|
|
| Author |
Thread Statistics | Show CCP posts - 0 post(s) |
KEROSIINI-TERO
Seraphim Initiative.
CRONOS.
471
   |
Posted - 2013.06.19 00:30:00 -
[1] - Quote
Math is the reason.
As far as I know stacking penalties haven't been implemented on reductive modifiers, and even if they would they wouldn't apply to first module (because it is a module stacking penalty).
Math, because if you would have three 50% resistance mods and calculated by adding percentages, it would go like this
"0,50+0,50+0,50 = 1,50 = 150%" which would mean that if you take damage you would get negative damage. Which can never happen ofc.
Now, where that 23,5% res comes from:
Every additional resistance bonus is calculated from the remaining amount. The following should clarify.
Base resistance 0% =0,00
with Skills 10% = 0,10
now as you calculate 15% resistance bonus, it reduces the remaining 'damage hole'.
Remaining amount in numbers: 1,00 - 0,10 = 0,90
Now the 15% is calculated on that remaining damage hole:
0,90 x 0,85 = 0,765 which is the new damage hole.
In order to see what is the reduction amount instead of the hole:
1,00 - 0,765 = 0,235 = 23,5% resistance.
As you see the math dictates the following:
- Each additional reduction bonus (resistance for example) has lesser effect, even without any balancing penalties
- It is impossible to reach 100% resistance (without 100% resistance mod, that is)
- Traditionally reductive bonuses haven't had penalties because of this
- Multiplying bonuses work exactly the opposite, following bonuses have increased effect so that's why penalties are needed for them |
KEROSIINI-TERO
Seraphim Initiative.
CRONOS.
472
|
Posted - 2013.06.19 16:53:00 -
[2] - Quote
Winsaucerer wrote:
.
.
So given your calculation we can recover the 23.5%. Now there's other situations I still cannot calculate.
New scenario: 2x 15% resistance plates, plus standard 10%, is listed as 33.47% in game.
If I do each plate separately, taking 10%, then 15% of the remaining, then 15% of the remaining, I get 34.975% (order doesn't matter). This is higher than 33.47%, of course :)
If I calculate the 15% as a resist on the remaining 0.9 after applying 10% damage reduction, it gives 13.5%. If I then add these together using the stacking penalty, it comes up with a total resist of 25.77%. This gives a final resist of 33.19%. Closer to 33.47%, but still not correct.
So not quite sure how to deal with situations where there are multiple resistance modules.
.
.
Ok, as I don't have access to games resistance results, I checked your value of 10%skill and 2x 15% res mod ==> 33,47% res
I tested with stacking penalty values from http://www.eve-wiki.net/index.php?title=Stacking_penalty
and found out that:
Base: 10% skill = 1-0,10 = 0,90
1st 15% res mod = 1-0,15 = 0,85
2nd 15% res mod WITH stacking penalty = (1- 0,15*0,87) = 0,8695
All those multiplied are
0,6651675
and
1-0,6651675 = 33,48325% resistance after skill, unpenalised mod and first penalised mod
This is so close to value (33,47%) you reported it is possible that it's some rounding issue. |
KEROSIINI-TERO
Seraphim Initiative.
CRONOS.
475
|
Posted - 2013.06.21 00:24:00 -
[3] - Quote
Yea the 'traditional' ccp ideology of the order of applying stacking penalties is indeed
- Best first (unpenalized)
- Second best then (first stacking penalty
- Third mod (second penalty)
etc
If that's not the case in dust then we can say that's a bug, or say "That's Dust". |
KEROSIINI-TERO
Seraphim Initiative.
CRONOS.
489
|
Posted - 2013.06.25 16:29:00 -
[4] - Quote
Garth Mandra wrote:I wonder if the code works correctly for modules that increase numbers (damage mods etc) and that the bug is that modules that reduce numbers (resistance mods) are treated the same way, ie. in the opposite order.
If the code orders the modules largest number first then:
For damage mods, if you had one of each:
complex 10% = 1.1 - no penalty
enhanced 5% = 1.05 - 1st penalty
basic 3% = 1.03 - 2nd penalty
For resistance mods (I'll make up some numbers):
basic 3% = 0.97 - no penalty since it has the largest number
enhanced 5% = 0.95 - 1st penalty
complex 10% = 0.9 - 2rd penalty since it has the smallest number
Hmm, I don't think that's the case as it's possible to calculate resistances without using the 1-x style (0,85) which I used in order to explain things easier.
But I won't be checking that as well.
As above poster, I hope CCP has a quick look at how dust handles.
they have the real values and they know how they intend things to be. Just hoping that they shed some light on it.... |
|
|
| |
|