Hear me out... We keep talking about glancing this and bring back evasion DR and whatever bell curve maths and such, but I wanted to add my suggestion that I think would not be frustrating to play against and would remove the need for gouging stats.
Instead of making hit and evade a direct comparison, we make them entirely independent.
There is no more evade value. Only evade DR and hit chance.
The attacker rolls to hit based on a bellcurve formula with his own hit. Skills and buffs can modify this chance. For instance, an evade buff or passive would mean attackers get reduced hit. This means you want accuracy even when fighting tanks, which makes sense. In actual combat, even heavily armored combatants will try to not get hit, but they will be far less effective at it.
Example:
50 hit: 10%
120 hit: 50%
170 hit: 75%
220 hit: 90%
etc. I dunno how the math would look like. But basically hitting high amounts of hit means good chance to hit but never guaranteed.
When an attack "misses", it applies evasion DR.
Your CEL determines how much damage reduction you get from evasion proccing.
10 CEL: 11%
50 CEL: 55%
so you can reach fairly good DR value with your CEL. Skills might increase this
Effects when an attack "misses"
- Defender always take some damage, reduced by evasion DR. So there is no "full evade"
- Skills do activate status resistance checks, but defender gains +1% bonus status resistance per evasion DR
- Weapon on-hit and on-crit effects don't proc
- Most skill effects that are guaranteed still activate, like fleur(?).
Evade from your gear and certain skills reduces enemy's hit value! Making your CEL more valuable. So evade builds would be centered around trying to have better chances of proccing evasion and increasing DR.
So for instance, maybe a guy has 200 hit, so his hit chance is 85%. The enemy has gear (25), a passive (20) and one buff (35) that reduce hit by 80, so he has 50% chance to hit instead of 85%. And his DR is 60% from raw CEL, so the evasion build is getting an averaged 30% damage reduction from evasion with, let's say, a medium amount of investment. Meanwhile, a tank build with no CEL invested would get about 1.5% DR from evasion in that scenario.
Balancing this part would be quite delicate but if the hit chance formula is solid and the ways to modify hit feel fair, the rest will fall into place.
Of course this is just an example but the idea would be that the common evasion build would be able to reach an evasion chance that gives them good DR without needing to focus their whole build around it, allowing them to get other kinds of defenses as well like good hp and DR. You just invest however much you want into each and it gives proportional results.
It would also put an element of ever-present RNG to common combat which would make plays less "guaranteed" even in tank matchups.
Skill would still be an important stat, but you wouldn't need to reach a specific amount or compete with gouged numbers on the other side. So overall people would be able to put stats a bit more spread out and focus on what they want. If they want high hit, they can invest more and there's skills that will help with that, but it would never completely invalidate evasion.
Evasion ignore abilities would still exist, but since evade no longer requires going all in, it would be more like a nice way to counter evaders instead of a way to completely destroy their build.
You would probably see many builds with something like 50 cel, 30 def, 30 res, and a bit of evade buffing, getting decent mileage out of hybriding. Or super dodge+tank builds who are very tough but aren't untouchable (still need SR versus DoTs...). Or maybe a tank gets like 30 CEL and a little bit of evasion passives to complement his tankiness.
Overall, it would take work to fine tune the math and balance, but if gotten right, I think it would be a great way to future-proof the balance of evade and hit and reduce the unfairness of certain matchups. No more feeling like your build is completely useless, whether that's because you can't hit a guy or you can't dodge a guy!
Of course, everything in this post is up for debate and ultimately it is not meant to be a suggestion to be implemented as is but something to offer alternatives and a new way to look at the evasion problem. I'd love to hear your feedback and discuss it!
Instead of making hit and evade a direct comparison, we make them entirely independent.
There is no more evade value. Only evade DR and hit chance.
The attacker rolls to hit based on a bellcurve formula with his own hit. Skills and buffs can modify this chance. For instance, an evade buff or passive would mean attackers get reduced hit. This means you want accuracy even when fighting tanks, which makes sense. In actual combat, even heavily armored combatants will try to not get hit, but they will be far less effective at it.
Example:
50 hit: 10%
120 hit: 50%
170 hit: 75%
220 hit: 90%
etc. I dunno how the math would look like. But basically hitting high amounts of hit means good chance to hit but never guaranteed.
When an attack "misses", it applies evasion DR.
Your CEL determines how much damage reduction you get from evasion proccing.
10 CEL: 11%
50 CEL: 55%
so you can reach fairly good DR value with your CEL. Skills might increase this
Effects when an attack "misses"
- Defender always take some damage, reduced by evasion DR. So there is no "full evade"
- Skills do activate status resistance checks, but defender gains +1% bonus status resistance per evasion DR
- Weapon on-hit and on-crit effects don't proc
- Most skill effects that are guaranteed still activate, like fleur(?).
Evade from your gear and certain skills reduces enemy's hit value! Making your CEL more valuable. So evade builds would be centered around trying to have better chances of proccing evasion and increasing DR.
So for instance, maybe a guy has 200 hit, so his hit chance is 85%. The enemy has gear (25), a passive (20) and one buff (35) that reduce hit by 80, so he has 50% chance to hit instead of 85%. And his DR is 60% from raw CEL, so the evasion build is getting an averaged 30% damage reduction from evasion with, let's say, a medium amount of investment. Meanwhile, a tank build with no CEL invested would get about 1.5% DR from evasion in that scenario.
Balancing this part would be quite delicate but if the hit chance formula is solid and the ways to modify hit feel fair, the rest will fall into place.
Of course this is just an example but the idea would be that the common evasion build would be able to reach an evasion chance that gives them good DR without needing to focus their whole build around it, allowing them to get other kinds of defenses as well like good hp and DR. You just invest however much you want into each and it gives proportional results.
It would also put an element of ever-present RNG to common combat which would make plays less "guaranteed" even in tank matchups.
Skill would still be an important stat, but you wouldn't need to reach a specific amount or compete with gouged numbers on the other side. So overall people would be able to put stats a bit more spread out and focus on what they want. If they want high hit, they can invest more and there's skills that will help with that, but it would never completely invalidate evasion.
Evasion ignore abilities would still exist, but since evade no longer requires going all in, it would be more like a nice way to counter evaders instead of a way to completely destroy their build.
You would probably see many builds with something like 50 cel, 30 def, 30 res, and a bit of evade buffing, getting decent mileage out of hybriding. Or super dodge+tank builds who are very tough but aren't untouchable (still need SR versus DoTs...). Or maybe a tank gets like 30 CEL and a little bit of evasion passives to complement his tankiness.
Overall, it would take work to fine tune the math and balance, but if gotten right, I think it would be a great way to future-proof the balance of evade and hit and reduce the unfairness of certain matchups. No more feeling like your build is completely useless, whether that's because you can't hit a guy or you can't dodge a guy!
Of course, everything in this post is up for debate and ultimately it is not meant to be a suggestion to be implemented as is but something to offer alternatives and a new way to look at the evasion problem. I'd love to hear your feedback and discuss it!
![[-]](https://neus-projects.net/forums/images/collapse.png)
