Plate Armor - Disadvantage for Paladins? - Page 2

CoLD MeTaL · 10-04-2005, 11:03 PM

First, I don't realy care about this one way or the other, there isn't a lot you can do to significantly affect either number in the discussion.

Kalera wrote:
Actually, it is that simple. It's just math.

The way I read mitigation is percent of damage, and avoidance is a percent to NOT TAKE damage.

It's on a blow by blow basis.

It doesn't matter if it is 'blow by blow.' You must not understand math very well.

So I still think that the equation is meaningless.

The equation is right on actually. It's your example that is off. In your first example, you list no avoidance. In the second example, you do.

In order to compare numbers and the relative merit of avoidance vs. mitigation, you have to compare two sets TOGETHER: avoidance AND mitigation for EACH sample. That's what Naari did above, and what you failed to do. Your argument is invalid because you left out valuable data in your calculations.

Let's look at Naari's numbers again, only this time we'll use the percentages to develop real numbers:

Example 1: 35% mitigation, 22% avoidance

Assume a mob outputs 1000 points over the course of a fight. Avoidance is applied first, at 22%, so you will avoid 220 points of damage. The remaining 780 points of damage is mitigated at 35%, so you will mitigate 273 additional points of damage, meaning you take, as a total, 507 points of damage.

No, avoidance means there is a 22% chance of NO DAMAGE, so either u get 1000 OR NONE.

Then mitigation is applied, so you have an 88% chance of getting 1000-350 (650) unless I am misunderstanding the term avoidance and the text in game.

Example 2: 33% mitigation, 24% avoidance

Assume the same mob outputs 1000 points over the course of a fight. Avoidance applied first at 24%, you will avoid 240 points of damage. The remaining 760 points of damage is mitigatd at 33%, meaning you will mitigate an additional 250.8 points of damage. So you will receive 509.2 points of damage.

No, avoidance means there is a 24% chance of NO DAMAGE, so either u get 1000 OR NONE.

Then mitigation is applied, so you have a 76% chance of getting 1000-330 (670) unless I am misunderstanding the term avoidance and the text in game.

Which is close to 1 in 4. If you lower that to 20% for a 35% mitigation, then 2 in 10 times you will take 650.

Take it to 40 strike attempts so that you have 10 in 40, and 8 in 40 to compare.

3x10x670 = 20100 (25% avoidance, 33% mitigation)

8x4x650= 20800 (20% avoidance, 35% mitigation)

700 points difference ON AVERAGE, for 5 points of avoidance, and 2 points of mitigation.

Now, it should surprise no one who understands math that those numbers of 507 and 509.2 match up exactlye with Narri's percentages of 50.7% and 50.92%.

So, as you can see, the two numbers together determine how much damage you are going to prevent. They work in tandem. And as Naari concluded from the results:

If you trade mitigation for avoidance 1 for 1 you will take more damage unless your avoidance is above 50%.

Of course, if you're able to make a 2-for-1 trade, or better, then sure - take avoidance. But if switching a piece of gear out nets you a 5% loss in mitigation and you only gain 5% in avoidance, then unless that avoidance is over 50%, you will take more damage.

BTW - Thanks for the sample Naari. I've been wondering about this for a while and was too lazy to work the math out. Glad to see you jump on it.

Comments in yellow.

If mitigation and avoidance work as in your post, really mitigation and avoidance are the same, and it only matters which is APPLIED FIRST. (Order Ops)

Which again I wonder how you KNOW which is applied first.

So the real question for me is, Does avoidance give you a percentage chance to take NO DAMAGE, or does it work like mitigation as a PERCENT of REAL DAMAGE ?

robusticus · 10-05-2005, 12:12 AM

If 1000 dmg is applied in 10 hits of 100, given 22% avoidance and 35% mitigation, on average you would avoid 2 hits for 200 dmg and mitigate 280 dmg so you would take 520. However, I believe avoidance is more random - in that it selects a number from 1 to 100 and if that number is over your avoidance % you take a hit. So in that scenario, you could conceivably avoid every hit or none of them. Of course amount of damage is also random, but mitigation is constant. That's why the equation is meaningless, it requires a sample size much larger than one fight. It might be valid if you measured all of the fights during your time progressing one level.

Plus that whole interrupt thing, which so far I'm hearing you get less interrupts with more avoidance. That's a particular item for paladins - as we are the only tanks who need to cast spells to improve our tanking ability.

Message Edited by robusticus on 10-04-2005 01:32 PM

Message Edited by robusticus on 10-04-2005 01:33 PM

Kale · 10-05-2005, 12:52 AM

CoLD MeTaL wrote:
If mitigation and avoidance work as in your post, really mitigation and avoidance are the same, and it only matters which is APPLIED FIRST. (Order Ops)

Which again I wonder how you KNOW which is applied first.

Oh man... this is priceless SMILEY

How do you know? Simple: if you avoid a blow, then there is no need to mitigate any damage. If you do not avoid a blow, then you must mitigate the damage.

Kale · 10-05-2005, 12:58 AM

robusticus wrote:

If 1000 dmg is applied in 10 hits of 100, given 22% avoidance and 35% mitigation, on average you would avoid 2 hits for 200 dmg and mitigate 280 dmg so you would take 520. However, I believe avoidance is more random - in that it selects a number from 1 to 100 and if that number is over your avoidance % you take a hit. So in that scenario, you could conceivably avoid every hit or none of them. Of course amount of damage is also random, but mitigation is constant. That's why the equation is meaningless, it requires a sample size much larger than one fight. It might be valid if you measured all of the fights during your time progressing one level.

That's what randomness does, and that is why a large sample size is necessary. We're not discussing one fight here (well, maybe ColdMetal is, but you can't measure the true benefit of Mitigation or Avoidance over the course of one fight). We're talking about what the advantage of avoidance and mitigation are over a long period of time. Armor selections only come along every 6-10 levels. If you purchase Fulginate at level 40, you can wear that to level 50 and beyond. So the question is, over time, is it better to have high mitigation, or high avoidance? The obvious answer is both, but the mechanics of the game prevent you from doing both well. And as Naari showed us, avoidance only overtakes Mitigation after it has reached 50%.

CoLD MeTaL · 10-05-2005, 12:58 AM

Kalera wrote:

Oh man... this is priceless

How do you know? Simple: if you avoid a blow, then there is no need to mitigate any damage. If you do not avoid a blow, then you must mitigate the damage.

Congratulations on not reading the post. And proving my point as well.

lisasdarr · 10-05-2005, 01:07 AM

CoLD MeTaL wrote:

Kalera wrote:Example 1: 35% mitigation, 22% avoidanceAssume a mob outputs 1000 points over the course of a fight. Avoidance is applied first, at 22%, so you will avoid 220 points of damage. The remaining 780 points of damage is mitigated at 35%, so you will mitigate 273 additional points of damage, meaning you take, as a total, 507 points of damage.
No, avoidance means there is a 22% chance of NO DAMAGE, so either u get 1000 OR NONE.

Then mitigation is applied, so you have an 88% chance of getting 1000-350 (650) unless I am misunderstanding the term avoidance and the text in game.

You have just clearly shown you have no concept of what we are discussing, this is not about a single blow, it is about the average result over a significant enough period of time for the randomness to be eliminated.

Kale · 10-05-2005, 01:18 AM

CoLD MeTaL wrote:

No, avoidance means there is a 24% chance of NO DAMAGE, so either u get 1000 OR NONE.

Then mitigation is applied, so you have a 76% chance of getting 1000-330 (670) unless I am misunderstanding the term avoidance and the text in game.

Yes, you do not understand avoidance. It is not a 24% change to take 'all or nothing'. If you are fighting a mob, you do not have a 24% chance to avoid all damage. You have a 24% chance per attack to avoid damage of that attack. This is why we take a large sample size, and apply the forumal, because it gives an accurate, fine-detail picture of the relationship between the two things. And your example is (once again) statistically wrong. You increase Mitigation by 2% in your second sample, but you lower avoidance by more than double, to 50%! How is that a fair comparison? If you want to see the real benefits, you have to make fair swaps.

Kale · 10-05-2005, 01:21 AM

CoLD MeTaL wrote:

Kalera wrote:Oh man... this is priceless How do you know? Simple: if you avoid a blow, then there is no need to mitigate any damage. If you do not avoid a blow, then you must mitigate the damage.

Congratulations on not reading the post. And proving my point as well.

I read your post. I don't mean this as an insult, but it is clear you don't have any experience with math or statistics, or else you would at least understand what the key points of this discussion are. I don't know how old you are, but I can only guess you're either young (and thus haven't had a formal education in mathematics yet, or statistics) or you simply haven't pursued a career in any field that requires math/statistics. I do not mean that as a flame in any way. I am simply saying, if you don't have the background to participate in the discussion in a meaningful way, then all of this math and analysis isn't going to make any sense to you. And that is clearly the case here.

robusticus · 10-05-2005, 01:59 AM

Instead of vague equations that don't really mean anything for a single fight, does anybody know what the various differences would be for say

Imbued Ebon Plate Chest/Legs, Cobalt Chain everywhere else

Imbued Ebon Plate Chest/Legs, Cobalt Plate everywhere else

All Cobalt chain

All Cobalt plate

What is the tier 6 rare leather armor and how do you get it and what would that do for MIT and avoidance?

Pre-update, which classes actually used chain armor?

Is it TOTAL weight of your armor, or if you have plate gloves and everything else cloth do you still get the whole penalty to avoidance?

One of their goals with this update was to make it not so much a math problem anymore. Ironic that we have to have all of this analysis over huge samples of data to figure out which armor to buy.

Lawman · 10-05-2005, 02:02 AM

CoLD MeTaL wrote:

I haven't seen anything posted by SOE on how they do damage.

LOL... and you never will...

Majorminor · 10-05-2005, 02:02 AM

robusticus wrote:

What is the tier 6 rare leather armor and how do you get it and what would that do for MIT and avoidance?

Think it's a safe bet to say your Mit would go down, your Avoid would go up.

I'll leave the math stuff the the other posters and I just can't do those nuber things

Naa · 10-05-2005, 04:04 AM

CoLD MeTaL wrote:
If mitigation and avoidance work as in your post, really mitigation and avoidance are the same, and it only matters which is APPLIED FIRST. (Order Ops)

Which again I wonder how you KNOW which is applied first.

So the real question for me is, Does avoidance give you a percentage chance to take NO DAMAGE, or does it work like mitigation as a PERCENT of REAL DAMAGE ?

Others have answered you more than adequately. However, let me say that mathematically, it matters little which is applied first. 7x5 = 5x7

Logically, avoidance is applied first and then mitigation. In real life, you are either hit by the weapon or not. If you are hit, then you take some amount of damage which is mitigated by your armor.

But mathematically, it matters not one iota if you figure it out in the reverse:

1. If he is hit, he would take X damage after the hit is mitigated.

2. Was he hit?

Of course, there are differences in other ways. With avoidance, damage is all or none. You are hit or not any particular round. Also with avoidance, it is only effective within a limited arc in front of the player.

The all or none nature of avoidance makes it streaky. In practice this means that a series of hits that get through can leave you dead if you have poor mitigation because your healers cannot keep up with the damage you are taking. On the other hand, you might sometimes beat something you shouldn't if you get lucky enough on the avoidance rolls.

Naa · 10-05-2005, 04:19 AM

robusticus wrote:
Naari, what is your avoidance, just for curiousity - since you have few interrupt problems? I generally prefer to stun-ward in offensive stance so my avoidance most of the time is 26% with SBS. Healers get extra work (xp) when grouped with me.

Sorry I didn't get back to you earlier on this question. I stepped out for a nice round of golf this afternoon.

I'm level 40. As I mentioned, I wear legendary imbued leather for my chest piece and legs.

Self buffed in offensive mode, my mitigation is 1281 (32%) and my avoidance is 22.9%. That is using my SBH.

Self buffed in defensive mode, my mitigation is still 1281 (32%) and my avoidance is 38.9%. That is using my Guard of Grumm Shield and Moonlight Mace.

I need to get a better shield and weapon. I haven't acquired all my new gear for tier 5 yet.

I could also boost my avoidance some by getting the appropriate ring and using it.

I normally run the offensive stance and either solo or duo with a Mystic. We generally do not take on Heroic mobs but do take on multiple linked mobs. If I was running in a fuller group against heroic mobs and I was the MT, I would run the defensive setup above.

CoLD MeTaL · 10-05-2005, 06:28 PM

I believe the formula you are looking for is over a large enough sample size is :

(Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage )))

Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed.

Message Edited by CoLD MeTaL on 10-05-2005 08:10 AM

Kale · 10-05-2005, 07:40 PM

CoLD MeTaL wrote:

I believe the formula you are looking for is over a large enough sample size is :
(Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage )))
Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed.
Message Edited by CoLD MeTaL on 10-05-2005 08:10 AM

Yep, that's basically the formula. However, the reason to take a large sample size is to negate the skew. You'll get more skewed numbers with a smaller sample size. But Naari brings up the best point of all: Avoidance can actually be unpredictable, due to the hit/no-hit nature of it. You can get pelted for a dozen hits in a row, or not. And, like she (he?) said, it has a limited arc. Mitigation, on the other hand, is rock-solid. You know you're going to mitigate X amount of damage every single hit. As a plate tank, our avoidance is never going to be very high anyway (we're never going to approach that 60% or more range that makes it truly effective). As a plate tank, we sort of have to assume we're going to get hit. That's our job. Our skill is really in mitigating the damage. That's the point of plate armor - to absorb the blow. Right now, if there is a flaw in the game, it is that avoidance and mitigation for a plate tank is even in question. If we were going to model fantasy warfare a bit more accurately, this shouldn't even be a question. We should be opting for plate at every chance, because we should be able to have better mitigation than we could ever have avoidance, due to our training with heavy armor. I think what might be lacking here is that we aren't trained in Heavy Armor. We can wear it (because our skill allows it), but we don't have any additional training with it. So we aren't any better fighters wearing plate than anyone else. What the developers might want to look into is a skill for armor that actually matters. Sure, a priest can wear plate, but should they be as skilled with it as a dedicated fighter, like a Paladin or Shadowknight? Not really. Our skill with plate armor should improve our mitigation, such that choosing to wear lighter armor is actually a bad decision.

Naa · 10-05-2005, 08:00 PM

CoLD MeTaL wrote:

I believe the formula you are looking for is over a large enough sample size is :
(Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage )))
Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed.

Message Edited by CoLD MeTaL on 10-05-2005 08:10 AM

Congrats on needlessly complicating things:

What you have written can be simplified as follows:

damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation% * 1)

damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation%)

damage taken = Full damage * (1 - Avoidance% * 1) * (1 - Mitigation%)

damage taken = Full damage * (1 - Avoidance%) * (1 - Mitigation%)

damage taken / full damage = 1 * (1 - Avoidance%) * (1 - Mitigation%)

damage taken / full damage = (1 - Avoidance%) * (1 - Mitigation%)

percent damage taken = (100% - Avoidance%) * (100% - Mitigation %)

percent damage taken = (100% - Mitigation%) * (100% - Avoidance%)

which by the way is identical to

% of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100%

Now where on earth or the heavens above have I seen that formula before?

Of course, if you really like doing the extra and needless calculations just because you like punching buttons on a calculator, have at it :smileyvery-happy:

Yrield · 10-05-2005, 08:17 PM

Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course.

robusticus · 10-05-2005, 08:25 PM

Who says we're plate tanks? I don't think we are anymore. Warriors are plate tanks.

I really wish people would answer questions instead of bashing on one another.

Looks like leather chest/legs gives 8-10% more avoidance. My avoidance in offensive stance with a 2 hander is about 14%...that's with full plate and AGI as my lowest attribute. SBS gives me about 12% more. I have noticed less interruptions since I re-equiped SBS (acquired SM finally)...

I wonder what the impact of the other slots (head, shoulders, arms, gloves, boots) is to the avoidance penalty and if there is a significant jump between chain and leather.

I'm starting to really regret selling my ancient slayer's ring of agility. Plat down the drain if I want it back.

Naa · 10-05-2005, 08:39 PM

robusticus wrote:

Who says we're plate tanks? I don't think we are anymore. Warriors are plate tanks.

I really wish people would answer questions instead of bashing on one another.

Looks like leather chest/legs gives 8-10% more avoidance. My avoidance in offensive stance with a 2 hander is about 14%...that's with full plate and AGI as my lowest attribute. SBS gives me about 12% more. I have noticed less interruptions since I re-equiped SBS (acquired SM finally)...

I wonder what the impact of the other slots (head, shoulders, arms, gloves, boots) is to the avoidance penalty and if there is a significant jump between chain and leather.

I'm starting to really regret selling my ancient slayer's ring of agility. Plat down the drain if I want it back.

The leather on chest and legs doesn't give you near that much. My agility stats are reasonably high although I haven't intentionally been buying equipment to boost it. Just checking my character profile it is 93.

http://eq2players.station.sony.com/en/pplayer.vm?characterId=146015104

That is prior to any buffing. Part of why it is that high is that I am a ratonga. I may want to get myself an agility ring to boost it even more. (If you aren't aware, you don't need to continue wearing the imbued rings after you buff using them - of course it is a pain swapping them in and out. If I recall correctly, the buff lasts 25 minutes).

Yrield · 10-05-2005, 08:49 PM

robusticus wrote:
Who says we're plate tanks? I don't think we are anymore. Warriors are plate tanks.

I really wish people would answer questions instead of bashing on one another.

Looks like leather chest/legs gives 8-10% more avoidance. My avoidance in offensive stance with a 2 hander is about 14%...that's with full plate and AGI as my lowest attribute. SBS gives me about 12% more. I have noticed less interruptions since I re-equiped SBS (acquired SM finally)...

I wonder what the impact of the other slots (head, shoulders, arms, gloves, boots) is to the avoidance penalty and if there is a significant jump between chain and leather.

I'm starting to really regret selling my ancient slayer's ring of agility. Plat down the drain if I want it back.

+12 agi give 0.5% avoidance 1 pts of defense give 0.3% avoidance dunno about parry

robusticus · 10-05-2005, 08:59 PM

Try equiping some plate and soloing some groups with that SBH. See how you like the interruption situation then. Who knows, maybe they fixed it stealth-like.

My AGI too is naturally high and I kept a COUPLE items that boost it but haven't equipped them in a while. I still have way too much WIS though... I need to swap hex dolls... more plat down the drain. I'm ok for now though because screaming mace is so much better than the BBC...but I have prismatic greatsword ambitions... SMILEY

CoLD MeTaL · 10-05-2005, 09:11 PM

Naari wrote:

CoLD MeTaL wrote:

I believe the formula you are looking for is over a large enough sample size is :

(Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage )))

Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed.

Message Edited by CoLD MeTaL on 10-05-200508:10 AM

Congrats on needlessly complicating things:

What you have written can be simplified as follows:

damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation% * 1)

damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation%)

damage taken = Full damage * (1 - Avoidance% * 1) * (1 - Mitigation%)

damage taken = Full damage * (1 - Avoidance%) * (1 - Mitigation%)

damage taken / full damage = 1 * (1 - Avoidance%) * (1 - Mitigation%)

damage taken / full damage = (1 - Avoidance%) * (1 - Mitigation%)

percent damage taken = (100% - Avoidance%) * (100% - Mitigation %)

percent damage taken = (100% - Mitigation%) * (100% - Avoidance%)

which by the way is identical to

% of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100%

Now where on earth or the heavens above have I seen that formula before?

Of course, if you really like doing the extra and needless calculations just because you like punching buttons on a calculator, have at it :smileyvery-happy:

No you are not taking into account that you only mitigate non-avoided damage

22% Avoidance, 35% mitigation, 100000 points over time.

Yours: % of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100%

(100000 - 35000) * (100000 - 22000) / 1

(65000) * (75000) /1 = 5,070,000,000

Your ORIGINAL formula is wrong, which is why no amount of algebra will help, because you are still taking mitigation against a number that doesn't factor in avoidance. You are also trying to work in percents, which stats should have taught you, averages of averages ARE meaningless.

My formula COULD have been simplified to

(1-Mitigation%) * (Full Damage - (Avoidance% * Full damage))

Mine: (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage )))

(100000 - ( .22 * 100000)) - (.35 * (100000 - (.22 * 100000)))

(100000 - 22000) - (.35 * (100000 - 22000))

78000 - (.35 * 78000)

78000 - 25740 = 52260 marginally different from your formula. (Off course statistically meaningless because comps don't generate truly random numbers)

Avoidance Mitigation Damage Avoidance AvoidedBase MitigationTotal Percent

22% 33% 100000 22000 78000 25740 52260 52.26%

23% 32% 100000 23000 77000 24640 52360 52.36%

24% 31% 100000 24000 76000 23560 52440 52.44%

25% 30% 100000 25000 75000 22500 52500 52.50%

26% 29% 100000 26000 74000 21460 52540 52.54%

27% 28% 100000 27000 73000 20440 52560 52.56%

28% 27% 100000 28000 72000 19440 52560 52.56%

29% 26% 100000 29000 71000 18460 52540 52.54%

30% 25% 100000 30000 70000 17500 52500 52.50%

31% 24% 100000 31000 69000 16560 52440 52.44%

32% 23% 100000 32000 68000 15640 52360 52.36%

33% 22% 100000 33000 67000 14740 52260 52.26%

34% 21% 100000 34000 66000 13860 52140 52.14%

35% 20% 100000 35000 65000 13000 52000 52.00%

36% 19% 100000 36000 64000 12160 51840 51.84%

1) I doubt you will find the majority of game items trade 1 for 1 avoidance for mitigation, haven't really paid that much attention to that. 2) there is a point of diminishing resturns on avoidance, but it isn't a hard 50%, it floats based on your mitigation. 3) Since stats like agility move avoidance without altering mitigation, jewelry, etc. will then change your real place on the curve.

CoLD MeTaL · 10-05-2005, 09:30 PM

Yrieldom wrote:
Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll.

if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit.

So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course.

And it is against a enemy your same level, all con infromation affects these calculations.

Naa · 10-05-2005, 09:49 PM

CoLD MeTaL wrote:

Naari wrote:

CoLD MeTaL wrote:

I believe the formula you are looking for is over a large enough sample size is :

(Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage )))

Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed.

Message Edited by CoLD MeTaL on 10-05-200508:10 AM

Congrats on needlessly complicating things:

What you have written can be simplified as follows:

damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation% * 1)

damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation%)

damage taken = Full damage * (1 - Avoidance% * 1) * (1 - Mitigation%)

damage taken = Full damage * (1 - Avoidance%) * (1 - Mitigation%)

damage taken / full damage = 1 * (1 - Avoidance%) * (1 - Mitigation%)

damage taken / full damage = (1 - Avoidance%) * (1 - Mitigation%)

percent damage taken = (100% - Avoidance%) * (100% - Mitigation %)

percent damage taken = (100% - Mitigation%) * (100% - Avoidance%)

which by the way is identical to

% of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100%

Now where on earth or the heavens above have I seen that formula before?

Of course, if you really like doing the extra and needless calculations just because you like punching buttons on a calculator, have at it :smileyvery-happy:

No you are not taking into account that you only mitigate non-avoided damage

22% Avoidance, 35% mitigation, 100000 points over time.

Yours: % of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100%

(100000 - 35000) * (100000 - 22000) / 1

(65000) * (75000) /1 = 5,070,000,000

Your ORIGINAL formula is wrong, which is why no amount of algebra will help, because you are still taking mitigation against a number that doesn't factor in avoidance. You are also trying to work in percents, which stats should have taught you, averages of averages ARE meaningless.

My formula COULD have been simplified to

(1-Mitigation%) * (Full Damage - (Avoidance% * Full damage))

Mine: (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage )))

(100000 - ( .22 * 100000)) - (.35 * (100000 - (.22 * 100000)))

(100000 - 22000) - (.35 * (100000 - 22000))

78000 - (.35 * 78000)

78000 - 25740 = 52260 marginally different from your formula. (Off course statistically meaningless because comps don't generate truly random numbers)

Avoidance Mitigation Damage Avoidance AvoidedBase MitigationTotal Percent

22% 33% 100000 22000 78000 25740 52260 52.26%

23% 32% 100000 23000 77000 24640 52360 52.36%

24% 31% 100000 24000 76000 23560 52440 52.44%

25% 30% 100000 25000 75000 22500 52500 52.50%

26% 29% 100000 26000 74000 21460 52540 52.54%

27% 28% 100000 27000 73000 20440 52560 52.56%

28% 27% 100000 28000 72000 19440 52560 52.56%

29% 26% 100000 29000 71000 18460 52540 52.54%

30% 25% 100000 30000 70000 17500 52500 52.50%

31% 24% 100000 31000 69000 16560 52440 52.44%

32% 23% 100000 32000 68000 15640 52360 52.36%

33% 22% 100000 33000 67000 14740 52260 52.26%

34% 21% 100000 34000 66000 13860 52140 52.14%

35% 20% 100000 35000 65000 13000 52000 52.00%

36% 19% 100000 36000 64000 12160 51840 51.84%

1) I doubt you will find the majority of game items trade 1 for 1 avoidance for mitigation, haven't really paid that much attention to that. 2) there is a point of diminishing resturns on avoidance, but it isn't a hard 50%, it floats based on your mitigation. 3) Since stats like agility move avoidance without altering mitigation, jewelry, etc. will then change your real place on the curve.

A) You don't understand the application of my formula.

B) You don't correctly apply your formula.

Point A demonstrated:

% of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100%

% of full Damage = (100% - 35%) * (100% - 22%) / 100%

% of full Damage = 65% * 78% / 100%

% of full Damage = 50.7%

50.7% * 100000 points of damage over the course of the battle = 50700 points of damage taken

Point B demonstrated:

Mine: (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage )))

(100000 - (22% * 100000)) - (35% * (100000 - (22% * 100000)))

(100000 - (22000)) - (35% * (100000 - (22000)))

(78000) - (35% * (78000))

78000 - (27300)

50,700

Low and Behold, the numbers match.

As for your non math points:

1) You are right most items don't trade 1 for 1 on avoidance vs. mitigation. Since that is true, you will need to run the math if you really want to see which is better. In other words, I agree with this point.

2) I never said there was a point of diminishing returns at 50% of avoidance. What I said, incorrectly, was that above 50% avoidance trumped mitigation 1 for 1. I later retracted that and redid my general rule to be correct. See earlier in the thread. I can admit I was wrong and actually did so before anyone pointed it out.

3) Stats on Jewelry (and on the armor too for that matter) do affect your avoidance. And as such, they certainly do affect your point on the curve. For example, they might be able to get your avoidance higher than your mitigation so that in a fight where the avoidance works (in the arc) it is more effective over the long haul than the mitigation. In otherwords, I agree with this point.

CoLD MeTaL · 10-05-2005, 10:06 PM

Naari wrote:

LOL I wrote 35%, and used 33% in my calculations.

Sorry, thanks for pointing that out.

Naa · 10-05-2005, 10:22 PM

CoLD MeTaL wrote:

Naari wrote:

LOL I wrote 35%, and used 33% in my calculations.

Sorry, thanks for pointing that out.

And thanks for admitting you were wrong.

lisasdarr · 10-06-2005, 02:52 AM

Yrieldom wrote:Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course.

40% avoidance means that over a 'sufficiantly large sample' you will avoid 40% of the hits. If 100 hits is a sufficiantly large sample to counter tha randomness then it doesn mean that you avoid 40 out of 100 hits.

Yrield · 10-06-2005, 07:38 PM

lisasdarren wrote:
Yrieldom wrote:Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course.
40% avoidance means that over a 'sufficiantly large sample' you will avoid 40% of the hits. If 100 hits is a sufficiantly large sample to counter tha randomness then it doesn mean that you avoid 40 out of 100 hits.

Well no. Because you dont have 40% avoidance, you have 40% CHANCE to avoid a hit every roll you do (chance is the keywork here) Sure you can reach 40% of effective avoidance after 100 hits, then redo the test against the same mob and reach 40% of effective avoidance after 3291 hits. To make it simple you can eleminate the randomness of a "% of chance" because every roll you do is not dependant of your previous roll. Its just like lotto, a basic lotto with 1 chance out of 100 to win, buying 100 tickets doesnt mean you have atleast 1 winning ticket, its just mean that you have 100 more chance to win

Rochir · 10-06-2005, 08:54 PM

Plate armor does in fact reduce our avoidance when equipped. When I craft, I take of my weapon, shield, greaves, curiass and gloves. Without my weapon or shield, with all of my armor on, my avoidance is 23.3%. When I unequip chest, legs and gloves, it goes up to 26.3%. 3 pieces of pristine ebon vangaurd amor cost me 3% avoidance. A full suit of plate armor will cost us about 7-8% avoidance. That is a significant hit. Our armor does in fact reduce our avoidance. The mitigation it adds is not sufficient to compensate for the loss of avoidance as our mitigation is capped (which effectively makes it go down as we level up) for each teir and armor type. Avoidance on the other hand is always a percentage and remains constant as we level up.

Whether or not our armor is a liability or a benefit depends on the strength of the mob we are fighting and the type of attacks it is using. Against a sufficiently strong mob, our mitigation becomes ineffective. There are certain attacks that ignore our mitigation. The advantage of avoidance is that it depends only on the level of the mob and works to some extent on all types of attacks and all types of mobs. The big problem with our avoidance is it is frontal only. On the other hand, brawler avoidance is 360 degrees. We must mitigate all damage from behind. It is sad but true that against certain types of mobs in certain situations, we would be better off without any armor on. Another big advantage that brawlers have is that their leather armor gives them the stats and resists they need without significantly reducing their avoidance. Our avoidance is capped lower than theirs so there is no point in our wearing cloth or leather armor to compensate. Our avoidance will never be as high as theirs.

No matter how you look at it, brawlers are in fact the uber tanks in eq2 SMILEY

Message Edited by Rochir on 10-06-2005 12:23 PM

Kale · 10-06-2005, 09:25 PM

Yrieldom wrote:
lisasdarren wrote:
Yrieldom wrote:Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course.
40% avoidance means that over a 'sufficiantly large sample' you will avoid 40% of the hits. If 100 hits is a sufficiantly large sample to counter tha randomness then it doesn mean that you avoid 40 out of 100 hits.
Well no. Because you dont have 40% avoidance, you have 40% CHANCE to avoid a hit every roll you do (chance is the keywork here) Sure you can reach 40% of effective avoidance after 100 hits, then redo the test against the same mob and reach 40% of effective avoidance after 3291 hits. To make it simple you can eleminate the randomness of a "% of chance" because every roll you do is not dependant of your previous roll. Its just like lotto, a basic lotto with 1 chance out of 100 to win, buying 100 tickets doesnt mean you have atleast 1 winning ticket, its just mean that you have 100 more chance to win

Some real rocket scientists at work here... It's not anything like the lotto. A statistics class would really help you understand why. A much better comparison is coin-flipping. 50% chance for heads, 50% chance for tails. Those are the actual percentage chances. However, getting exactly 5 heads and 5 tails doesn't always happen. If you take 10 coin flips, there are 10 possibe outcomes (heads/tails): 0/10, 1/9, 2/8, 3/7, 4/6, 5/5, 6/4, 7/3, 8/2, 9/1, 10/0. Those outcomes, in that order, will actually form a bell curve, with the 5/5 being the highest point on the graph, and the 0/10 and 10/0 being the lowest point on a graph. Meaning, if you were to flip 10 coins, you would have a much better chance of getting 4 heads and 6 tails and you would of getting 0 heads and 10 tails. This is where sample size comes into play. Sufficiently large sample sizes will remove the 'streak nature' from the results, and come closer to the real percentage chance of 50/50. It does no good to look at small sample sizes because anything can happen. That's the random nature of things. So instead of looking at a very small sample (after all, it's possible to avoid 10 hits in a row - not likely, but possible) we look at a very large sample size, to get a much more realistic picture of what will happen on an average intance. The middle portion of the bell curve is the average instance. In terms of avoidance - if you're avoidance is 40%, that doesn't mean you are going to avoid 40% of the hits every fight. But what it does mean is that, on the bell curve, you're probably going to be avoiding close to 40% most of the time. You're going to be closer to avoiding 40% than you are 100% or 0%.

10-04-2005, 11:03 PM	#31
CoLD MeTaL Loremaster Join Date: Apr 2005 Posts: 3,217	First, I don't realy care about this one way or the other, there isn't a lot you can do to significantly affect either number in the discussion. Kalera wrote: Actually, it is that simple. It's just math. The way I read mitigation is percent of damage, and avoidance is a percent to NOT TAKE damage. It's on a blow by blow basis. It doesn't matter if it is 'blow by blow.' You must not understand math very well. So I still think that the equation is meaningless. The equation is right on actually. It's your example that is off. In your first example, you list no avoidance. In the second example, you do. In order to compare numbers and the relative merit of avoidance vs. mitigation, you have to compare two sets TOGETHER: avoidance AND mitigation for EACH sample. That's what Naari did above, and what you failed to do. Your argument is invalid because you left out valuable data in your calculations. Let's look at Naari's numbers again, only this time we'll use the percentages to develop real numbers: Example 1: 35% mitigation, 22% avoidance Assume a mob outputs 1000 points over the course of a fight. Avoidance is applied first, at 22%, so you will avoid 220 points of damage. The remaining 780 points of damage is mitigated at 35%, so you will mitigate 273 additional points of damage, meaning you take, as a total, 507 points of damage. No, avoidance means there is a 22% chance of NO DAMAGE, so either u get 1000 OR NONE. Then mitigation is applied, so you have an 88% chance of getting 1000-350 (650) unless I am misunderstanding the term avoidance and the text in game. Example 2: 33% mitigation, 24% avoidance Assume the same mob outputs 1000 points over the course of a fight. Avoidance applied first at 24%, you will avoid 240 points of damage. The remaining 760 points of damage is mitigatd at 33%, meaning you will mitigate an additional 250.8 points of damage. So you will receive 509.2 points of damage. No, avoidance means there is a 24% chance of NO DAMAGE, so either u get 1000 OR NONE. Then mitigation is applied, so you have a 76% chance of getting 1000-330 (670) unless I am misunderstanding the term avoidance and the text in game. *Which is close to* 1 in 4. If you lower that to 20% for a 35% mitigation, then 2 in 10 times you will take 650. Take it to 40 strike attempts so that you have 10 in 40, and 8 in 40 to compare. 3x10x670 = 20100 (25% avoidance, 33% mitigation) 8x4x650= 20800 (20% avoidance, 35% mitigation) 700 points difference ON AVERAGE, for 5 points of avoidance, and 2 points of mitigation.** Now, it should surprise no one who understands math that those numbers of 507 and 509.2 match up exactlye with Narri's percentages of 50.7% and 50.92%. So, as you can see, the two numbers together determine how much damage you are going to prevent. They work in tandem. And as Naari concluded from the results: If you trade mitigation for avoidance 1 for 1 you will take more damage unless your avoidance is above 50%. Of course, if you're able to make a 2-for-1 trade, or better, then sure - take avoidance. But if switching a piece of gear out nets you a 5% loss in mitigation and you only gain 5% in avoidance, then unless that avoidance is over 50%, you will take more damage. BTW - Thanks for the sample Naari. I've been wondering about this for a while and was too lazy to work the math out. Glad to see you jump on it. Comments in yellow. If mitigation and avoidance work as in your post, really mitigation and avoidance are the same, and it only matters which is APPLIED FIRST. (Order Ops) Which again I wonder how you KNOW which is applied first. So the real question for me is, Does avoidance give you a percentage chance to take NO DAMAGE, or does it work like mitigation as a PERCENT of REAL DAMAGE ? __________________

10-05-2005, 12:52 AM	#33
Kale Join Date: Jan 2005 Posts: 222	CoLD MeTaL wrote: If mitigation and avoidance work as in your post, really mitigation and avoidance are the same, and it only matters which is APPLIED FIRST. (Order Ops) *Which again I wonder how you KNOW which is applied first.* Oh man... this is priceless How do you know? Simple: if you avoid a blow, then there is no need to mitigate any damage. If you do not avoid a blow, then you must mitigate the damage.

10-05-2005, 12:58 AM	#34
Kale Join Date: Jan 2005 Posts: 222	robusticus wrote: If 1000 dmg is applied in 10 hits of 100, given 22% avoidance and 35% mitigation, on average you would avoid 2 hits for 200 dmg and mitigate 280 dmg so you would take 520. However, I believe avoidance is more random - in that it selects a number from 1 to 100 and if that number is over your avoidance % you take a hit. *So in that scenario, you could conceivably avoid every hit or none of them.* Of course amount of damage is also random, but mitigation is constant. That's why the equation is meaningless, it requires a sample size much larger than one fight. It might be valid if you measured all of the fights during your time progressing one level. That's what randomness does, and that is why a large sample size is necessary. We're not discussing one fight here (well, maybe ColdMetal is, but you can't measure the true benefit of Mitigation or Avoidance over the course of one fight). We're talking about what the advantage of avoidance and mitigation are over a long period of time. Armor selections only come along every 6-10 levels. If you purchase Fulginate at level 40, you can wear that to level 50 and beyond. So the question is, over time, is it better to have high mitigation, or high avoidance? The obvious answer is both, but the mechanics of the game prevent you from doing both well. And as Naari showed us, avoidance only overtakes Mitigation after it has reached 50%.

10-05-2005, 12:58 AM	#35
CoLD MeTaL Loremaster Join Date: Apr 2005 Posts: 3,217	Kalera wrote: Oh man... this is priceless How do you know? Simple: if you avoid a blow, then there is no need to mitigate any damage. If you do not avoid a blow, then you must mitigate the damage. Congratulations on not reading the post. And proving my point as well. __________________

10-05-2005, 01:07 AM	#36
lisasdarr Join Date: Dec 2004 Posts: 186	CoLD MeTaL wrote: Kalera wrote:Example 1: 35% mitigation, 22% avoidanceAssume a mob outputs 1000 points over the course of a fight. Avoidance is applied first, at 22%, so you will avoid 220 points of damage. The remaining 780 points of damage is mitigated at 35%, so you will mitigate 273 additional points of damage, meaning you take, as a total, 507 points of damage. No, avoidance means there is a 22% chance of NO DAMAGE, so either u get 1000 OR NONE. Then mitigation is applied, so you have an 88% chance of getting 1000-350 (650) unless I am misunderstanding the term avoidance and the text in game. You have just clearly shown you have no concept of what we are discussing, this is not about a single blow, it is about the average result over a significant enough period of time for the randomness to be eliminated. __________________

10-05-2005, 12:12 AM	#32
robusticus Loremaster Join Date: Apr 2005 Posts: 220	If 1000 dmg is applied in 10 hits of 100, given 22% avoidance and 35% mitigation, on average you would avoid 2 hits for 200 dmg and mitigate 280 dmg so you would take 520. However, I believe avoidance is more random - in that it selects a number from 1 to 100 and if that number is over your avoidance % you take a hit. So in that scenario, you could conceivably avoid every hit or none of them. Of course amount of damage is also random, but mitigation is constant. That's why the equation is meaningless, it requires a sample size much larger than one fight. It might be valid if you measured all of the fights during your time progressing one level. Plus that whole interrupt thing, which so far I'm hearing you get less interrupts with more avoidance. That's a particular item for paladins - as we are the only tanks who need to cast spells to improve our tanking ability. Message Edited by robusticus on 10-04-2005 01:32 PM Message Edited by robusticus on 10-04-2005 01:33 PM

10-05-2005, 01:18 AM	#37
Kale Join Date: Jan 2005 Posts: 222	CoLD MeTaL wrote: No, avoidance means there is a 24% chance of NO DAMAGE, so either u get 1000 OR NONE. Then mitigation is applied, so you have a 76% chance of getting 1000-330 (670) unless I am misunderstanding the term avoidance and the text in game. Yes, you do not understand avoidance. It is not a 24% change to take 'all or nothing'. If you are fighting a mob, you do not have a 24% chance to avoid all damage. You have a 24% chance per attack to avoid damage of that attack. This is why we take a large sample size, and apply the forumal, because it gives an accurate, fine-detail picture of the relationship between the two things. And your example is (once again) statistically wrong. You increase Mitigation by 2% in your second sample, but you lower avoidance by more than double, to 50%! How is that a fair comparison? If you want to see the real benefits, you have to make fair swaps.

10-05-2005, 01:21 AM	#38
Kale Join Date: Jan 2005 Posts: 222	CoLD MeTaL wrote: Kalera wrote:Oh man... this is priceless How do you know? Simple: if you avoid a blow, then there is no need to mitigate any damage. If you do not avoid a blow, then you must mitigate the damage. Congratulations on not reading the post. And proving my point as well. I read your post. I don't mean this as an insult, but it is clear you don't have any experience with math or statistics, or else you would at least understand what the key points of this discussion are. I don't know how old you are, but I can only guess you're either young (and thus haven't had a formal education in mathematics yet, or statistics) or you simply haven't pursued a career in any field that requires math/statistics. I do not mean that as a flame in any way. I am simply saying, if you don't have the background to participate in the discussion in a meaningful way, then all of this math and analysis isn't going to make any sense to you. And that is clearly the case here.

10-05-2005, 01:59 AM	#39
robusticus Loremaster Join Date: Apr 2005 Posts: 220	Instead of vague equations that don't really mean anything for a single fight, does anybody know what the various differences would be for say Imbued Ebon Plate Chest/Legs, Cobalt Chain everywhere else Imbued Ebon Plate Chest/Legs, Cobalt Plate everywhere else All Cobalt chain All Cobalt plate What is the tier 6 rare leather armor and how do you get it and what would that do for MIT and avoidance? Pre-update, which classes actually used chain armor? Is it TOTAL weight of your armor, or if you have plate gloves and everything else cloth do you still get the whole penalty to avoidance? One of their goals with this update was to make it not so much a math problem anymore. Ironic that we have to have all of this analysis over huge samples of data to figure out which armor to buy.

10-05-2005, 02:02 AM	#40
Lawman Loremaster Join Date: Nov 2004 Posts: 76	CoLD MeTaL wrote: I haven't seen anything posted by SOE on how they do damage. LOL... and you never will...

10-05-2005, 02:02 AM	#41
Majorminor Loremaster Join Date: Mar 2005 Posts: 372	robusticus wrote: What is the tier 6 rare leather armor and how do you get it and what would that do for MIT and avoidance? Think it's a safe bet to say your Mit would go down, your Avoid would go up. I'll leave the math stuff the the other posters and I just can't do those nuber things __________________

10-05-2005, 04:04 AM	#42
Naa Join Date: Nov 2004 Posts: 37	CoLD MeTaL wrote: If mitigation and avoidance work as in your post, really mitigation and avoidance are the same, and it only matters which is APPLIED FIRST. (Order Ops) Which again I wonder how you KNOW which is applied first. So the real question for me is, Does avoidance give you a percentage chance to take NO DAMAGE, or does it work like mitigation as a PERCENT of REAL DAMAGE ? Others have answered you more than adequately. However, let me say that mathematically, it matters little which is applied first. 7x5 = 5x7 Logically, avoidance is applied first and then mitigation. In real life, you are either hit by the weapon or not. If you are hit, then you take some amount of damage which is mitigated by your armor. But mathematically, it matters not one iota if you figure it out in the reverse: 1. If he is hit, he would take X damage after the hit is mitigated. 2. Was he hit? Of course, there are differences in other ways. With avoidance, damage is all or none. You are hit or not any particular round. Also with avoidance, it is only effective within a limited arc in front of the player. The all or none nature of avoidance makes it streaky. In practice this means that a series of hits that get through can leave you dead if you have poor mitigation because your healers cannot keep up with the damage you are taking. On the other hand, you might sometimes beat something you shouldn't if you get lucky enough on the avoidance rolls.

10-05-2005, 04:19 AM	#43
Naa Join Date: Nov 2004 Posts: 37	robusticus wrote: Naari, what is your avoidance, just for curiousity - since you have few interrupt problems? I generally prefer to stun-ward in offensive stance so my avoidance most of the time is 26% with SBS. Healers get extra work (xp) when grouped with me. Sorry I didn't get back to you earlier on this question. I stepped out for a nice round of golf this afternoon. I'm level 40. As I mentioned, I wear legendary imbued leather for my chest piece and legs. Self buffed in offensive mode, my mitigation is 1281 (32%) and my avoidance is 22.9%. That is using my SBH. Self buffed in defensive mode, my mitigation is still 1281 (32%) and my avoidance is 38.9%. That is using my Guard of Grumm Shield and Moonlight Mace. I need to get a better shield and weapon. I haven't acquired all my new gear for tier 5 yet. I could also boost my avoidance some by getting the appropriate ring and using it. I normally run the offensive stance and either solo or duo with a Mystic. We generally do not take on Heroic mobs but do take on multiple linked mobs. If I was running in a fuller group against heroic mobs and I was the MT, I would run the defensive setup above.

10-05-2005, 06:28 PM	#44
CoLD MeTaL Loremaster Join Date: Apr 2005 Posts: 3,217	I believe the formula you are looking for is over a large enough sample size is : (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage ))) Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed. Message Edited by CoLD MeTaL on 10-05-2005 08:10 AM __________________

10-05-2005, 07:40 PM	#45
Kale Join Date: Jan 2005 Posts: 222	CoLD MeTaL wrote: I believe the formula you are looking for is over a large enough sample size is : (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage ))) Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed. Message Edited by CoLD MeTaL on 10-05-2005 08:10 AM Yep, that's basically the formula. However, the reason to take a large sample size is to negate the skew. You'll get more skewed numbers with a smaller sample size. But Naari brings up the best point of all: Avoidance can actually be unpredictable, due to the hit/no-hit nature of it. You can get pelted for a dozen hits in a row, or not. And, like she (he?) said, it has a limited arc. Mitigation, on the other hand, is rock-solid. You know you're going to mitigate X amount of damage every single hit. As a plate tank, our avoidance is never going to be very high anyway (we're never going to approach that 60% or more range that makes it truly effective). As a plate tank, we sort of have to assume we're going to get hit. That's our job. Our skill is really in mitigating the damage. That's the point of plate armor - to absorb the blow. Right now, if there is a flaw in the game, it is that avoidance and mitigation for a plate tank is even in question. If we were going to model fantasy warfare a bit more accurately, this shouldn't even be a question. We should be opting for plate at every chance, because we should be able to have better mitigation than we could ever have avoidance, due to our training with heavy armor. I think what might be lacking here is that we aren't trained in Heavy Armor. We can wear it (because our skill allows it), but we don't have any additional training with it. So we aren't any better fighters wearing plate than anyone else. What the developers might want to look into is a skill for armor that actually matters. Sure, a priest can wear plate, but should they be as skilled with it as a dedicated fighter, like a Paladin or Shadowknight? Not really. Our skill with plate armor should improve our mitigation, such that choosing to wear lighter armor is actually a bad decision.

10-05-2005, 08:00 PM	#46
Naa Join Date: Nov 2004 Posts: 37	CoLD MeTaL wrote: I believe the formula you are looking for is over a large enough sample size is : (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage ))) Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed. Message Edited by CoLD MeTaL on 10-05-2005 08:10 AM Congrats on needlessly complicating things: What you have written can be simplified as follows: damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation% * 1) damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation%) damage taken = Full damage * (1 - Avoidance% * 1) * (1 - Mitigation%) damage taken = Full damage * (1 - Avoidance%) * (1 - Mitigation%) damage taken / full damage = 1 * (1 - Avoidance%) * (1 - Mitigation%) damage taken / full damage = (1 - Avoidance%) * (1 - Mitigation%) percent damage taken = (100% - Avoidance%) * (100% - Mitigation %) percent damage taken = (100% - Mitigation%) * (100% - Avoidance%) which by the way is identical to % of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100% Now where on earth or the heavens above have I seen that formula before? Of course, if you really like doing the extra and needless calculations just because you like punching buttons on a calculator, have at it :smileyvery-happy:

10-05-2005, 08:17 PM	#47
Yrield General Join Date: Jan 2005 Posts: 267	Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course.

10-05-2005, 08:25 PM	#48
robusticus Loremaster Join Date: Apr 2005 Posts: 220	Who says we're plate tanks? I don't think we are anymore. Warriors are plate tanks. I really wish people would answer questions instead of bashing on one another. Looks like leather chest/legs gives 8-10% more avoidance. My avoidance in offensive stance with a 2 hander is about 14%...that's with full plate and AGI as my lowest attribute. SBS gives me about 12% more. I have noticed less interruptions since I re-equiped SBS (acquired SM finally)... I wonder what the impact of the other slots (head, shoulders, arms, gloves, boots) is to the avoidance penalty and if there is a significant jump between chain and leather. I'm starting to really regret selling my ancient slayer's ring of agility. Plat down the drain if I want it back.

10-05-2005, 08:39 PM	#49
Naa Join Date: Nov 2004 Posts: 37	robusticus wrote: Who says we're plate tanks? I don't think we are anymore. Warriors are plate tanks. I really wish people would answer questions instead of bashing on one another. Looks like leather chest/legs gives 8-10% more avoidance. My avoidance in offensive stance with a 2 hander is about 14%...that's with full plate and AGI as my lowest attribute. SBS gives me about 12% more. I have noticed less interruptions since I re-equiped SBS (acquired SM finally)... I wonder what the impact of the other slots (head, shoulders, arms, gloves, boots) is to the avoidance penalty and if there is a significant jump between chain and leather. I'm starting to really regret selling my ancient slayer's ring of agility. Plat down the drain if I want it back. The leather on chest and legs doesn't give you near that much. My agility stats are reasonably high although I haven't intentionally been buying equipment to boost it. Just checking my character profile it is 93. http://eq2players.station.sony.com/en/pplayer.vm?characterId=146015104 That is prior to any buffing. Part of why it is that high is that I am a ratonga. I may want to get myself an agility ring to boost it even more. (If you aren't aware, you don't need to continue wearing the imbued rings after you buff using them - of course it is a pain swapping them in and out. If I recall correctly, the buff lasts 25 minutes).

10-05-2005, 08:49 PM	#50
Yrield General Join Date: Jan 2005 Posts: 267	robusticus wrote: Who says we're plate tanks? I don't think we are anymore. Warriors are plate tanks. I really wish people would answer questions instead of bashing on one another. Looks like leather chest/legs gives 8-10% more avoidance. My avoidance in offensive stance with a 2 hander is about 14%...that's with full plate and AGI as my lowest attribute. SBS gives me about 12% more. I have noticed less interruptions since I re-equiped SBS (acquired SM finally)... I wonder what the impact of the other slots (head, shoulders, arms, gloves, boots) is to the avoidance penalty and if there is a significant jump between chain and leather. I'm starting to really regret selling my ancient slayer's ring of agility. Plat down the drain if I want it back. +12 agi give 0.5% avoidance 1 pts of defense give 0.3% avoidance dunno about parry

10-05-2005, 08:59 PM	#51
robusticus Loremaster Join Date: Apr 2005 Posts: 220	Try equiping some plate and soloing some groups with that SBH. See how you like the interruption situation then. Who knows, maybe they fixed it stealth-like. My AGI too is naturally high and I kept a COUPLE items that boost it but haven't equipped them in a while. I still have way too much WIS though... I need to swap hex dolls... more plat down the drain. I'm ok for now though because screaming mace is so much better than the BBC...but I have prismatic greatsword ambitions...

10-05-2005, 09:11 PM	#52
CoLD MeTaL Loremaster Join Date: Apr 2005 Posts: 3,217	Naari wrote: CoLD MeTaL wrote: I believe the formula you are looking for is over a large enough sample size is : (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage ))) Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed. Message Edited by CoLD MeTaL on 10-05-200508:10 AM Congrats on needlessly complicating things: What you have written can be simplified as follows: damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation% * 1) damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation%) damage taken = Full damage * (1 - Avoidance% * 1) * (1 - Mitigation%) damage taken = Full damage * (1 - Avoidance%) * (1 - Mitigation%) damage taken / full damage = 1 * (1 - Avoidance%) * (1 - Mitigation%) damage taken / full damage = (1 - Avoidance%) * (1 - Mitigation%) percent damage taken = (100% - Avoidance%) * (100% - Mitigation %) percent damage taken = (100% - Mitigation%) * (100% - Avoidance%) which by the way is identical to % of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100% Now where on earth or the heavens above have I seen that formula before? Of course, if you really like doing the extra and needless calculations just because you like punching buttons on a calculator, have at it :smileyvery-happy: No you are not taking into account that you only mitigate non-avoided damage 22% Avoidance, 35% mitigation, 100000 points over time. Yours: % of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100% (100000 - 35000) * (100000 - 22000) / 1 (65000) * (75000) /1 = 5,070,000,000 Your ORIGINAL formula is wrong, which is why no amount of algebra will help, because you are still taking mitigation against a number that doesn't factor in avoidance. You are also trying to work in percents, which stats should have taught you, averages of averages ARE meaningless. My formula COULD have been simplified to (1-Mitigation%) * (Full Damage - (Avoidance% * Full damage)) Mine: (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage ))) (100000 - ( .22 * 100000)) - (.35 * (100000 - (.22 * 100000))) (100000 - 22000) - (.35 * (100000 - 22000)) 78000 - (.35 * 78000) 78000 - 25740 = 52260 marginally different from your formula. (Off course statistically meaningless because comps don't generate truly random numbers) Avoidance Mitigation Damage Avoidance AvoidedBase MitigationTotal Percent 22% 33% 100000 22000 78000 25740 52260 52.26% 23% 32% 100000 23000 77000 24640 52360 52.36% 24% 31% 100000 24000 76000 23560 52440 52.44% 25% 30% 100000 25000 75000 22500 52500 52.50% 26% 29% 100000 26000 74000 21460 52540 52.54% 27% 28% 100000 27000 73000 20440 52560 52.56% 28% 27% 100000 28000 72000 19440 52560 52.56% 29% 26% 100000 29000 71000 18460 52540 52.54% 30% 25% 100000 30000 70000 17500 52500 52.50% 31% 24% 100000 31000 69000 16560 52440 52.44% 32% 23% 100000 32000 68000 15640 52360 52.36% 33% 22% 100000 33000 67000 14740 52260 52.26% 34% 21% 100000 34000 66000 13860 52140 52.14% 35% 20% 100000 35000 65000 13000 52000 52.00% 36% 19% 100000 36000 64000 12160 51840 51.84% 1) I doubt you will find the majority of game items trade 1 for 1 avoidance for mitigation, haven't really paid that much attention to that. 2) there is a point of diminishing resturns on avoidance, but it isn't a hard 50%, it floats based on your mitigation. 3) Since stats like agility move avoidance without altering mitigation, jewelry, etc. will then change your real place on the curve. __________________

10-05-2005, 09:30 PM	#53
CoLD MeTaL Loremaster Join Date: Apr 2005 Posts: 3,217	Yrieldom wrote: Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course. And it is against a enemy your same level, all con infromation affects these calculations. __________________

10-05-2005, 09:49 PM	#54
Naa Join Date: Nov 2004 Posts: 37	CoLD MeTaL wrote: Naari wrote: CoLD MeTaL wrote: I believe the formula you are looking for is over a large enough sample size is : (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage ))) Given the fact that no computer generates a truly statistically random number, the real value over time will be skewed. Message Edited by CoLD MeTaL on 10-05-200508:10 AM Congrats on needlessly complicating things: What you have written can be simplified as follows: damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation% * 1) damage taken = (Full damage - (Avoidance% * Full damage)) * (1 - Mitigation%) damage taken = Full damage * (1 - Avoidance% * 1) * (1 - Mitigation%) damage taken = Full damage * (1 - Avoidance%) * (1 - Mitigation%) damage taken / full damage = 1 * (1 - Avoidance%) * (1 - Mitigation%) damage taken / full damage = (1 - Avoidance%) * (1 - Mitigation%) percent damage taken = (100% - Avoidance%) * (100% - Mitigation %) percent damage taken = (100% - Mitigation%) * (100% - Avoidance%) which by the way is identical to % of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100% Now where on earth or the heavens above have I seen that formula before? Of course, if you really like doing the extra and needless calculations just because you like punching buttons on a calculator, have at it :smileyvery-happy: No you are not taking into account that you only mitigate non-avoided damage 22% Avoidance, 35% mitigation, 100000 points over time. Yours: % of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100% (100000 - 35000) * (100000 - 22000) / 1 (65000) * (75000) /1 = 5,070,000,000 Your ORIGINAL formula is wrong, which is why no amount of algebra will help, because you are still taking mitigation against a number that doesn't factor in avoidance. You are also trying to work in percents, which stats should have taught you, averages of averages ARE meaningless. My formula COULD have been simplified to (1-Mitigation%) * (Full Damage - (Avoidance% * Full damage)) Mine: (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage ))) (100000 - ( .22 * 100000)) - (.35 * (100000 - (.22 * 100000))) (100000 - 22000) - (.35 * (100000 - 22000)) 78000 - (.35 * 78000) 78000 - 25740 = 52260 marginally different from your formula. (Off course statistically meaningless because comps don't generate truly random numbers) Avoidance Mitigation Damage Avoidance AvoidedBase MitigationTotal Percent 22% 33% 100000 22000 78000 25740 52260 52.26% 23% 32% 100000 23000 77000 24640 52360 52.36% 24% 31% 100000 24000 76000 23560 52440 52.44% 25% 30% 100000 25000 75000 22500 52500 52.50% 26% 29% 100000 26000 74000 21460 52540 52.54% 27% 28% 100000 27000 73000 20440 52560 52.56% 28% 27% 100000 28000 72000 19440 52560 52.56% 29% 26% 100000 29000 71000 18460 52540 52.54% 30% 25% 100000 30000 70000 17500 52500 52.50% 31% 24% 100000 31000 69000 16560 52440 52.44% 32% 23% 100000 32000 68000 15640 52360 52.36% 33% 22% 100000 33000 67000 14740 52260 52.26% 34% 21% 100000 34000 66000 13860 52140 52.14% 35% 20% 100000 35000 65000 13000 52000 52.00% 36% 19% 100000 36000 64000 12160 51840 51.84% 1) I doubt you will find the majority of game items trade 1 for 1 avoidance for mitigation, haven't really paid that much attention to that. 2) there is a point of diminishing resturns on avoidance, but it isn't a hard 50%, it floats based on your mitigation. 3) Since stats like agility move avoidance without altering mitigation, jewelry, etc. will then change your real place on the curve. A) You don't understand the application of my formula. B) You don't correctly apply your formula. Point A demonstrated: % of full Damage = (100% - Mitigation%) * (100% - Avoidance%) / 100% % of full Damage = (100% - 35%) * (100% - 22%) / 100% % of full Damage = 65% * 78% / 100% % of full Damage = 50.7% 50.7% * 100000 points of damage over the course of the battle = 50700 points of damage taken Point B demonstrated: Mine: (Full damage - (Avoidance% * Full damage )) - (Mitigation% * (Full damage - (Avoidance% * Full damage ))) (100000 - (22% * 100000)) - (35% * (100000 - (22% * 100000))) (100000 - (22000)) - (35% * (100000 - (22000))) (78000) - (35% * (78000)) 78000 - (27300) 50,700 Low and Behold, the numbers match. As for your non math points: 1) You are right most items don't trade 1 for 1 on avoidance vs. mitigation. Since that is true, you will need to run the math if you really want to see which is better. In other words, I agree with this point. 2) I never said there was a point of diminishing returns at 50% of avoidance. What I said, incorrectly, was that above 50% avoidance trumped mitigation 1 for 1. I later retracted that and redid my general rule to be correct. See earlier in the thread. I can admit I was wrong and actually did so before anyone pointed it out. 3) Stats on Jewelry (and on the armor too for that matter) do affect your avoidance. And as such, they certainly do affect your point on the curve. For example, they might be able to get your avoidance higher than your mitigation so that in a fight where the avoidance works (in the arc) it is more effective over the long haul than the mitigation. In otherwords, I agree with this point.

10-05-2005, 10:06 PM	#55
CoLD MeTaL Loremaster Join Date: Apr 2005 Posts: 3,217	Naari wrote: LOL I wrote 35%, and used 33% in my calculations. Sorry, thanks for pointing that out. __________________

10-05-2005, 10:22 PM	#56
Naa Join Date: Nov 2004 Posts: 37	CoLD MeTaL wrote: Naari wrote: LOL I wrote 35%, and used 33% in my calculations. Sorry, thanks for pointing that out. And thanks for admitting you were wrong.

10-06-2005, 02:52 AM	#57
lisasdarr Join Date: Dec 2004 Posts: 186	Yrieldom wrote:Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course. 40% avoidance means that over a 'sufficiantly large sample' you will avoid 40% of the hits. If 100 hits is a sufficiantly large sample to counter tha randomness then it doesn mean that you avoid 40 out of 100 hits. __________________

10-06-2005, 07:38 PM	#58
Yrield General Join Date: Jan 2005 Posts: 267	lisasdarren wrote: Yrieldom wrote:Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course. 40% avoidance means that over a 'sufficiantly large sample' you will avoid 40% of the hits. If 100 hits is a sufficiantly large sample to counter tha randomness then it doesn mean that you avoid 40 out of 100 hits. Well no. Because you dont have 40% avoidance, you have 40% CHANCE to avoid a hit every roll you do (chance is the keywork here) Sure you can reach 40% of effective avoidance after 100 hits, then redo the test against the same mob and reach 40% of effective avoidance after 3291 hits. To make it simple you can eleminate the randomness of a "% of chance" because every roll you do is not dependant of your previous roll. Its just like lotto, a basic lotto with 1 chance out of 100 to win, buying 100 tickets doesnt mean you have atleast 1 winning ticket, its just mean that you have 100 more chance to win

10-06-2005, 08:54 PM	#59
Rochir Loremaster Join Date: Mar 2005 Location: Lucan Dlere Posts: 510	Plate armor does in fact reduce our avoidance when equipped. When I craft, I take of my weapon, shield, greaves, curiass and gloves. Without my weapon or shield, with all of my armor on, my avoidance is 23.3%. When I unequip chest, legs and gloves, it goes up to 26.3%. 3 pieces of pristine ebon vangaurd amor cost me 3% avoidance. A full suit of plate armor will cost us about 7-8% avoidance. That is a significant hit. Our armor does in fact reduce our avoidance. The mitigation it adds is not sufficient to compensate for the loss of avoidance as our mitigation is capped (which effectively makes it go down as we level up) for each teir and armor type. Avoidance on the other hand is always a percentage and remains constant as we level up. Whether or not our armor is a liability or a benefit depends on the strength of the mob we are fighting and the type of attacks it is using. Against a sufficiently strong mob, our mitigation becomes ineffective. There are certain attacks that ignore our mitigation. The advantage of avoidance is that it depends only on the level of the mob and works to some extent on all types of attacks and all types of mobs. The big problem with our avoidance is it is frontal only. On the other hand, brawler avoidance is 360 degrees. We must mitigate all damage from behind. It is sad but true that against certain types of mobs in certain situations, we would be better off without any armor on. Another big advantage that brawlers have is that their leather armor gives them the stats and resists they need without significantly reducing their avoidance. Our avoidance is capped lower than theirs so there is no point in our wearing cloth or leather armor to compensate. Our avoidance will never be as high as theirs. No matter how you look at it, brawlers are in fact the uber tanks in eq2 Message Edited by Rochir on 10-06-2005 12:23 PM __________________ Eldatur Aduial, High-Elf (Koada'Dal) Paladin & Armorer Demolition

10-06-2005, 09:25 PM	#60
Kale Join Date: Jan 2005 Posts: 222	Yrieldom wrote: lisasdarren wrote: Yrieldom wrote:Formula doesnt work very well because the % of avoidance show in your Persona screen is versus a single attack roll. if you have 40% avoidance you don't avoid 40 of 100 hits. you have 40% chance to avoid ONE hit. So with 40% avoidance in the the persona screen you can end up fight1 with 22% of effective avoidance and fight2 with 63% of effective avoidance, fight1 and 2 are against the same mob of course. 40% avoidance means that over a 'sufficiantly large sample' you will avoid 40% of the hits. If 100 hits is a sufficiantly large sample to counter tha randomness then it doesn mean that you avoid 40 out of 100 hits. Well no. Because you dont have 40% avoidance, you have 40% CHANCE to avoid a hit every roll you do (chance is the keywork here) Sure you can reach 40% of effective avoidance after 100 hits, then redo the test against the same mob and reach 40% of effective avoidance after 3291 hits. To make it simple you can eleminate the randomness of a "% of chance" because every roll you do is not dependant of your previous roll. Its just like lotto, a basic lotto with 1 chance out of 100 to win, buying 100 tickets doesnt mean you have atleast 1 winning ticket, its just mean that you have 100 more chance to win Some real rocket scientists at work here... It's not anything like the lotto. A statistics class would really help you understand why. A much better comparison is coin-flipping. 50% chance for heads, 50% chance for tails. Those are the actual percentage chances. However, getting exactly 5 heads and 5 tails doesn't always happen. If you take 10 coin flips, there are 10 possibe outcomes (heads/tails): 0/10, 1/9, 2/8, 3/7, 4/6, 5/5, 6/4, 7/3, 8/2, 9/1, 10/0. Those outcomes, in that order, will actually form a bell curve, with the 5/5 being the highest point on the graph, and the 0/10 and 10/0 being the lowest point on a graph. Meaning, if you were to flip 10 coins, you would have a much better chance of getting 4 heads and 6 tails and you would of getting 0 heads and 10 tails. This is where sample size comes into play. Sufficiently large sample sizes will remove the 'streak nature' from the results, and come closer to the real percentage chance of 50/50. It does no good to look at small sample sizes because anything can happen. That's the random nature of things. So instead of looking at a very small sample (after all, it's possible to avoid 10 hits in a row - not likely, but possible) we look at a very large sample size, to get a much more realistic picture of what will happen on an average intance. The middle portion of the bell curve is the average instance. In terms of avoidance - if you're avoidance is 40%, that doesn't mean you are going to avoid 40% of the hits every fight. But what it does mean is that, on the bell curve, you're probably going to be avoiding close to 40% most of the time. You're going to be closer to avoiding 40% than you are 100% or 0%.