|
|
| Author |
Thread Statistics | Show CCP posts - 0 post(s) |
Synner Zerg
SVER True Blood
Unclaimed.
3
   |
Posted - 2013.02.13 16:44:00 -
[1] - Quote
Skill levels are stacked, but different skills' bonuses are compounded, and so are the mods.
For example:
Assault Rifle Base Damage: 30
Weaponry: LVL4 (2% per level)
Assault Rifle Proficiency: LVL2 (3% per level)
2 x Damage Modules: 10% each
1) First, the individual skills are stacked.
It is NOT
1.02 * 1.02 * 1.02 * 1.02 like was posted in OP. This would have given a multiplier of 1.08243215 for Weaponry
It IS
1 + (0.02 * 4) = 1.08 for Weaponry
1 + (0.03 * 2) = 1.06 for Proficiency
This may seem like a small enough difference (especially with CPU/PG rounding down), but this would produce completely different results on higher base stats and bonuses
2) Next, different skills' bonuses are compounded
It is NOT 8% + 6% = 14% = 1.14 multiplier (if you would just add the bonuses)
It is NOT 1.02 * 1.02 * 1.02 * 1.02 * 1.03 * 1.03 = 1.488352278544 (if you would compound everything)
It IS
1.08 * 1.06 = 1.1448 for Weaponry and Proficiency. Note we got the individual multipliers in previous step.
For a full calculation:
(1 + (0.02 * 4)) * (1 + (0.03 * 2)) = 1.1448
3) Last, modules' bonuses are compounded
It is NOT 8% + 6% + 10% + 10% = 1.34 multiplier (if you would just add all the bonuses)
It IS 1.08 * 1.06 * 1.1 * 1.1 = 1.385208 multiplier
Or with stacking penalty (if it would work). Note that stacking applies to mods bonuses only. Assuming penalty of 1 for first mod (i.e. nothing), and 0.869 for second mod
1.08 * 1.06 * (1 + (0.1 * 1) ) * (1 + (0.1 * 0.869) ) = 1.368711432
So, as you can see, having more different skills/modules affecting the same attribute produces more bonuses than comparative higher levels of a single skill. Just for example, lets assume 2 skills, both giving 10% bonus to something.
Only Skill 1 at LVL4 = 1 + (0.1 * 4) = 1.4
Skill 1 at LVL3 and Skill 2 at LVL1 = (1 + (0.1 * 3) ) * (1 +(0.1 * 1) ) = 1.43
In most cases, it will be cheaper to train another skill to LVL1, than to bring the first skill to LVL4 (or higher)
Yes: 2 modules with 10% bonus each is better than 1 module with 20% bonus (in attribute benefit only... of course you are spending 1 extra slot with 2 modules) |
Synner Zerg
SVER True Blood
Unclaimed.
3
|
Posted - 2013.02.13 16:49:00 -
[2] - Quote
Altina McAlterson wrote:Training level V of a skill that gives you 5% per level gives you a bonus of 25%.
Correct, cause individual skill levels are stacked.
Quote:A 10% mod and a 5% mod would give you 15% bonus. Doesn't seem like that's what is being said here.
Incorrect, cause multiple mods/skills are compounded. So in this case 10% mod and 5% mod give you 15.5% bonus
Quote:I know the fitting screen says 2 damage mods gives a 1.21
This just proves the above about compounding
This is the way things have worked in EvE for a long time
|
Synner Zerg
SVER True Blood
Unclaimed.
3
|
Posted - 2013.02.13 22:07:00 -
[3] - Quote
Altina McAlterson wrote:
No it doesn't. It only proves that the damage mods are broken.
Modules are not compounded nor are skills. Weaponry V plus Proficiency V would be 25% bonus to weapon damage, not 26.5%. The only compounding is between the total bonus from all mods and the total bonus from all skills.
Well, you can play EvE and figure it out for yourself (cause in EvE, the fitting screen actually works, and shows all your bonuses with skills, mods, ammo loaded, booster and implant affects applied), or you can continue to believe what you want to believe
|
Synner Zerg
SVER True Blood
Unclaimed.
3
|
Posted - 2013.02.14 03:59:00 -
[4] - Quote
Altina McAlterson wrote:
Modules are not compounded nor are skills
OK, lets dispute that right there. I will show the stats shown in-game backed up by math
Buy a Sica. You don't need skills to use it, and it's only 136k for this one test.
Then buy 3x Militia Shield Regenerator. Again, no skills to use it. It provides 10% bonus
I have shield Boost Systems at LVL1, so that's 3% Shield Recharge Rate to the base of Sica
Base: 22.0 <- from market screen
After 3% skill bonus: 22 * 1.03 = 22.66 <- from my fitting screen
After 1 Shield Regenerator added = 24.93 <- from my fitting screen
After 2 Shield Regenerators added = 27.09 <- from my fitting screen
After 3 Shield Regenerators added = 28.64 <- from my fitting screen
Now, lets see which math supports the above.
If bonuses for modules are not compounded, but added up, then it would be
10% + 10% + 10% = 30% bonus
22.66 * 30% =
22.66 * 1.30 = 29.458 (29.46 rounded)
Above is wrong, that's not what the stats page shows.
Let's try to apply stacking penalty:
10% + (86.91% of 10%) + (57.05% of 10%) =
10% + 8.691% + 5.705% = 24.396% bonus
22.66 * 24.396% =
22.66 * 1.24396 = 28.1881336 (28.19 rounded)
Above is wrong, that's not what the stats page shows.
Now, lets try to compound module bonuses with stacking
(Remember, that percentage bonuses act as multipliers)
(1+0.1) * (1+(0.1*0.8691)) * (1+(0.1*0.5705)) =
1.1 * 1.08691 * 1.05705 = 1.26381003705 compound multiplier (or 26.381003705% bonus)
22.66 * 1.26381003705 = 28.637935439553 (28.64 rounded)
Above matches exactly what the fitting screen shows
I don't know why we had to go through all this, we already know this from 2x 10% damage modifiers showing up as 1.21 (1.1 * 1.1), not 1.20 (10% + 10%). Damage mods aren't broken, the stacking penalty on them is not applied. But since you are convinced that they are broken, i gave you another example above with different module and attribute. Is the Shield Regenerator also broken?
Fine, lets take yet another example.
On the same Sica
Base speed: 24.00 <- from fitting screen
After 1 Nanofiber Structure I (10% each) = 26.40 <- from fitting screen
After 2 Nanofiber Structure I (10% each) = 28.69 <- from fitting screen
With your additive math:
24 * (10% + 10%) =
24 * 20% = 28.80
Not what the fitting screen shows
With your additive math including stacking penalty:
24 * (10% + 8.691%) =
24 * 18.691% = 28.48584 (28.49 rounded)
Not what the fitting screen shows
With compounding and stacking penalty:
24 * 1.1 * 1.08691 = 28.694424 (28.69 rounded)
Again, matches fitting screen exactly
Not enough? Let's take another module and attribute
Sica base armor: 770
Field Mechanics 3x 5% = 15% bonus from skills
Sica armor after skill = 885.5 (885 rounded down)
Shield/Armor/CPU/PG always seems to be rounded down, rather than up, and fitting screen shows no decimals
Same Nanofiber Structure I takes away 10% from armor
Armor 885 <- from fitting screen
After 1 Nanofiber Structure I (-10% each) = 796 <- from fitting screen
After 2 Nanofiber Structure I (-10% each) = 716 <- from fitting screen
With your math:
885 - 20% = 708
Not what the fitting screen shows
With compounding
885 * (1+ -0.1) * (1+ -0.1) =
885 * 0.9 * 0.9 = 716.85 (716 rounded down)
Again, exactly matches the fitting screen
Note that while bonuses are stacking penalized, the drawbacks are not.
So, either everything single module is "broken", and the fact that EvE math (that has been studied and documented for a decade) fits perfectly is just a coincidence... or the fact that EvE math fits perfectly means that this is actually what is happening.
I am not here to convince you.
But to the OP, i hope this clears up things for you.
|
Synner Zerg
SVER True Blood
Unclaimed.
3
|
Posted - 2013.02.14 04:35:00 -
[5] - Quote
Sigh.... read and comprehend
Synner Zerg wrote:Skill levels are stacked, but different skills' bonuses are compounded, and so are the mods
Just so that it's clear:
LVL2 of 10% skill = 1.10 * 2 = 1.20 (20% bonus)
2 different skills each at LVL1 giving 10% each = 1.1 * 1.1 = 1.21 (21% bonus)
Altina McAlterson wrote:Heavy Type-II has base armor stat of 650. With Mechanics V it has 812 armor.
650 x ( 1 + (.05 x 5)) = 812
812.5 actually, which is rounded down to 812.
Yes. Correct. Nothing i said contradicts this! Individual skill levels are stacked (additive)
Altina McAlterson wrote:
Assault Type-II has base shield rate of 25. With Shield Boost Systems III the shield rate is 27.25. With 2 Basic Shield Rechargers the rate is 35.42.
25 * ( 1 + ( .03 * 3 )) * ( 1 + ( .15 * 2 )) = 35.42
Close result, wrong calculation. The modules are compounded between each other, and don't forget the stacking penalty
25 * ( 1 + (.03 * 3 )) * ( 1 + .15) * ( 1 + (.15 * .8691 )) = 35.4228131875 = 35.42 rounded
You are showing me that with sufficiently small numbers (we are talking about percentages of percentages here after all), and with the help of rounding, either calculation is close enough.
However, your numbers don't contradict my calculations. Yet i've shown that with larger numbers, my calculations still hold true, while yours are no longer accurate. Just try the numbers with a Sica like i explained and see for yourself
Altina McAlterson wrote:
You need to learn to check things out for yourself
Oh yeah, i did. The proof is above and in my findings with a Sica
Altina McAlterson wrote:
And you can shut the f*ck up the next time adults are talking or you can continue to be an idiot.
LOL! 
|
Synner Zerg
SVER True Blood
Unclaimed.
3
|
Posted - 2013.02.14 06:00:00 -
[6] - Quote
Altina McAlterson wrote:Ok, I've completely lost my mind on this thread. I'm sorry I told you to shut the **** up. I've acted like a total ass and it was mostly uncalled for.
Hey, no hard feelings. I accept your apology and apologize for a harsh approach myself. My reason is that i just finished the maths on my spreadsheet, figuring all this out with limited testing that we can do with Dust in-game fitting tool. It doesn't help that the in-game fitting tool does have bugs too:
https://forums.dust514.com/default.aspx?g=posts&t=58050&find=unread
Plus, like we saw, in small numbers and small quantity of mods, both calculations are same after rounding, which can be misleading. Easy oversight to make.
Glad we cleared this up. I'd wish you to "fly safe", but what's the Dust equivalent? Kill lots?  . Either way, have a good one |
|
|
| |
|