+- SWTOR Mechanics Forums (http://mmo-mechanics.com/swtor/forums)
+-- Forum: General (/forum-1.html)
+--- Forum: Game Mechanics (/forum-5.html)
+--- Thread: Damage and Healing formula (/thread-285.html)
RE: Damage and Healing formula - Ezmode - 09-30-2011 03:01 PM
Bonus damage does not include Force Damage. Force damage on the other hand, does increase from the Power rating that contributes to Bonus damage, so that should be fine.
RE: Damage and Healing formula - Kaedis - 10-11-2011 02:15 PM
After an exhaustive session messing with the data, I have a formula that works for at least the data points MK pulled for ranks 3 and 4 Viscous Slash using a Lightsaber. Note that these tests were performed at level 14, tested single-weapon with both Lightsaber and a training sword (more on that later), with a couple dozen data points for each rank with differing weapon damages and bonus damages.
Tooltip = (WpnDmg + StdDmg*0.115)*WpnCoeff + Bonus*BnsCoeff
For rank 3, WpnCoeff = 1.06467, and for rank 4, 1.19669. BnsCoeff = 1.54 for both ranks.
Given the formula itself, I doubt this is perfect, but it works for all of the data points MK dug up.
Here's the spreadsheet we've been using. Of interest is the VS tab, which has the data points I tested. Columns D and F are secondary calculations, they don't apply to my formula.
Another point of interest is that physical weapons (which deal kinetic damage) apparently work slightly differently, as my formula under-estimates their tooltip damage by about 1-2 damage.
RE: Damage and Healing formula - Freehugs - 10-15-2011 01:06 PM
Bit of an update: It looks like the damge and healing of abilites with ranks is worked out from the level the ability becomes avaliable, not your characters level. Which explains why we have ranks of abilities. You can see the rank levels under tbl / pkgabilitiestable
RE: Damage and Healing formula - Freehugs - 10-20-2011 04:17 PM
Looking at the new numbers of talents, I'm thinking that the damage and healing points are no longer based on the characters level. It's possible that damage and healing abilities from talents have ranks now and so the numbers for abilities higher up the tree are actually inflated by their high rank. This would explain for example why Revivification in the Sorc healing tree appeared to get a huge boost for no apparent reason.
RE: Damage and Healing formula - Freehugs - 10-24-2011 04:20 PM
Is anyone working on this stuff other than me? After messing around with the XML data for a while I decided to take a different approach. I looked at the way Knight abilities scale with their melee/force damage bonus. I was able to use between 3 and 14 data points for each ability.
The results:
All abilities had linear scaling within a margin of error +/- 0.5 tooltip damage.
Only 8 of 19 abilities had a coefficient range that matched the coefficient listed in the XML data.
There were 5 abilities that are not listed in the XML data at all.
Hilt Strike - XML Coefficient 2.520 , Actual Coefficient between 1.634 and 1.671 . Matches XML? No
Force Exhaustion - XML Coefficient 1.200 , Actual Coefficient between 1.189 and 1.263 . Matches XML? Yes
Force Exhaustion DoT - XML Coefficient 0.452 , Actual Coefficient between 0.446 and 0.481 . Matches XML? Yes
Zealous Leap - XML Coefficient 1.840 , Actual Coefficient between 1.839 and 2.020 . Matches XML? Yes
Plasma Brand - XML Coefficient 1.160 , Actual Coefficient between 1.154 and 1.226 . Matches XML? Yes
Plasma Brand DoT (4 ticks) - XML Coefficient 0.365 , Actual Coefficient between 0.743 and 0.761 . Matches XML? No
Burning Purpose (6 ticks) - XML Coefficient 0.085 , Actual Coefficient between 0.087 and 0.096 . Matches XML? No
Overhead Slash - XML Coefficient 2.030 , Actual Coefficient between 2.248 and 2.452 . Matches XML? No
Burning Blade (6 ticks) - XML Coefficient 0.057 , Actual Coefficient between 0.050 and 0.062 . Matches XML? Yes
Force Leap - XML Coefficient 0.910 , Actual Coefficient between 0.866 and 1.022 . Matches XML? Yes
Force Sweep - XML Coefficient 1.350 , Actual Coefficient between 1.411 and 1.443 . Matches XML? No
Slash - XML Coefficient 1.540 , Actual Coefficient between 1.533 and 1.732 . Matches XML? Yes
Blade Storm - XML Coefficient 1.870 , Actual Coefficient between 1.969 and 2.041 . Matches XML? No
Riposte - XML Coefficient 1.010 , Actual Coefficient between 1.124 and 1.226 . Matches XML? No
Master Strike - XML Coefficient 1.300 , Actual Coefficient between 4.040 and 4.190 . Matches XML? No
Force Stasis - XML Coefficient 0.680 , Actual Coefficient between 0.669 and 0.722 . Matches XML? Yes
Cyclone Slash - XML Coefficient 1.76 , Actual Coefficient between 1.737 and 1.876 . Matches XML? Yes
Saber Throw - XML Coefficient 1.520 , Actual Coefficient between 1.533 and 1.635 . Matches XML? No
Force Push - XML Coefficient 1.410 , Actual Coefficient between 1.443 and 1.486 . Matches XML? No
Pommel Strike - XML Coefficient - , Actual Coefficient between 3.120 and 3.194 . Matches XML? -
Opportune Strike - XML Coefficient - , Actual Coefficient between 2.646 and 2.712 . Matches XML? -
Dispatch - XML Coefficient - , Actual Coefficient between 2.886 and 2.963 . Matches XML? -
Guardian Slash - XML Coefficient - , Actual Coefficient between 2.165 and 2.350 . Matches XML? -
Guardian Slash alt - XML Coefficient - , Actual Coefficient between 2.759 and 2.886 . Matches XML? -
Note: Master Strike and Cyclone Slash are actually double attacks that only show one damage number on the tooltip. In this case I have just combined their two coefficients.
RE: Damage and Healing formula - Masterkiller - 10-25-2011 03:04 AM
(10-24-2011 04:20 PM)Freehugs Wrote: Is anyone working on this stuff other than me?
I have been living this formula night and day, so yes I definitely am. Kore I dont think will respond to me on IM any more! Unfortunately I've been in Miami for the last week and won't get back to work on it until Wednesday.
Here is my current theory:
[(X * rank modifier) + Weapon + (Bonus * Coef)] * (1+amountmodifier)
Having (1+amountmodifier) as the final multiplier is how Assault is around 45-50% of Weapon+Bonus damage. Prior to (1+amountmodifier) there appears to be 3 terms. If you only use 2 terms, being weapon damage with multipler and bonus damage with multiplier, they scale much to quickly when you increase or decrease weapon damage or bonus damage. Meaning if you have 44 bonus damage and you have a static multiplier of 1.5 or whatever, the result will deviate quickly when you start adding or subtracting bonus damage. Same with weapon damage.
Kore posted a formula that had the same basic principle, although I feel pretty strongly that standarddamageperlevel is not a part of the melee damage (I could be wrong I just haven't seen how it would be possible). Using the formula above, I am within 1 point of tooltip damage for rank 3 Vicious Slash with increasing and decreasing bonus damage and weapon damage. It works. When I go to rank 4, I adjust the X amount to fit the formula and it once again works. I don't think rank touches the bonus damage or weapon damage multiplier .
One reason I believe weapon damage isn't modified, is when switching out weapons the final tooltip amount changes almost directly to the amount of change on the weapon. (1+amountmodifier) for Vicious Slash is 1.01 or so I believe (dont have info in front of me) which is why the above holds true.
I also believe that bonus damage is multiplied directly by the coefficient xml amount. When I had a scale of rank 3 VS with changing bonus damage, I could take the tooltip amount of the lowest bonus damage datapoint, and the highest amount of bonus damage, multiply the rest of equation and then multiply the bonus damage * coef and it works (for any bonus damage datapoint).
Maybe I can post some of the spreadsheet page I was using for this from the hotel room tonight. But no Freehugs, you aren't the only one working on it!
-edit- The X in the equation I have not determined where it comes from yet. For rank 3 VS it was 96. For rank 4 VS it was 121.
Example 1 Rank 3 VS:
Max tooltip: 241
Max Weapon damage: 95
Bonus Damage: 28.8
amountmodifierpercent: 0.0199
coefficient: 1.53999
X = 97
[(X * rank modifier) + Weapon + (Bonus * Coef)] * (1+amountmodifier)
[(97) + 95 + (28.8 * 1.53999)] * (1+.0199)
(192 + 44.349) * 1.0199
241.05
Example 2 Rank 3 VS (with different bonus damage amount):
Max tooltip: 249
Max Weapon damage: 95
Bonus Damage: 34
amountmodifierpercent: 0.0199
coefficient: 1.53999
X = 97
[(X * rank modifier) + Weapon + (Bonus * Coef)] * (1+amountmodifier)
[(97) + 95 + (34 * 1.53999)] * (1+.0199)
244.356 * 1.0199
249.219
Example 3 Rank 3 VS (with different weapon and bonus damage amount):
Max tooltip: 220
Max Weapon damage: 83
Bonus Damage: 23.6
amountmodifierpercent: 0.0199
coefficient: 1.53999
X = 97
[(X * rank modifier) + Weapon + (Bonus * Coef)] * (1+amountmodifier)
[(97) + 83 + (23.6 * 1.53999)] * (1+.0199)
(180 + 36.3416) * 1.0199
220.646
This last example is barely off when rounded up, which most likely has to do with decimals somewhere along the way.
RE: Damage and Healing formula - Kaedis - 10-25-2011 08:48 AM
Quote:Kore I dont think will respond to me on IM any more!
It's been a super busy couple of weeks, this is the first time I've been able to check the forums since the 17th.
Also, I'm still convinced we're treating the amountmodifierpercent wrong. I see no reason they would have overcomplicated the equation with it, especially since the Force equation has no equivalent value. That doesn't mean they haven't, I'll just need a more convincing argument than an equation with it that requires an empirically determined static variable. We also have to take into account that our XML is badly out of date.
RE: Damage and Healing formula - Freehugs - 10-25-2011 10:51 AM
You're wrong about weapon damage not being modified Masterkiller, it's just that the ability you're working with scales roughly one for one. Take a look at Force Leap, it scales much less than that.
I agree that we need to take a step back from the XML data, it's too out of date and we may never get an updated version.
So what we have to go on really is numbers collected from beta, and from the numbers I have we can be pretty certain of the following:
Each ability has a base min and max damage for each rank, and increasing the rank increases both the min and max damage in proportion to each other (but not necessarily the same amount). This may be determined by a formula, but it may also just be a table of values.
Each abilities min and max damage values scale in a linear way, using the same coefficient for both. e.g. 'bonus damage' x 'coefficient' = damage bonus for both min and max values.
Abilities do not scale in a linear way with weapon damage OR different types of weapons effect damage in a different way (i.e. light sabers and vibro swords have a different modifier). In fact, weapon scaling seems very erratic, even an ability like Slash where the values are close, sometimes the damage goes up more than the weapon damage and sometimes it goes up less. There's not even a strong correlation between the way min and max values scale.
I have no data on how damage scales with off-hand weapon's.
So we can determine the formula for Force and Tech powers easily by changing the damage bonus and using that to calculate the coefficient, then use the coefficient to determine the base damage of a rank.
Force/Tech power = Base damage + damage bonus x coefficient
Weapon damage is still to be determined, we might need a better data set than we have.
RE: Damage and Healing formula - Masterkiller - 10-25-2011 11:48 AM
(10-25-2011 08:48 AM)Kore Wrote:I was just joking aroundQuote:Kore I dont think will respond to me on IM any more!
It's been a super busy couple of weeks, this is the first time I've been able to check the forums since the 17th.
(10-25-2011 10:51 AM)Freehugs Wrote: You're wrong about weapon damage not being modified Masterkiller, it's just that the ability you're working with scales roughly one for one. Take a look at Force Leap, it scales much less than that.
I agree that we need to take a step back from the XML data, it's too out of date and we may never get an updated version.
So what we have to go on really is numbers collected from beta, and from the numbers I have we can be pretty certain of the following:
Each ability has a base min and max damage for each rank, and increasing the rank increases both the min and max damage in proportion to each other (but not necessarily the same amount). This may be determined by a formula, but it may also just be a table of values.
Each abilities min and max damage values scale in a linear way, using the same coefficient for both. e.g. 'bonus damage' x 'coefficient' = damage bonus for both min and max values.
Abilities do not scale in a linear way with weapon damage OR different types of weapons effect damage in a different way (i.e. light sabers and vibro swords have a different modifier). In fact, weapon scaling seems very erratic, even an ability like Slash where the values are close, sometimes the damage goes up more than the weapon damage and sometimes it goes up less. There's not even a strong correlation between the way min and max values scale.
I have no data on how damage scales with off-hand weapon's.
So we can determine the formula for Force and Tech powers easily by changing the damage bonus and using that to calculate the coefficient, then use the coefficient to determine the base damage of a rank.
Force/Tech power = Base damage + damage bonus x coefficient
Weapon damage is still to be determined, we might need a better data set than we have.
Hey hey, if you are going to flatly state that "I'm wrong" instead of saying "I dont think that is correct and here is why" at least please be sure to provide some evidence why.
Looking at Force Charge/Leap, the amountmodifier is -0.39000 which means each point of weapon damage is worth .61 in my formula. Do you have any datapoints I can plug this in at? From the Force Charge datapoints I had, I believe I just had to modify X to the proper amount and the rest scaled accurately. I'll see if I can get a hold of my spreadsheet (it's 300 miles away).
RE: Damage and Healing formula - Kaedis - 10-25-2011 12:03 PM
Quote:So we can determine the formula for Force and Tech powers easily by changing the damage bonus and using that to calculate the coefficient, then use the coefficient to determine the base damage of a rank.
Force/Tech power = Base damage + damage bonus x coefficient
Functionally, we can do the same with weapon damage attacks, we just have to independently determine the weapon coefficient, the bonus coefficient, and the static additive. The amountmodifierpercent is only really relevant if we're trying to convert the XML to ability damage. If we're assuming an empirical test is needed to determine coefficient and base, the formula simplifies to WpnDmg*WpnCoeff + BonusDmg*BonusCoeff + C.