Ember Online

Could someone explain how unaffect rates are calculated? I think I understand the variables that go into it (aggressor int, level of staff - not sure if blessings/enchants impact, defender wis) but is the unaffect rate impacted by anything else? It seems like the rate bounces around too wildly because 2nd/3rd standard deviation events happen way too often. For example, I estimate I get unaffects in dreads about 33% of the time over the long run. However, the frequency with which 3/3 or 5/6 unaffects (or 3/3 5/6 hits) occur is to high. If a mob has 33% resists, they should resist 3/3 only ~3% of the time but I feel like I see it 25% of the time.

For defenders its: Race Modifier * Class Modifier * Wisdom * Level * 0.13 * SpellToughnessMultiplier
For attackers its: Race Modifier * Class Modifier * Intelligence * Level * 0.1 * (Random(WeaponMin*0.2, WeaponMax*0.2))

Where weapon can only be a staff to effect pierce rate.
If attacker roll > defender roll then it hits, otherwise resist.
SpellToughness is typically set at 1. Blind, sap, etc normally its 1.15 or 1.1. makes it harder to pierce.

i think thats what you want, i was curious about the same thing last night and looked for it lol

You are using old data and formulas. We talked about it being changed rather recently, and it was.

Defenders = Race Modifier * Class Modifier * Wisdom * Level * 0.125 * SpellToughnessMultiplier
Attackers = (Intelligence * Level * 0.05) + (Random(WeaponMin*0.225, WeaponMax*0.225))) + 30

SpellToughness is typically set at 1. No spells get a nerf to this anymore, left it in in case it does need to be. For smite it is actually 0.8, which means it is easier to penetrate with smite, though this is only on undead.

Race/Class modifier would be for example, 20% would be 1.2, 0% would be 1, etc. So start with a (base of 1 and add Percent/100)

Attackers never had a racial or class modifier from what I know. So worthless to state it. The formula above shaves the unaffect rate around half or little more. Based on 23 int and going down from that, aka you lose less unaffect per INT.

So the only intra-battle variability in unaffect rates is the max/min dmg of the staff?

CGI wrote:So the only variability in unaffect rates is the max/min dmg of the staff?

Yeah.

You roll and the defender will roll. The person with the highest number wins. The roll is:

Min Value = CalculationAbove/6
Max Value = CalculationAbove

so if the Defender was a Human Knight with 20 wisdom at level 25, it would be something like 1.1*1*20*25*0.125=68.75

So there max roll is 68.75, their min roll is 68.75/6= 11.46.

So again, Min roll in this case is 11.46 and max roll is 68.75.

Oh. No wonder its so streaky. That's a lot of volatility.

is that really how it works? because something seems very off on that.

Defenders = Race Modifier * Class Modifier * Wisdom * Level * 0.125 * SpellToughnessMultiplier
Attackers = (Intelligence * Level * 0.05 * (Random(WeaponMin*0.225, WeaponMax*0.225))) + 30

defender
gnome monk
16.4713541667 - 98.828125

gnome mage 23 wis (17.2200520833 - 103.3203125)

10 wis horc fighter
5.2 - 31.25

attacker
gnome mage 23 int/void staff 22.5 ad
29.2578125 - 175.546875 (just input avg on weapon damage)

everything seemed right until i pitted the resist rates versus the hit rates; (range)

-worst MR crit possible (half orc knight 10 wis)

horc rolls of 5.20 -29.24 cant resist
gnome mage 23 int attack rolls 31.26-175.54 cant miss
leaving rolls 29.25-31.25 as a possible resist(which will likely 50/50 )
which is an average of 1 resist out of 167 attacks
disregarding fizzles that's a 99.9941% hit rate for the attacker

-best MR crit possible 23 wis gnome mage
defender rolls 17.22 - 29.24 wont succeed
attacker rolls 103.33 - 175.54 wont fail
leaving rolls 29.25-103.32 (which will likely 50/50)
which is an average of 35 resists per 156 attacks
disregarding fizzles that is a hit rate of 77.57% for the attacker



this seems like insanity, or do i have to factor that at 40% because of "pvp hit rate"

Terron wrote:is that really how it works? because something seems very off on that.

Defenders = Race Modifier * Class Modifier * Wisdom * Level * 0.125 * SpellToughnessMultiplier
Attackers = (Intelligence * Level * 0.05 * (Random(WeaponMin*0.225, WeaponMax*0.225))) + 30

defender
gnome monk
16.4713541667 - 98.828125

gnome mage 23 wis (17.2200520833 - 103.3203125)

10 wis horc fighter
5.2 - 31.25

attacker
gnome mage 23 int/void staff 22.5 ad
29.2578125 - 175.546875 (just input avg on weapon damage)

everything seemed right until i pitted the resist rates versus the hit rates; (range)

-worst MR crit possible (half orc knight 10 wis)

horc rolls of 5.20 -29.24 cant resist
gnome mage 23 int attack rolls 31.26-175.54 cant miss
leaving rolls 29.25-31.25 as a possible resist(which will likely 50/50 )
which is an average of 1 resist out of 167 attacks
disregarding fizzles that's a 99.9941% hit rate for the attacker

-best MR crit possible 23 wis gnome mage
defender rolls 17.22 - 29.24 wont succeed
attacker rolls 103.33 - 175.54 wont fail
leaving rolls 29.25-103.32 (which will likely 50/50)
which is an average of 35 resists per 156 attacks
disregarding fizzles that is a hit rate of 77.57% for the attacker



this seems like insanity, or do i have to factor that at 40% because of "pvp hit rate"

Sorry weapon values are added, not multiplied. Otherwise no weapon would destroy it.

With 23 int and a oad staff it is 64.375 as max.

This would be min/max with a gnome sorc
(25*23*0.05) + (21*0.225) + 30
(25*23*0.05) + (25*0.225) + 30
Weapon includes enchants.

But in max cases, it is 64.375 there, so 64.375/6 and 64.375 for min/max



Half orc knight with 12 wis = 37.5.... VS a gnome int 23 you have 22.25% MR rate.
Half orc knight with 10 wis = 31.25 .... VS a gnome int 23 you have 15.45% MR rate.
Assassin HO with 10 wis = 28.9 .. VS gnome int 23 you have 11.41% MR rate.
Goblin Knight with 16 wis = 59.5.. VS gnome int 23 you have 44.89% mr rate.
Human Knight with 20 wis = 68.75.. Vs gnome int 23 you have 53.53% mr rate.
Human Knight with 20 wis = 68.75.. Vs gnome int 23 you have 53.53% mr rate.
Elf Monk with 20 wis = 82.75.. Vs gnome int 23 you have 64.64% MR rate
Gnome Sorc with 22 wis = 98.72.. vs gnome int 23 you have 71.45% MR rate.


Edit: Updated formula above.

Most classes see a noticeable improvement as they improve their gear, but it's kind of impossible for spellcasters to gain any advantage using a top-end staff vs. storebought staff. For example:

Level 25, 23 int w/ decked void staff average max roll would be 63.81.
Level 24, 23 int w/ decked void staff (if could equip) average max roll would be 62.66.
Level 25, 22 int w/ decked void staff average max roll would be 62.56
Level 25, 23 int w/ decked store bought staff average max roll would be 62.35.

So going from a store bought staff to the best staff currently i/g is a delta of ~1.5 on your max roll, compared to ~1.2 when level 24 vs. 25 or losing an int point. I know the comparison isn't going to be equal, but going from store bought to OAD weapon on a melee class has such a huge impact vs. gaining a level or a str point.

I don't have exact #s but I think these are in the ballpark. Say the delta between store bought vs. OAD weapon for melee is 6 AD and you add 5-10 damage per AD, you're adding 30-60 damage per attack. On a caster, if you're adding 1.5 points to your max roll, maybe that adds a 1% boost to your unaffect rate? That seems incredibly meager in comparison. Even in an extreme case, 23int gnome vs. 10wis assassin you cut that 11.41% unaffect rate to 10.41% which is a ~10% boost in expected damage or 15 damage vs. the 30-60 damage for melees. On the other end, 23int gnome vs. 20wis elf monk you cut unaffect from 64% to 63% which is like adding 2 points of expected damage.

CGI wrote:Most classes see a noticeable improvement as they improve their gear, but it's kind of impossible for spellcasters to gain any advantage using a top-end staff vs. storebought staff. For example:

Level 25, 23 int w/ decked void staff average max roll would be 63.81.
Level 24, 23 int w/ decked void staff (if could equip) average max roll would be 62.66.
Level 25, 22 int w/ decked void staff average max roll would be 62.56
Level 25, 23 int w/ decked store bought staff average max roll would be 62.35.

So going from a store bought staff to the best staff currently i/g is a delta of ~1.5 on your max roll, compared to ~1.2 when level 24 vs. 25 or losing an int point. I know the comparison isn't going to be equal, but going from store bought to OAD weapon on a melee class has such a huge impact vs. gaining a level or a str point.

I don't have exact #s but I think these are in the ballpark. Say the delta between store bought vs. OAD weapon for melee is 6 AD and you add 5-10 damage per AD, you're adding 30-60 damage per attack. On a caster, if you're adding 1.5 points to your max roll, maybe that adds a 1% boost to your unaffect rate? That seems incredibly meager in comparison. Even in an extreme case, 23int gnome vs. 10wis assassin you cut that 11.41% unaffect rate to 10.41% which is a ~10% boost in expected damage or 15 damage vs. the 30-60 damage for melees. On the other end, 23int gnome vs. 20wis elf monk you cut unaffect from 64% to 63% which is like adding 2 points of expected damage.

That is because the damage is based on INT and INT only. the staff is just a bonus though but its about 5% from having a weapon to not. Weapons will not and should not give a major bonus to unaffect. They never will. The damage that sorcs do is on the high rather than low. You can't really base melee exactly with spell damage. Most of a sorc is based on it's stats and weapon does not effect damage in this regard.

I'm not sure if you are claiming sorcs are weaker in general, but lately alot of people been screaming that sorcs destroy even with moderate wisdom. The whole unaffect rate needs a looking at though in general.

I don't want to make weapons a base for spell casters. It's a little bonus, that's it.