Dexterity, We Have a Problem.
|
Two Ideas:
Add Physical Resistance and Reduce Armor Rating. So a Marauder, which had 75% Physic. Damage reduce from Armor before now will have about 50% Physic Damage Reduce from Armor + 50% from Resistence. U can get the Physical Res. from Items, maybe from Passives. (ok this is a little copy from Diablo 3) Second Idea: Give Ranger items Armor. For example: Marauder normal lvl 60 Armor: 1000 Armor needs 100 Str Duelist normal lvl 60 Armor: 666 Armor and 333 Evasion, needs 50str + 50Dex Ranger normal lvl 60 Armor: 333 Armor and 666 Evasion, needs 100Dex. IGN: kReiZy Last edited by sYkoDe4d#0481 on Jun 21, 2012, 12:32:48 PM
|
|
|
I didn't see this post at first so I made a threadnought somewhere else. Unfortunately it got locked, but I'll link it again here.
Sorry if it sounds detached from the conversation. I simply hadn't read this thread before, so here I analyze this problem from the beginning. If you wish you can go straight to the solutions. Anyway, here goes: We're far into the closed beta, and I know it's wishful thinking to expect GGG to change major gameplay mechanics at this point. But it's not impossible. I think this is the only moment I can try and express those concerns. --------What's the point of this thread? Dexterity is a flawed stat, from its very core. It's flawed in its mechanics, its flawed in its psychological impact on players, and it's flawed on its scaling. GGG are doing their best to balance it by tweaking it numerically, but that will never overcome the major design issues that it implies. Dexterity needs a major overhaul if it's ever to be a viable stat, and more so if it wishes to be viable for hybridization. -------------------------------------------------------------------------------- --------What are those problems? -------------------------------------------------------------------------------- a) Accuracy and miss chance are bad mechanics.
Spoiler
These mechanics please nobody, neither on the giving nor the receiving end.
First, accuracy. Building to minimize a problem that keeps growing through the game (miss chance) has a negative psychological effect. It's working to compensate a failure, not to improve. Aside from the problem with investment, the mere act of missing is frustrating. Missing is okay in games without manual aiming, but in games like this it's just absurd. If you shoot a perfect arrow that lands square on an enemy, it should hit, period, not miss because of a numerical roll. Tweaking the numerical impact on dexterity over hit chance is not a solution. If it's overdone, non-dex builds will suffer from very low hit chances and the huge frustration it involves. If it's kept low, dedicated DEX builds will gain little offensive advantage. b) Evasion is a bad mechanics.
Spoiler
Evasion is not a good mechanic either, as presented right now. Its main issue is that it has terrible synergy with the character concepts Dexterity is supposed to enforce (low HP glass cannons). Evasion is random by nature, which implies the possibility of huge damage bursts. This would be okay if the DEX characters had a respectable HP buffer, but ironically they're worse in that aspect than INT chars themselves, which thanks to their ES enjoy a sizeable buffer.
Second, Evasion is bad because it doesn't protect against spells, and this is a bigger issue than what it looks like at first. It is bad because it makes DEX the only stat that provides no defense at all against spellcasting. INT gives an ES buffer, STR an HP buffer, but DEX leaves the character completely unable to resist magical assaults in any shape or form (aside from magical resistances which are universal and not being taken into account for this reasoning). Third, Evasion DEX bonus is percentual over armor. This is a huge deal. It means in order for DEX to shine as a defensive stat, full evasion gear should be worn, which kills any attempts to hybridize. Impact of these issues:
Spoiler
The conclusion of Accuracy and Evasion being bad is bad news for DEX. First, it's pointless to invest in DEX exclusively (after a certain point, the accuracy bonus is negligible, and as was pointed before it leaves you completely open to burst damage and magic), but even worse, DEX is a terrible hybrid attribute.
Let me make a difference here between "hybrid character" and "hybrid attribute building". I know very well there are successful characters that take nodes from the DEX area as well as from INT and STR. But be aware that they are there for specific nodes and effects, not for the stats themselves. There is no point in building DEX (as a stat) alongside STR or INT either, for two reasons: INT/DEX: Accuracy doesn't affect spellcasting. DEX is entirely useless offensively if you wish to use spells. Evasion is percentual over armor values so it's wasted on the hybrid / int gear. STR/DEX: Accuracy might mean a marginal increase in DPS, but substantially lower than STR bonuses if some accuracy nodes are respected. Evasion is wasted, as, yet again, it's percentual over armor and the base values of hybrid/STR armor are from bad to nonexistant. -------------------------------------------------------------------------------- --------Proposed solutions -------------------------------------------------------------------------------- There is an alternate interpretation of Dexterity and evasion mechanics that would solve this problem and make DEX a viable stat, both offensively, defensively, and in terms of stat hybridization. Without further ado: Accuracy dissapears. DEX provides AS and CS
Spoiler
Accuracy disappears as a stat, for good. There is no such a thing as miss chance anymore. Attacks that land are attacks that hit, and the only way to "avoid" an attack in the game becomes positioning and manual evasion. Accuracy is replaced by attack and cast speed as the offensive side of the DEX stat. 10 DEX gives +2% Attack Speed / Cast Speed. This makes it interesting for all builds and brings it in line with the other stats in terms of offensive usefulness. Also, this creates a psychological positive reinforcement for building Dexterity. This makes Dexterity a valid offensive option for dedicated building and hybridization, bringing it in line with the other stats. To recap: -STR Is a good offensive choice because it increases physical damage without a cap. -INT is a good offensive choice because it increases mana/Mana regen, which makes you able to use more powerful spells more often, without a cap. -DEX is a good offensive choice because it lets you attack or cast faster, without a cap (unlike now, where it is limited to attacks, not spell, and is affected by a cap). Any combination of these is also useful. INT/DEX spellcasters are now possible, since DEX gives spells a positive effect, unlike the current iteration of accuracy. Evasion overhaul. Reduction of both spells and attacks biased towards bust prevention
Spoiler
Evasion is radically reworked. Now, evasion provides damage reduction, similar to armour, applied to both attacks and spells. Unlike armour, the formula for reduction has more effect on attacks that take higher percentage of HP, rather than small attacks. On average, it's about 50% as effective as armour. In layman's terms, if you're being hit by several small attacks, you'll notice evasion less. However, if there is something heavy swinging slowly at you, the reduction will be much higher. So you're probably wondering... Doesn't that make Armour obsolete? This new evasion is like Armour on steroids! It protects against attacks AND magic, and it also handles burst damage! Not quite. To begin with, even though the formula is skewed towards protecting against strong bursts of damage (which is consistent with DEX characters), it's worse on average, and should bring less damage reduction than what a dedicated Armour guy would. In exchange for that, it's directly improved by the stat DEX (which armor isn't). Why this difference? Easy: as stated before, DEX is the only stat that DOESN'T provide a "HP-like" buffer. Therefore, it is fair that DEX/evasion is the only stat that can improve your resistance against BOTH kinds of damage, from attacks and spells. As you can see now, the defensive part of DEX (evasion) now becomes a very valid mechanic that has interesting synergy with both strength and intelligence. Again, let's make a comparative: STR: Defends against physical and spell damage, by providing a HP buffer. Armour scaling is biased towards protecting against small hits, and its weak to burst. INT: Defends against physical and spell damage, by providing an ES buffer. It is weak to Chaos. DEX: Defends against physical and spell damage, by reducing both. It is weak to multiple small hits (unlike now, where DEX only protects against physical strikes and does it in a neglectable way). Again, you'll notice hybridization is worthwhile. STR/DEX stat hybrids using their corresponding armor is useful as the complementary scaling formulas fit each other nicely protecting against all the damage spectrum. INT/DEX is also correct because evasion helps mitigate the squishiness while INT backs it up with a buffer. Last, pure DEX is viable. -------------------------------------------------------------------------------- --------Conclusions --------------------------------------------------------------------------------
Spoiler
As you can see in the explanations of the two changes, all stat combinations and builds become viable. But I think the most important conclusion of my DEX overhaul is that the defensive aspect of each character becomes not only viable but loyal to what the concepts behind the stats represent in a roleplaying perspective.
As you can see, a DEX character is naturally prepared to deal with slow, hard hitting opponents, which feels about right with the idea we have of a nimble rogue, rather than the current "get one shot when you commit a mistake" situation. STR, heavy armoured characters, make use of the current scaling formula (which could be exaggerated even more), ignoring small hits that bounce against their armor. Instead, they're weak to powerful, armor-crushing blows. INT specialists are the all rounders, defending against both types of damage as long as they have space to breathe, and as long as no special anti-shield measures are used against them (chaos). The cool thing about my change is that it deepens even more the rock-paper-scissors orientation of PoE's upcoming PvP. Now there is the possibility to build characters with the objective to exploit other's defenses. For instance, a slow hitting evasion character (a Ranger with a slow longbow) would be ideal to bring down a heavy armored, slow character, as it would be easy to avoid the damage while dealing a lot of it. Similarly, a heavy armored dagger wielding duelist would be the perfect Ranger hunter, and many other combinations arise. Anyway, regardless of what improvements this suggestion would bring, I think it's important to put our sight on how flawed DEX is and how badly it needs attention. Again, I know we're already far in the design process, but GGG is still in time to fix this. RIOT did a similar thing while taking away their own evasion and RNG based abilities. I did put a lot of effort in this thread, so please read it carefully if you wish to write a rebuttal. Otherwise, I'll be looking forward to your opinion and arguments. See you! |
|
|
(reposted from the other thread)
I think this is really poorly thought out. Why should all attacks hit? Shouldn't you become more 'accurate' as you level up? Why should the same amount of sword swings or bow shots at level 1 land as when I am level 50? Why should I NOT be able to dodge attacks more with more experience (evasion). My lvl 49 duelist has 380 dex currently; by your math of 10 dex = 2% AS / CS, my increased speed JUST from dexterity would be 76%. That's absurd. I WOULD like to see very high dexterity characters get a bonus to avoiding critical hits, however, as it 'makes sense' and helps the glass cannon builds from getting one shotted by powerful critters. I do like some of the OP's ideas on page one, however. |
|
|
I still like my suggestion at the bottom of the last page, but I doubt it is possible at this point of development, because it would push the relase date back I guess.
I just wonder if and how GGG will approach this problem, because without an overhaul likes this I can only imagine cheap solutions and workarounds. Last edited by golem09#3277 on Jun 27, 2012, 4:37:59 PM
|
|
|
Well, the thread got unlocked and moved, so it's duplicated now. Sorry for the hassle.
|
|
|
I love Accuracy and miss chances Falcord, and most of the people do so, and they all are a must for a arpg too ! I found your thoughts pretty much self-dislikes with very insufficient reasons behind it.
Edit: I was forgetting... Evasion should make us have percentile block chance against spells, if this is ever get implemented, sure evasion would be worthwhile, now it is pretty fragile and useless imo. "This is too good for you, very powerful ! You want - You take" Last edited by BrecMadak#3812 on Jun 27, 2012, 7:40:47 PM
|
|
|
I think Falcord is mostly right in his (or her?) analysis of the Dex problem, but I do not believe we should get rid of accuracy.
Why not just keep the bonus to accuracy from dex and add something like every 10 dex gives +1% (or .5% ...or lower depending on how the balance works out) attack speed and/or maybe some similarly small-valued bonus to crit chance? That would help make Dex useful for offense capabilities on par with Str or Int. As for defense, I think the dead horse of Evasion not being optimally implemented has been sufficiently beaten: we want something that makes Dex valuable for defense but we don't want it to just become armor or ES. Falcord's defense solution is one that bears greater scrutiny. It might be something to try before release (maybe even during open beta) if the devs can do it. But, there may be a simpler solution (or at least the solution seems simpler to me): right now we have Str-based defense in the form of armor that encourages one to charge in and tank damage (except for maybe in upper difficulties) and Int-based def in the form of ES which provides a buffer for damage but requires time to recharge. Armor reflects the physical endurance of its wearer and ES is literally a shield of mental energy. In order to be similarly fitting for Dex, any Dex-based defense should rely on movement and speed. Right now Evasion is just a percentage chance to receive no damage from an attack, which, since that percentage can never get to 100%, leaves the Dex-based defender with a bit of a problem. I propose, then, that we allow the percentage chance to reach 100%...sort of. We could keep the same evasion values on armor and everything, but instead of giving one flat evasion rate, we allow evasion rating to be determine on the basis of a sequence of hits, rather than on a flat per-hit basis, with the probabilities changing as more enemies hit in a succession. I will try to explain more thoroughly: first suppose you were a dex-based character with evasion rating x fighting a monster. The monster attacks, with the usual numerical checks going in to play to see if the monster hits. But, here we have a numerical system such that evasion can go up to 100% (actually, it can go higher, which I shall explain momentarily), so if the monster is sufficiently inaccurate or you sufficiently evasive, it might be completely impossible for the monster to hit you...on the first strike. When you successfully evade an attack, x (your evasion rating) would be diminished, rendering you more likely to be hit by subsequent attacks. The amount by which x is diminished would be determined by some function of the accuracy and damage of the attacker. So, if the monster strikes again, the computer checks its accuracy against your reduced evasion, now x-j. x-j may still be sufficient to give you 100% evasion chance against this enemy, in which case they would miss again, reducing your evasion to x-2j...and so on until eventually the monster scores a hit. After the monster reduces your evasion to the point that your evasion chance against it is not 100%, evasion works exactly as it does now in that you have some non-zero probability of getting hit. This mechanic would reflect the fact that many different attacks from different angles or a series of attacks are more difficult to dodge. I would propose that we keep the minimum of 5% dodge chance. Also, the block chance would come into effect after evasion is checked so they would not become redundant. Dex should also have some small bonus to block chance (like the same values I gave above for attack speed bonuses or even lower). In this way we give Dex-based characters a way to stay viable in melee combat, provided that they keep in motion and do not get locked down and surrounded. Also, this would play into the notion that slow but heavy hitters should be less dangerous than fast movers and fast attackers. A fast mover/attacker can attak many times in rapid succession, making it more likely to reduce your evasion to levels at which you can be hit even by slower more ponderous enemies, but a group of slow enemies could also pose a serious threat if they manage to envelope the player. Now, as for the evasion value itself, unlike ES I think it would be best if there were no delay in regenerating evasion. That way, while a Dex character may need to disengage occasionally, they can move into and out of combat more fluidly. The drawback of evasion vs ES, on the flip side, would be the ability of ES to mitigate spell damage. Obviously the numbers would need to be crunched and tested in such a way that this would be both viable and balanced. Ideally pure Dex defenders would be able to spec themselves such that they could evade a few hits from any enemies they may face. Passives that grant bonus to evade chance would easily be reworked into this mechanism, as would the one that puts all of evasion into armor. There could even be passives that increase the rate at which you regenerate evasive capacity, or some nodes that offer a tradeoff between your max evasion and the regen rate. It would also be good if there were some curses or abilities that lowered enemy accuracy ratings (I think a smoke bomb or some such device for a rogue, shadow, or duelist would be quite useful). In this way we could make evasion useful to hybrids as well. Let me know what you guys think. Tl,dr: Not sure how to make this short, but here goes: Make evasion a resource kind of like ES but (very) different that gets used when you evade attacks. You may be able to 100% dodge an enemy, but the more attacks that are aimed at you the more likely you get hit. |
|
|
Falcord has the right idea. Remove accuracy from players and evasion from mobs. Use Dex instead for Attack Speed and Cast Speed. I would say .15% per point of dex would be perfect.
His Evasion rework is less important IMO, but well thought out. The point is the Dex pretty much sucks as a stat. |
|
|
I'm glad to see that this thread has become so large. While the ideas that have been presented are vast and varying, the overall message that I believe has been delivered is exactly the reason that I made this thread to begin with.
Rather than push my own ideas, I wanted to make it apparent to GGG that the Dex stat could use some work. If none of the ideas within this thread are used, I am fine. At least the creative minds at GGG have been alerted to the problems that the community sees with Dexterity. Please feel free to use this thread to offer more suggestions on Dex or to comment on existing ones! :) Last edited by YoMicky#3367 on Jul 1, 2012, 1:06:57 AM
|
|
|
Another idea for making dex a little bit more useful and attractive is to give a run speed bonus per 1o dex points.
So, if you have around 3oo dexterity points, you actual run speed would be increased by around, let's say 20% (0.67% increase per 1o dex) And please don't remove accuracy, I am compeletely with BrecMadak on that point. My little tribute to Diablo 1 aka why Diablo 3 is the worst part of the series.
http://www.youtube.com/watch?v=zP2ejhudUlU ig: Witchfire_The_Unholy |
|
