The defensive stats are actually pretty balanced

Ok...

I'm going to address the faults in the OP one point at a time.

"

NotRegret a écrit :

-Health: Universal defense. Bad against multiple targets good against burst.

Being hit by multiple targets IS burst. I completely do not see how health is "bad for multiple targets".

What's good for multiple targets, if not health?

"

-Health regen/life gain on hit: Universal defense. Better than max hp for sustained damage worst for burst. Optimized for tanky builds

Yes. Agreeable and accurate position on healing efficiency. As long as your health CAPACITY is enough to survive attacks, it then becomes more important to begin working on your health recovery. Anyone who plays video games smartly knows this.

"

-Life leech: Similar to health regen but more optimized for glass cannon builds

But then you go ahead and say this?

It's not SIMILAR to health regen. It's exactly the same. I understand you're saying it's more optimized for glass cannons because they deal more damage thus leech much more - but it's faulty to think this way. A glass cannon won't have the necessary health capacity for health restoration to be worthwhile. The reality of it is, even life leech is more optimal for tanky builds. The difference between life leech and raw health regen is very simple and disappointing: You take whichever one gives you more restoration.

Example... This ring has 3.2 health regen, but this other ring has 2% life leech. Well, as a level 70 Marauder, you take the 3% life leech because that will get you much more health return.

However, now you can choose between a 4% life leech gem, or a 1% heal per second Vitality gem. You're about to embark on a map where the enemies have increased armor, dodge, and natural physical reduction, plus you're cursed with Enfeeble. Clearly, your life leech won't be returning much life. Better to go with the regen.

However, NONE of this is optimal for glass cannons. This game as a whole has nothing that makes it worth going glass cannon.

"

-ES: Specialized defense for kiting hard countered by chaos. CI arguably makes it the best defensive stat but is balanced by a need for very expensive gear

If you're stacking only ES with no Chaos resist or health, obviously you're going CI. If not, then you DO have health and/or Chaos resist. So that first clause was pointless.

Beyond that, it definitely is not "arguably the best defensive stat, balanced by being expensive". In fact, that's spoken from a true player who's never played as CI.

We get Shocked from a level 1 Spark, we get chain Frozen by a solo white caster mob and killed through 9k ES, and we get flat-out framelocked at anything whose attack makes physical contact with our character. We are forced into using Eye of Chayula as our amulet to avoid the stunlocks, and forced into Wanderlust, Dream Frags, or that new Alpha helm to avoid being Frozen (and we STILL get perma-Chilled). All these uniques hold back some of our late-game scaling, whether in defenses like resists, or in offensive capabilities. We don't even use Tricorne to immune Shocks, because we can't AFFORD to have another unique in use. (Uniques are inherently weak items only used for a specific effect.)

In order to gain the ES regen, we have to avoid damage for a few seconds. If little tiny things are poking us for 12 damage, that still resets our regen timer. We have to clear a fight entirely. And when our gear generally lacks evasion and armor because it's ES-stacked, this generally means we have to jump into a fight with Quadra Buffs blazing - Granites, Blood Lust, Arctic Armour, maybe even Molten Shell... and god knows what else. And then when it gets a little dangerous, we have to bail, and then come back in later with our buffs blazing once again.

Health-stackers can endure to stay in a fight a bit longer, because their gear sets will naturally have loads of other stats like armor & whatnot. Plus they can use life flasks and life gain on hit effects. (ES users can also use life flasks if they get Zealot's Oath, but that doesn't quite qualify to be a standard-issue keystone because of its location.)

"

-Armor: Scales better than hp vs physical but worthless vs spells. Best defense against swarms of weak enemies, worthless vs burst or 1 strong enemy because it becomes less effective the higher the damage

I don't even know what you're arguing anymore. Of course armor doesn't effect magic damage. That's what resist does.

And health is for those big bone-crushing hits. I thought you were trying to prove that health isn't as overpowered as people are saying, now you're backing up the notion.

"

-Evasion: scales better than hp vs physical but worthless vs spells. Bad against swarms of weak enemies you end up being stun locked. Great against burst since it doesnt scale backwards like armor.

ACTUALLY...

Evasion is great against swarms of enemies, and ARGUABLY less effective against single large burst.

Against swarms, it allows the "large numbers" rule to take effect, letting your 30% evasion ACTUALLY evade 30% of the incoming damage more accurately to its odds.

Against single burst, who knows? Maybe you'll dodge the 1-shot. Maybe you won't. But the odds are 70:30 against you, and you simply die if you don't get lucky.

Oh-... Unless you stacked health, of course.

You also don't get framelocked if you stacked health. Read up on stun mechanics.

"

-Block chance: A physical reduction stat that scales better than armor/evasion but you sacrifice dps for the abiltiy to block

Doesn't this support the notion that defensive stats are overpowered more than it does your attempted point that they're not? In fact, when are you going to get to the offfff..........






-- You know what? I just understood the point of your thread. It's not comparing defensive stats to offensive stats. It's comparing defensive stats to OTHER defensive stats.

rofl...

I don't really see the point in doing this at all. Everyone knows you're supposed to mix armor and evasion. After that, everyone knows life is more effective than Energy Shield, if for NO other reason because of stun and status ailment mechanics. (And even after that reason, life simply has a lot more support than ES.)


Why did I think you were comparing defensive to offensive stats? It's not even like this game HAS offensive stats in the first place...

"

FacetiousTomato a écrit :

I just want to put in my 2 cents that NONE of these stats protect you from physical burst. I don't understand how evasion is "great against burst" why by definition evasion means
0 damage
0 damage
0 damage
5k hit

Burst protection what?

It's good, but you need enough of a base hit point pool to survive a few hits, so that you can react against multiple high-accuracy monsters before they reduce you to zero. In your example, as long as you can survive the one hit, you can heal in between each hit with flasks and the like.

"

NotRegret a écrit :

There are 2 types of damage

Sustained (done by swarms of weak monsters and dots)
Burst (done by crits and single strong monsters)

They come in 3 flavors

Physical Elemental and Chaos



A good defensive build will have answers to all of them. Relaying on just Evasion means you will die to sustained

Interesting way to put it. I've been recently thinking of the defense types in terms of Buffer (life, ES), Mitigation (most things) and Sustain (leech, gain on hit, gain on kill, regen, flasks).

Thanks ScrotieMcB for the nice graphs, even though I think you left quite a few relevant factors out of the calculation (Blind, Enfeeble, Armor Penetration effect). It gets really annoying to compute, however, some trends are insightful I think.

PoE defenses build on top of Life or ES. That's how it is and you can't build around it. Armor, Block, Evasion and Endurance charges (read: mitigation strategies) are supplemental and you always have to have enough buffer to take the one big hit. Armor doesn't protect you from the one big crit and evasion doesn't save you from the one blow that will eventually penetrate your layers of avoidance.

All those strategies work. I've played armor-heavy and evasion-acrobatics and while they produce completely different arch nemesi and playstyles they both work (with loads of hp of course). The questions are:

- Should the (by design) secondary defenses actually be usable as a primary defense?
- How strong should they be as a secondary defense compared to HP in terms of investment in nodes and/or EHP calculations?
- Are the different challenges in the game balanced regarding the different secondary mitigation strategies?

The HP-whining will lead nowhere, it's anchored deep within the game systems and thus it's bound to stay. There might and probably will be adjustments but HP/ES are the center of the world of defense - something has to be.

"

FacetiousTomato a écrit :

"

NotRegret a écrit :

-Armor: Scales better than hp vs physical but worthless vs spells. Best defense against swarms of weak enemies, worthless vs burst or 1 strong enemy because it becomes less effective the higher the damage
-Evasion: scales better than hp vs physical but worthless vs spells. Bad against swarms of weak enemies you end up being stun locked. Great against burst since it doesnt scale backwards like armor.
-Block chance: A physical reduction stat that scales better than armor/evasion but you sacrifice dps for the abiltiy to block

I just want to put in my 2 cents that NONE of these stats protect you from physical burst. I don't understand how evasion is "great against burst" why by definition evasion means
0 damage
0 damage
0 damage
5k hit

Burst protection what?

There is no hard counter to physical burst in this game. A hard counter would be game breaking and remove any chance of dying.

Evasion is the best counter to physical burst simply because the other options do nothing.
Armor will only reduce a big crit hit by 10%
Regeneration/life steal doesnt stop burst at all
Resist doesnt work on physical
Health prevents you from being 1 shot but wont stop the damage.
A 30-40% chance to dodge the big crit is the best option you have.

Evasion, kiting, and blocking are the only way to avoid face tanking a 2k damage crit

"

Khazrad a écrit :

Thanks ScrotieMcB for the nice graphs, even though I think you left quite a few relevant factors out of the calculation (Blind, Enfeeble, Armor Penetration effect). It gets really annoying to compute, however, some trends are insightful I think.

PoE defenses build on top of Life or ES. That's how it is and you can't build around it. Armor, Block, Evasion and Endurance charges (read: mitigation strategies) are supplemental and you always have to have enough buffer to take the one big hit. Armor doesn't protect you from the one big crit and evasion doesn't save you from the one blow that will eventually penetrate your layers of avoidance.

All those strategies work. I've played armor-heavy and evasion-acrobatics and while they produce completely different arch nemesi and playstyles they both work (with loads of hp of course). The questions are:

- Should the (by design) secondary defenses actually be usable as a primary defense?
- How strong should they be as a secondary defense compared to HP in terms of investment in nodes and/or EHP calculations?
- Are the different challenges in the game balanced regarding the different secondary mitigation strategies?

The HP-whining will lead nowhere, it's anchored deep within the game systems and thus it's bound to stay. There might and probably will be adjustments but HP/ES are the center of the world of defense - something has to be.

By design diablo clones have it so you can only truely die from burst.
You basically have unlimited sustained thanks to never-ending hp pots. Even if your pots do run dry it just costs 1 tp scroll to fill them back up. The only way you will die from sustained damage is if you are completely under geared/leveled.

That leaves burst as the main way that players die (and thus are challenged). If the developers were to give us a hard counter to burst (like say removing the dimensioning returns on armor) than players would never be challenged and never die.

Health is the natural counter to burst and works on every flavor of damage.

As long as players have infinite potions the major threat will come from burst, as long as the major threat comes from burst, health being the #1 burst counter will be a required stat.

"

Khazrad a écrit :

Thanks ScrotieMcB for the nice graphs, even though I think you left quite a few relevant factors out of the calculation (Blind, Enfeeble, Armor Penetration effect). It gets really annoying to compute, however, some trends are insightful I think.

• Blind is straight-up "enemies hit you with attacks 25% as often." It doesn't play with the evasion formula; if someone has zero evasion rating and blinds some monsters, they will hit 23.75% of the time (it does, however, use evasion entropy). Just don't take Unwavering Stance, and it's 4x EHP regardless of your other defenses.
• Enfeeble would benefit armour far more than evasion, due to how the armour formula works. Maybe should have included a graph for it. I think it's safe to say, though, that if you're going Enfeeble + endurance charges, you'll cap out very quickly, and you'd want to start speccing evasion after you cap.

"

Khazrad a écrit :

The questions are:

- Should the (by design) secondary defenses actually be usable as a primary defense?
- How strong should they be as a secondary defense compared to HP in terms of investment in nodes and/or EHP calculations?
- Are the different challenges in the game balanced regarding the different secondary mitigation strategies?

• No, because it would make the differences between defenses too extreme. A pure armour character would get faceroll everything physical but find non-physical opponents almost impossible to overcome, and evasion characters would survive only as long as their luck held out (except with the entropy system, it wouldn't). Life should be the backbone of EHP, while defenses should be EHP modifiers that make the character much better at certain tasks -- that is, at preventing the dangers the build is likely to expose itself to while doing its thing, whether it's melee or ranged.
• Defenses should have more of an impact than they do now, but without going to the extremes of pure defense builds. The only way specialized defenses will be the optimal choice (meaning: actually get used) is if the EHP gain in that specific category (such as vs physical damage for armour) is large enough to justify not getting a life node's worth of increase in all types of EHP; beating the life node by only a small amount (10% vs 8%) isn't enough to compensate for the degree of specialization. As it so happens, life nodes are competing very heavily with everything, including DPS nodes, which is why I have this long thread in Suggestions titled "Nerf life nodes... while boosting base life so life itself isn't nerfed." The point of the suggestion isn't to remove the life backbone from the game; like you said, it's not going anywhere. I just want it to only increase by 4%, not 8%, on the passive tree, while starting out with more life in the first place.
• In terms of defense balance, the key thing is whether the defense is tailored towards the way that class does their job. As long as most physical damage is from melee monsters, armour is in a good spot, because ranged shouldn't be strong against melee, they should rely on movement and their range to avoid those confrontations. Energy shield is in a good spot, because casters should be able to withstand some burst DPS, but not sustained, and ES recovery cooldown is a mechanic that hates sustained damage; Ghost Reaver kind of throws this out the window, but luckily casters aren't good at leeching, so it mostly is a non-caster keystone, while Zealot's Oath is specialized to the non-CI type of ES user. However, both give ES some sustainability, which is kind of bothersome; ES is definitely the strongest defense right now and by far the easiest to build around.
  
  Evasion should be a good defense against incoming ranged attacks, again relying on movement and range to nullify melee opponents. However, the number of "archers" (ranged non-caster) doesn't seem that large compared to the number of ranged casters, which render evasion useless. Although it would be a major redesign, it would probably be more fair if evasion also worked on spells (even non-damaging ones -- imagine how cool it would be to dodge a curse), which I feel wouldn't make ES and evasion feel too similar at all, even if they did approximately the same job (ES hates little hits that prevent recharging, evasion hates big hits). You'd have to take auto-hit out of spellcasting, redesign accuracy so that each attribute gives accuracy for its fun tools, bump spell damage numbers up slightly to compensate for chance to miss, then give Dexterity something to replace the accuracy that you just spread out among the three attributes. A lot of work. Probably will never happen, so I guess I'll just accept that evasion will be the third wheel. (Note that buffing what evasion does now is potentially dangerous by taking away from armour's identity, since most physical damage is done by attacks and evasion currently works only on attacks.)

If they were balanced you wouldnt see every build either going CI or full health nodes

Show me a viable HC build that focuses on armor more then other defenses

Show me a viable build at all that would rather take any other node if there was 1 point left and they take armor, block or evasion over health.



They need to improve evasion

They need to improve armor

They need to improve block

They could fix them all with tiny adjustments.



Create nodes so that armor and evasions can do more to block spells and abilities.

Block animation is a diminishing return so that there is no stun lock.

Problem solved, and yes there is a problem.


Until I see a build in which a marauder picks more armor nodes then health or a templar eschewing health for block then the game is not balanced.

"

ScrotieMcB a écrit :

"

Khazrad a écrit :

Thanks ScrotieMcB for the nice graphs, even though I think you left quite a few relevant factors out of the calculation (Blind, Enfeeble, Armor Penetration effect). It gets really annoying to compute, however, some trends are insightful I think.

• Blind is straight-up "enemies hit you with attacks 25% as often." It doesn't play with the evasion formula; if someone has zero evasion rating and blinds some monsters, they will hit 23.75% of the time (it does, however, use evasion entropy). Just don't take Unwavering Stance, and it's 4x EHP regardless of your other defenses.
• Enfeeble would benefit armour far more than evasion, due to how the armour formula works. Maybe should have included a graph for it. I think it's safe to say, though, that if you're going Enfeeble + endurance charges, you'll cap out very quickly, and you'd want to start speccing evasion after you cap.

The graphs were indeed nice, thank you.

I had a question about whether lowered accuracy from Enfeeble helps high evasion build more than low evasion build but I was able to figure it out from the formulas included with the graph. At 10k mark, evasion had 72% of the EHP of evasion+armour. This went up to 74% after 20% lowered accuracy but down to 65% after 25% reduced damage per hit. See below about standard monster damage value being too small compared to practice though.

Unlike Blind and Enfeeble, is Arrow Dodging a true evasion build tool then? I thought doubling your chance to evade would have increasing returns with evasion but either I don't understand how to add it to the formula or it ends up increasing armor+evasion EHP by same multiplier.

Overall, your first graph was informative but did not make me conclude that pure evasion is undesirable. The premise included a monster with standard damage for lvl 68 but I find those character sheet amounts laughably small in practice. Danger situations come from added phys/substantial/elemental damage mods and auras, magical/rare/unique monsters (that I believe get inherent bonus life and damage), reflect and hard hitting moves like shield charge. Higher damage-per-hit lowers the value of armor, shifting related graphs down while keeping their shape right? Not sure if including crits would actually change the shape.

Players would also consider elemental attack damage and physical spells (counting shield charge here due to autohit) in their build choices of course. Shield Charge is a huge worry for me regarding the viability of pure evasion, it should be at least dodgeable if not evadeable as well. Same for Rain of Arrows I suppose.

The second graph is more worrying for game balance. Endurance charges are just too amazing together with their armor synergy. Yet the hassle they cause makes people often pretend they don't even exist. The increasing returns don't look healthy though.

EDIT:
My best effort to compare arrow dodging at 10k defenses mark.

evasion EHP multiplier: 1/(1-(1-632/(632+2500^0.8))*2) = 10.58
armor+evasion: (1/(1-(1-632/(632+1250^0.8))*2)/(1-5000/(12*582+5000)) = 4.82
Pure evasion is 123% of armor+evasion against projectile attacks compared to 72% before arrow dodging.

Problem: I don't see how the evasion formula in the graph pic reached its current form. Or I do see that 1/(1-(1-632/(632+(x/4)^0.8))) = (632+(x/4)^0.8)/632 but the arrow dodging version is not as easy to convert into that form and the .95 multiplier you also had (which I assume represents the minimum evade chance) seems dubious. The evade chance is always being reduced by 5%?

The evasion EHP formula assumes that 5% chance to evade is the baseline, since a character with zero evasion still have 5% chance to evade (unless Unwavering Stance). Thus, 10% chance to get evade is a 5.5% increase, not a 11% increase.

However, I forgot about Arrow Dodging.

Any combination of Blind and Arrow Dodging is 95% to evade projectile attacks, even on zero evasion. Unless you take Unwavering Stance -- the drawback is stronger than we thought.

I don't have time to make a graph now, but I anticipate that Arrow Dodging without blind will look a little like the Armour graph with 7 endurance charges; it essentially makes it much more reasonable for evasion to reach a 95% evasion breakpoint.

Given the strength of Arrow Dodging, perhaps giving evasion the ability to evade spells wouldn't be a good idea after all -- evasion has the tools to almost completely nullify projectile attacks, and granting the same power to spells would be OP.

Arrow Dodging is really, really good. More narrow, but much stronger than Acrobatics, with no downside.