PDA

View Full Version : [Balance][Combat Mechanics] Resistance Calculation


DeathZer0
01-11-2009, 06:22 PM
--------------------------------1.2 Update--------------------------------------
Since the 40% soft cap change, this topic is out of date and the problem will probably be smaller.

Resists - cap seems to be about 700. Hard numbers :

776 = 42.1%
813 = 42.8%
1191 = 50.3%This is from the Sorc 1.2 PTR LIVE THREAD

According to this info the resistance needed for +1% mitigation past 40% will now be around 50 resistance, roughly 3 times as much as below 40%. This means the resistance cap of 75% mitigation will now be reached at ~2422. This is assuming that the increase is still linear (which it seems to be) and that the 50 resi per 1% mitigation is valid for all remaining 35%.

The resistance talisman nerf will also help to diminish the problem.

However, the calculation is still linear which means big survivability changes for small resistance/mitigation changes at high resistance/mitigation levels.

--------------------------------End of 1.2 Update-------------------------------

Disclaimer: This topic not only tries to explain how mitigation gain from the resistance stat works, but also why the current system might create some imbalances and presents a possible solution. I hope this is in the right forum.

While looking through a lot of discussion threads about resistance I noticed that the way how it is calculated seems to give the mitigation a big advantage for high values of resistance.
The problem is that the percentage gain is linear. I believe ~17 points of resist give you 1% mitigation at 40 (sorry I haven't personally tested the values yet).
This means if you already have a high percentage you'll live considerably longer if you gain only a little more resistance.

Some calculations:
5000 Max HP, 1000 damage per second before mitigation.

Now with 0% it is clear that you will live 5 seconds. If you now gain 5% (+85 Resist @ 40) it looks like this:
1000 * 0.95 = 950 dps after mitigation.
5000 / 950 = 5.26 seconds alive.
5.26/5 = 1.052

This means you'll live 5.2% longer.

Now the same thing with 70% starting mitigation (~1190).
1000 * 0.30 = 300
5000 / 300 = 16.666...

However, after another 5% (~1275 or 75%) it looks like this:
1000 * 0.25 = 250
5000 / 250 = 20
The difference: 20 / 16.66... = 1.2
With the same +85 resistance you'll be able to increase your lifetime by 5.2% or by as high as 20%. It only depends on how much resist you already have.

Personally I think the life time increase should be linear and not the mitigation percentage increase. This would definately make it harder to stack resistances with such devastating results. I'd recommend something like 1% additional life time per 15 or maybe even 10 resistance at level 40.

CALCULATIONS FOR PROPOSED CHANGE

This would be the formula for a lifetime increase:

The Constant is the Resistance needed for your character to be alive 1% longer.

Damage * Multiplier = Final Damage
Multiplier = ( 1 / ( 1 + Resistance / Constant / 100 ) )

So if 10 resist is needed for 1% longer life 1000 resist would increase your life expectancy by 100%. Thus the Multiplier will be 1/2 or 50%
2000 resist would increase your life by 200%, thus multiplying incoming damage of that type by 1/3 or giving you a mitigation of 66.6%
3000 resist equals 300% more life, which means a multiplyer of 1/4. Now you'd reach the cap of 75% mitigation.

Now the same for 15 resist needed for 1% longer lifetime:
500 Resist = Damage * 0.75 = 25% Mitigation
1000 Resist = Damage * 0.6 = 40%
1500 Resist = Damage * 0.5 = 50%
2000 Resist = Damage * 0.429 = 57.14%
2500 Resist = Damage * 0.375 = 62.5%
3000 Resist = Damage * 0.333 = 66.66%
3500 Resist = Damage * 0.3 = 70%
4000 Resist = Damage * 0.273 = 72.72%
4500 Resist = Damage * 0.25 = 75%

1% per 10 would turn a now 75% resistance (~1275) into a 56.04% resistance. Or in other words: your 1000 fireball that would have hit for 250 now hits for ~440 or 1.75 times harder.
1% per 15 would turn a now 75% resistance into a 45.96% resistance.

Obviously, the purpose of the new system would be to make stacking resistance a lot harder than it is now, so that more tradeoffs had to be considered to reach similar values of mitigation.

Hope this gave you an insight of how resistance work, how it could work and how to get the most out of the system (i.e. stack even more resistance).

Rok
01-11-2009, 10:02 PM
With the same +85 resistance you'll be able to increase your lifetime by 5.2% or by as high as 20%. It only depends on how much resist you already have.

Quite right except it's 16.8 Reset per 1% at level 40. So at 1260 you have 75% mitigation, but enemies can debuff as much as 378 and often debuff their own damage type so you need 1638 resistance to really have that level of mitigation.

So my question is, can that level on resistance be achieved with current gear? What has to be sacrificed to get this amount of resistance and is it worth the trade when you're only mitigating 1 damage type?

DeathZer0
01-11-2009, 10:32 PM
Deffinately possible. I think there've been several cases of people achieving resists which would be equal to 100% mitigation, which should basically still give them the full 75% after being debuffed (however, I don't know how or even if resist debuffs stack. of some I know that they don't).

From my experience 10 resist points are equal to 1 stat point. At least that's how it works on talismans. +4 strength and +40 corporal resistance can be crafted from the same material level. Which means the most powerful resistance talisman that can be made would be +200 (will cost you several hundreds of gold though).
I think +150 ones should be fairly available, too. Also +100 can be crafted from regular 38+ green items that have that stat (tier 8, skill level 175, common fragment; with a couple of t9, skill level 200 ingridents yields a common 8h t9 talisman; if it's only t8 will still give you +90 resist).

Also there's usually quite some resistance on the right gear. A lot of ~40 gear, especially rings, have 150-200 resistance against one or two types of damage.
So I'd say around 1000 isn't hard to achieve currently and won't have a lot of tradeoffs.

Edit: Most items with 150+ resists are either drops or are from public quests according to WARDB. And of course talismans.

Alkirin
01-12-2009, 02:45 PM
100+ resist talismans are rather very easy to get. They actually get a bit costly beyond that, but the 100 ones are common, and drop often enough from Level 34 range mobs/quests/pqs.

With the +100 to multiple types given from all manner of accessories and most armors at that level and beyond, stacking to 1000 resist is cheap, easy, and leaves room for minor tweaking.

There are sorcs and BWs (including myself) who make it a point to stack against each other's damage. Though, BWs still have a pretty nasty corp resist that stacks with their corp dots, alot don't bother to target each other anymore.

DeathZer0
02-17-2009, 11:26 PM
Updated the first post with new 1.2 info.

Khaelann
03-11-2009, 02:49 AM
2008 Resist is 66.5%, dont think adding 400 to that would get you to 75%.

DeathZer0
03-11-2009, 10:36 AM
Just did the calculations again using your data.

It's 16.8/1% mitigation pre and 50.415/1% post 40%.

I got the 50.415 using this formula:

(2008 - 16.8 * 40) / (66.5 - 40) = 50.415

400 more resist will give you 66.5 + (400/50.415) = 74.43

So yes adding 400 resistance will get you almost 75% mitigation.

This would improve your EHP by about 40%. This means if you have 5000 hp 66.5% mitgation will mean you effectively have ~14000 HP against that damage type. 75% would be 20000 HP. That's about 40% more HP.

If you allready have 2008 resistance, I'd definitly get the additional 400 if possible.
40% more HP even if only against a single damage type are still pretty good in my opinion.

Rok
04-16-2009, 05:02 PM
It's 16.8/1% mitigation pre and 50.415/1% post 40%.

Looking at the numbers in your first and last post I get different values past 40%. At first I thought the formula had a slight curve to it but then I realised it's just rounding errors: The game rounds off to the nearest 0.1%. +/- 0.05% gives you a range between 50.32 and 50.75.
If we collect some more hard numbers we could probably narrow that down.