For the love of god, please, rework DEXTERITY. (Updated)

"

FaceLicker wrote:

You seem to be disregarding how Damage Reduction from Armor actually works. The character sheet gives only an ESTIMATION and is by no means absolute. Here is how Damage Reduction works:

"

Malice wrote:

Armour / Damage Reduction
Damage Reduction reduces physical damage taken. Elemental damage and damage-over-time are not affected. The amount of damage reduction depends on the defender's armour total, and the attacker's attack damage:

reduction = armour / (armour + 20*damage)

The amount of reduction is capped cannot be more than 75%.

The fact that damage reduction scales with the amount of damage means it is difficult to know exactly how much damage is being reduced, however the maximum effect seems to be capped at around 4-5% of total armour rating.
An easy to remember rule of thumb is that to achieve 50% damage reduction, you will need an armour rating equal to twenty times that of the damage being dealt. For example, to achieve 50% damage reduction against a 100 damage hit, you'll need 2000 armour.

Monsterss in Chaos do a lot more than 100 Damage. Let's say they ONLY do twice that amount, 200 damage. That's quite a conservative number but let's go with it. That means you need a total of 4000 Armor just reduce that damage by 50% or 100 Damage. If the monsters did 400 damage, you would need a whopping 8000 Armor just to only take half of that. But wait.. 8000 armor, you say? Haha, my character sheet says I have 75% Damage Reduction. Your calculations are False!

You seem to think just because your character sheet might say 50% with 2k Armor at lvl 60, that you're reducing ALL physical Damage by half. That is simply not true and if you read how the mechanics actually work, you'd know that.

Not only does Damage Reduction not work the way you think it does, but you also have to take into consideration that Armor does not protect against anything BUT physical damage. Evasion, on the other hand, can protect against EVERYTHING in its own way. Because it protects against everything, you need much more Evasion to get the same %Evade as %Damage Reduction. The same reasoning can generally be seen in local and global mods. Local mods only affect the item they are on and so larger values don't affect the whole as much. Where as global mods affect everything and are generally seen as much smaller values. Armor protects you from only one source and that's why the number is larger. Evasion can protect from all, sure it's a gamble, but that's the nature of the stat.

Ok first up I welcome your post, its the first constructive and worthwhile addition anyone whos opposed this thread had made imo.

Ill try to address your post in a somewhat ordered manner to avoid confusion.

I never suggested that the numbers on the character sheet are an absolute figure that translates into the real game. Im not the most knowledgeable man about POE but I know that much.

You are right of course that when it says XX% damage reduction that this is based on a somewhat unrealistic set of numbers which are unlikely to occur while actually playing.

The exact same thing can be said of evasion though. Yeah so damage reduction will rarely function as well as it suggests, but neither will evasion. Just as damage reduction is reduced or ignored so is evasion. Most higher level creatures have very high accuracy ratings, this directly modifies (in a negative way) your evasion rating.

Thats why im comparing the character sheet number. Cause by and large they give good grounds for comparison. Im not suggesting they directly translate in game, but they are both (DR and Evasion) based on a similar standard equation. So if its true that DR is going to be lower in real terms, its also true that evasion will itself be lower.

Add to this that DR is easier to attain and you start to see where the inbalance is. A character with 500 Armour has a higher DR than a dex based character has evasion with the same numbers.

They are both negatively affected by all sorts of situational things in the real game, and in some rare circumstances Im sure evasion is more useful for very brief periods. But currently Armour is just way superior. All evidence points to this as fact... the fact the raw numbers point to it, and the fact that nearly all high level rangers either wear armour pieces OR take iron reflexes. If evasion wasn't broken you'd expect to see alot more rangers wearing full evade gear.

Where as in contrast how often do you see a marauder wearing evasion gear?? I appreciate this is by no means conclusive proof but its just another nail in the coffin. The only people who dont think evasion is broken whom actually use it at high levels have supplemented it with either hybrid armour builds, IR or massive block percentages.

Armour users dont need to do any of that. Also compounding the issue currently is that the predominant damage type in the game is physical. Even elite monsters which do extra elemental damage still do a larger percentage amount of physical.

There are very few creatures of any note that do elemental damage. Now im aware this may change as the beta progresses and if so evasion may start to come into its own a bit more, but currently its just not very good, on any level. The situation for its only benefit occurs so rarely its not worth the agro to bother with it. Take end of act 2, the one area of the game with a fair amount of elemental damage flying about.

Thats fine except mixed with those are physical damage archers whose projectiles are faster, are more accurate and who occur in greater numbers. Add to that the fact a blind man could dodge the elemental ranged skeletons and again I wonder where the benefit of evasion is?

Finally to clarify, im not moaning here. Im not whining or wanting rangers to be some uber tank class, I fully support them being glass cannons (of sorts).

But I dont support the fact that evasion as a pure armour class is not viable unless you support it with a secondary line of defense. Not unless they give the other two types of armour similar drawbacks.

EDIT: Just checked the ladder. In the top 10 there is only 1 ranger. He is using iron reflexes with max frenzy charges, and in a thread hes talking in even he is saying his survivability is very low, and thats with iron reflexes! In the top 20 of the ladder there are a total of 4 rangers and im willing to bet that at least 75% of those either use hybrid armour builds or Iron Reflexes. Where as im fairly sure not many witches use anything except energy shield and not many marauders use anything except damage reduction.

Those numbers (just 4 in the top 20) tell their own story imho. As its not like rangers are a traditionally unpopular class, quite the contrary they are usually one of the most popular.

"

and party size (damage scales with party size, accuracy doesn't).

Unless it was changed back again, it shouldn't be the case.

Edit: Also mark, you forgot that armour isn't doing anything against attacks that deal elemental damage while evasion do.

"

Mark_GGG wrote:

So while both "chance to evade" and "estimated physical damage reduction" are both estimates based on an expected average for monsters of your level, the estimation of chance to evade is reasonably accurate, where the estimation of damage reduction is likely to be wildly off for some hits. When we changed to the current damage reduction mechanics, we stopped estimating physical damage reduction on the character screen because we didn't feel it was accurate enough to bother doing - it's back now not because it's more accurate, but because people apparently prefer even a very inaccurate estimation to none at all (and because we put in some work to make an estimate of average damage for monster level). We never considered taking out the chance to evade estimation, as it never had this problem.

If the estimation for of the average damage for monster level is accurate then the average damage reduction that is shown is also accurate. Just have to make sure people know that it is an average.

"

Mark_GGG wrote:

The average monster damage estimation is accurate for a standard hit from a default attack of a monster of that level. I don't know if it accounts for critical strikes, I'm certain it doesn't account for any of the things I mentioned in 4) in my post.

Yeah.

Would it be possible to get the average damage from recording data from actual gameplay or simulations?

"

Sickness wrote:

"

Mark_GGG wrote:

So while both "chance to evade" and "estimated physical damage reduction" are both estimates based on an expected average for monsters of your level, the estimation of chance to evade is reasonably accurate, where the estimation of damage reduction is likely to be wildly off for some hits. When we changed to the current damage reduction mechanics, we stopped estimating physical damage reduction on the character screen because we didn't feel it was accurate enough to bother doing - it's back now not because it's more accurate, but because people apparently prefer even a very inaccurate estimation to none at all (and because we put in some work to make an estimate of average damage for monster level). We never considered taking out the chance to evade estimation, as it never had this problem.

If the estimation for of the average damage for monster level is accurate then the average damage reduction that is shown is also accurate. Just have to make sure people know that it is an average.

I'm not sure how meaningful an average can be here.

Average the damage of all the monster types and you'd have a number too low for blues, yellows, and monsters under aura effects.

Taking a weighted average to account for the appearance of blues and yellows might not even be feasible since their proportions change from area to area (see .9.6d patch notes).

Related question: how exactly are blue and yellow mobs enhanced (not counting the special mods they might have)? Do they always have more hp? deal more damage? hit with higher accuracy?

"

Sickness wrote:

"

Mark_GGG wrote:

The average monster damage estimation is accurate for a standard hit from a default attack of a monster of that level. I don't know if it accounts for critical strikes, I'm certain it doesn't account for any of the things I mentioned in 4) in my post.

Yeah.

Would it be possible to get the average damage from recording data from actual gameplay or simulations?

You can do it yourself if you have the patience:

Watch how much damage you take from a hit, check your armor rating (not reduction %), work backwards with the armor mitigation formula.

I've only done this a few times w/ my marauder to check the worst cases for armor, but my numbers were very rough estimates due to massive hp regen and I never bothered writing them down.