Khumak
04-24-2003, 12:20 PM
This is the best explanation I have seen thus far on the actual effect of defensive skill mods. Maybe something similar has already been posted here but I haven't seen it. Anyway here it is, shamefully stolen from the PoN boards:
http://pub148.ezboard.com/fpaladinsofnorrathfrm13.showMessage?topicID=2082.t opic
The effect on 'mod' items is on the skill number, not the actual rate at which the skill works. For example if parry is (parry_skill + level / constant) (and the consensus was that skills such as DW and DA did indeed use that format) then it is clear to see that a +10% parry would not equate to a 10% increase in parrying of attacks.
Let's assume that is the case for now. Take a 65 paladin with 200 default parry. Let's say that equates to 5% of blows being parried. I believe Graal had previously posted fairly convincing evidence that d/p/r checks are made on incoming attacks when the attack isn't from the rear BEFORE 'hit/miss' checks are made. Furthermore that its a straight check of your skill and that the mob is irrelevent (same d/p/r rate on Orc Pawn as on Uber_01). I dont not recall if it has ever been shown that each of the skills is checked in any particular turn. As Caladel mentioned an increase in any avoidance skill makes a small but non-zero reduction in the benefits of other avoidance skills since there is a smaller percentage of attacks passing through said skill to be later avoided by the remaining skills.
We will take the simplest model first, that being no consideration given to order of avoidance operation. This is not the most accurate case (which I may present later if time permits), but simply a starting point to familiarize oneself with some concepts.
Let's take his numbers for d/p/r for an example, slightly rounded for simplicity.
Skill, Percent, Skill Number, Skill Number + 65
D, 3.75%, 155, 220
P, 5.00%, 205, 270
R, 3.75%, 200, 265
One then achieves these as the constants for the skill check (from rate = (skill+65)/Constant)
D = 5867
P = 5300
R = 7200
(There is no need to obtain the actual value of these constants; I merely did so for those more familiar with actual numbers in their calculations. They are easily kept as merely Cd, Cp, and Cr.)
We will use the 'miss' rate as 37.75%. The 'hit' rate is 100% - the sum of misses,d,p,r which equals 50.25% in this case.
Now, lets say we have a 10% mod item. Then we have the following rates (obtained from trunc((default_skill)*(mod+1))/Constant)
D = 4.01%
P = 5.38%
R = 4.03%
It is logical to assume that attacks that are d/p/red would be 'missed' the same rate as attacks that get through d/p/r skill checks. The total attacks that were d/p/r represented 12.5% of mob attacks made, misses represented 37.75%, and hits represented 50.25% of OVERALL attacks, but hits represented 57.1% of attacks that made it past the d/p/r checks. Let's just take Parry for now. Remember that Parry is checked before the hit/miss calc is made, so 42.9% of the newly parried blows would have missed anyway so instead of an increase of 38 parries per 10,000 mob attacks the actual increase in attacks avoided is 38*(0.571) = 21.69 newly avoided attacks per 10,000 mob attacks made.
So, with the default skill you are being hit 5025 times, with the 10% parry mod you are being hit 5003.3 times per 10k mob attacks. In other terms you are taking 99.568% as much damage with the skill mod in comparison to without it. Now, to put some perspective to that amount you have to factor in your relevent hp figure. Whatever it is you will now 'tank' (take frontal melee damage) as if you had 0.432% more HP than before. So, a 10% Parry mod is 'worth' (note that this is in terms of 'net' hp, not raw hp that are then modified by ND/PE/PD).
To a 7k hp paladin = 30.23 hp
To a 8k hp paladin = 34.54 hp
To a 9k hp paladin = 38.86 hp
(with all of nd/pe/pd these would be equivalent to 25.94, 29.65, and 33.36 raw HP as displayed on items)
In actuality you be somewhat better off as completely avoiding an attack brings benefits such as not being stunned that the mere possession of those extra HPs does not (going to avoid a long discussion as to why more raw hp are more or less valuable than their hp-equivalent avoidance).
Hope that has been helpful,
Kether
Arete - E'ci
http://pub148.ezboard.com/fpaladinsofnorrathfrm13.showMessage?topicID=2082.t opic
The effect on 'mod' items is on the skill number, not the actual rate at which the skill works. For example if parry is (parry_skill + level / constant) (and the consensus was that skills such as DW and DA did indeed use that format) then it is clear to see that a +10% parry would not equate to a 10% increase in parrying of attacks.
Let's assume that is the case for now. Take a 65 paladin with 200 default parry. Let's say that equates to 5% of blows being parried. I believe Graal had previously posted fairly convincing evidence that d/p/r checks are made on incoming attacks when the attack isn't from the rear BEFORE 'hit/miss' checks are made. Furthermore that its a straight check of your skill and that the mob is irrelevent (same d/p/r rate on Orc Pawn as on Uber_01). I dont not recall if it has ever been shown that each of the skills is checked in any particular turn. As Caladel mentioned an increase in any avoidance skill makes a small but non-zero reduction in the benefits of other avoidance skills since there is a smaller percentage of attacks passing through said skill to be later avoided by the remaining skills.
We will take the simplest model first, that being no consideration given to order of avoidance operation. This is not the most accurate case (which I may present later if time permits), but simply a starting point to familiarize oneself with some concepts.
Let's take his numbers for d/p/r for an example, slightly rounded for simplicity.
Skill, Percent, Skill Number, Skill Number + 65
D, 3.75%, 155, 220
P, 5.00%, 205, 270
R, 3.75%, 200, 265
One then achieves these as the constants for the skill check (from rate = (skill+65)/Constant)
D = 5867
P = 5300
R = 7200
(There is no need to obtain the actual value of these constants; I merely did so for those more familiar with actual numbers in their calculations. They are easily kept as merely Cd, Cp, and Cr.)
We will use the 'miss' rate as 37.75%. The 'hit' rate is 100% - the sum of misses,d,p,r which equals 50.25% in this case.
Now, lets say we have a 10% mod item. Then we have the following rates (obtained from trunc((default_skill)*(mod+1))/Constant)
D = 4.01%
P = 5.38%
R = 4.03%
It is logical to assume that attacks that are d/p/red would be 'missed' the same rate as attacks that get through d/p/r skill checks. The total attacks that were d/p/r represented 12.5% of mob attacks made, misses represented 37.75%, and hits represented 50.25% of OVERALL attacks, but hits represented 57.1% of attacks that made it past the d/p/r checks. Let's just take Parry for now. Remember that Parry is checked before the hit/miss calc is made, so 42.9% of the newly parried blows would have missed anyway so instead of an increase of 38 parries per 10,000 mob attacks the actual increase in attacks avoided is 38*(0.571) = 21.69 newly avoided attacks per 10,000 mob attacks made.
So, with the default skill you are being hit 5025 times, with the 10% parry mod you are being hit 5003.3 times per 10k mob attacks. In other terms you are taking 99.568% as much damage with the skill mod in comparison to without it. Now, to put some perspective to that amount you have to factor in your relevent hp figure. Whatever it is you will now 'tank' (take frontal melee damage) as if you had 0.432% more HP than before. So, a 10% Parry mod is 'worth' (note that this is in terms of 'net' hp, not raw hp that are then modified by ND/PE/PD).
To a 7k hp paladin = 30.23 hp
To a 8k hp paladin = 34.54 hp
To a 9k hp paladin = 38.86 hp
(with all of nd/pe/pd these would be equivalent to 25.94, 29.65, and 33.36 raw HP as displayed on items)
In actuality you be somewhat better off as completely avoiding an attack brings benefits such as not being stunned that the mere possession of those extra HPs does not (going to avoid a long discussion as to why more raw hp are more or less valuable than their hp-equivalent avoidance).
Hope that has been helpful,
Kether
Arete - E'ci