maryland, United States
PostedOct 17, 2012 7:18 pm
Thanks for the information Trainer772, I really appreciate your time and effort into making this guide for AVA. Hope you can continue your work.

faintshadow6

Rank 0
Joined
25 Sep 2012
Posts
16
Location
United States
PostedOct 20, 2012 9:00 am
thanks a lot for the guide, this explains my issues with the striker so much. I just never understood why i could snipe a guy from across and right after tag at point blank with the gun, even though i dont use the rando barrel. On a side note, should I use the random damage barrel or not since the gun has so much variance?

trainer772

Rank 5
trainer772
Joined
15 Feb 2012
Posts
3316
Location
Bowlofrice Korea, South
PostedOct 20, 2012 3:33 pm
faintshadow6 wrote:
thanks a lot for the guide, this explains my issues with the striker so much. I just never understood why i could snipe a guy from across and right after tag at point blank with the gun, even though i dont use the rando barrel. On a side note, should I use the random damage barrel or not since the gun has so much variance?  
I recommend it for the awesome 10-meter shots lol. Just zoom in and always be ready to fire off another shot.

XiaoJyun

Rank 5
XiaoJyun
Joined
25 Jul 2012
Posts
3573
Location
Slovenia
PostedOct 23, 2012 5:05 am
TL DR game is pay2win due to armours, guns have huge variance to help noobs, so a baddie can OHK with his SVD Zebra, while tanking PGM/ASW/FRF2

i guess that are the main reasons for silly stuff like variance

to add up, overpriced ghostgear that only benefits PM, already weakest class isnt free, how it pretty much should be since PM needs defense (after all hes supposedly CQC class)

nonetheless interesting info, but not really helpfull this time
Display posts from previous:   Sort by: