MaxQuest

Hmm, what if it would be possible to balance such a system, without post-release band-aids?Also, yes, JS is aware of the potential 17/1 problem. So at least he is looking for a solution. Maybe we could propose something as well: Afaik it was confirmed that class starting abilities will scale.I was explaining the need in such scalling here. Plus here is the confirmation from Josh: link Good point. Possible counter-point could be: if in PoE1 we had viable and optimal builds, in PoE2 we could end up with optimal and very-optimal)As long as there is possibility of squeezing more, I am happy) Opportunity cost.Having a lvl 17 monk and hitting level-up screen, should bring thoughts about what else you could take instead. I.e. the benefits of taking: - 1st level in barbarian - 1st level in cipher - 18th level in monk should be approximatively the same; and just like in PoE1 depend mostly on our playstyle preference and character's niche in the group; without a clear winner for all possible party compositions. That saddened me the most about that game.There is a single most-optimal build. 2-3 possible contenders. Aaand, that's all. Plus there is no decent hard-cc. 1 lvl dip in barbarian for carnage1 lvl dip in chanter for chants 1 lvl dip in monk for wounds 1 lvl dip in cipher for 1st rank powers + dmg increase via whip 1 lvl dip in rogue for sneak attack So far so good. As all above options are wanted. I.e. there is an opportunity cost. Now there: - will be added scaling to make pure class options also optimal (under certain conditions) or at least viable. - and little bit of tweaking for stuff that gets too above or too low under the power curve. Lets take for example carnage. We can immediately spot potential problems due to several synergies: - cipher with poe1_carnage could potentially overflow with focus - rogue with poe1_carnage could potentially transform in AoE killing machine, especially if also meets condition for Deathblows - wizard with poe1_carnage could cause over-abundance of hits with Citzal Lance, and leading to enemies being insta-melted via combusting-wounds + blasts. - plus - poe1_carnage limits design-space for weapons with on-hit/on-crit effects especially if they are chance bassed. (otherwise there would be no reason to nerf The Unlabored Blade proc rate) As you can see there are 2 main problems: - big damage increase - hit count increase Keeping in mind that there will be scaling, both problems could be mitigated in the following fashion: V1. - scaling damage: carnage dmg coefficient ranges either from -0.90 to -0.34 (additively); or from 10% to 66% multiplicatively; depending on class or power level - tweaking a): carnage hits do not proc on-hit/on-crit effects AND carnage does no longer trigger citzal's blast-like aoe - tweaking b): same as above, but also do not generate focus (although this would make barb useless for a cipher) note: scaling damage won't be a problem because low-level carnage will still deal somewhat decent damage, with DR being substituted by Penetration system. V2. - instead of scaling damage, we could scale carnage proc rate: i.e. on each swing barbarian has a chance to deal cleave damage to a ll enemies around the target. and this chance would scale with power level from x% to 100% (linearly or not) - tweaking: not needed; as hit rate is already reduced. Tbh I like V2 more, as it keeps the current spirit of barbarian (a character who is able to apply weapon effects to many enemies simultaneously). Plus it doesn't kill Combusting wizard/barb from the straight-go. And even if you would take 7/11 wizard/barb for Citzal lance + carnage, it would not be such an issue anymore, because: - 1. carnage would not have a 100% proc rate, rather closer to 70%. So much less hits. (specifically down from 125 to ~95 vs 5 targets; versus 50 hits while dual-wielding) - 2. your wizard spells would deal damage based on power_level 6; not 9; Maybe cc + self-buffs would also have shorter duration; - Citzal's lance is a summoned weapon, it's damage could scale as well - 3. your barbarian abilities would deal damage based on power_level 6; not 9 - 4. your general dps output would be somewhat equal to a pure wizard who just spams spells. And potentially even less vs a single target. - 5. Heart of Fury has an AoE of 2m; and it does NOT scale with INT. Most of the time you are very unlikely to hit even 5 enemies with it. ------------- Now, be it carnage scales according to V1, or be it according to V2. If you are a lvl 1 fighter, you could: - take 2nd level in fighter, for a fighter talent. Let's say Weapon Mastery which would add +0.1 to dmg coefficient - take 1st level in barbarian, for carnage (at 10% chance or 10% dmg). Which would roughly increase your dmg vs single target by 0%; vs 2 targets by 10%; vs 3 targets by 20%, and so on. - take 1st level in rogue; for +0.15 dmg coefficient from sneak attack - take 1st level in cipher; for +0.1 dmg coefficient from soul whip - and so on. But these do scale. Imagine the same fighter at lvl 17. On next level-up he could: - take 18th level in fighter, for a lategame talent/specialization - take 1st level in barbarian, for carnage (at 30% chance or 30% dmg). - take 1st level in rogue; for +0.45 dmg coefficient from sneak attack - take 1st level in cipher; for +0.3 dmg coefficient from soul whip If that lategame fighter's talent provides a benefit of +0.35 dmg coefficient magnitude (or it highly increases the base damage of some of fighter's abilities) than it might be worth it to go pure fighter. The idea is to make that choice to not be a no-brainer. All we need is to go through all 55 combinations, consider all the op combinations, and find the common golden mean for scaling (plus tweak those that end up too weak or powerful) P.S. Khm, looking at that wall, it ended up longer than expected.

It indeed takes a lot of effort to tweak and balance. And JS acknowledges it. Hmm, the trade-off is: (debatable ofc, but) - power-builders delight, as being able to squeeze more out of your character, and plan at which levels to start multi-classing and which order, is also sort of a pleasure) - JS personal preference for trying something new, in an attempt to push the genre further. And if the system turns out to be great, be the one who invented it. That would be Witcher 2

Knew about the drug. But didn't know that Draining Whip was giving focus-per-hit instead of focus-per-damage. Well yeah, it would be hard to balance quick-switching blunderbuss, especially if you also skip recovery. That would result x6 focus gain from the talent compared to dual-wielding daggers at zero-recovery. (for first 3-4 shots ofc) Tbh, I like that cipher's focus gain scales with dps. You optimize your dps. You get more focus. You deal even more damage via damaging powers; or cc more often and your party deals even more damage. Which spirals further. The thing was that it was not that easy to figure out proper weapons/armor/stat-spread/party-buffs-debuffs-composition for the maximum dps, especially when you also had to take DR into account. With penetration system this seems to get quite simplified through.

Agreed, it's not simple. If I was to balance such a system, I would probably start from the end. One needs to know exactly what he wants to achieve first (could be some axiomatic sentences) and then work back from there, adding variables, and arriving to formulas for power_levels, damage coefficients and spell uses amount. "Axioms" or rather starting points could be our expectations: - 1. we want multi-class characters to be picked (55 possibilities) - 2. we want a pure class characters to be picked too (11 possibilities) - 3. we want the picking to be based on rule of cool or personal preference, rather than how powerful they are. - 4. we want to take into account that a 19/0 character at level up; has to decide between 20th level (and respective goodies) in class x, or 1st level in class y. Both options should be approximately equally desirable. Because if one is clearly better than other - it kills diversity, and partially the point in having such a big amount of class combinations. - 5. we want to take into account that 1/0 character at level up; has to decide between 2nd level (and respective goodies) in class x, or 1st level in class y. And in a balanced environment both options should be approximately equally desirable. These could be rewritten further: - 1. a level max/0 class x character should not be stronger in [utility + damage_output] than the most optimal combination of x/y. - 2. a level max/0 class x character should be somewhat equally strong in [utility + damage_output] as other optimal combinations of x/y. - 3. all remaining x/y combinations should at least be viable. But max/0 and and a few of x/y combinations should be optimal. - 4. X20 ~= Y1 when making a choice at lvl 19->20 - 5. X2 ~= Y1 when making a choice at lvl 1->2 From 4 and 5, it is clear that stuff added by dipping in a 2nd class, should scale. Be it with character level, class level or both. This scaling could linear or not; It could start at higher value but have lower increase with level; or vice-versa. So there are at least two axis to balance dipping. I understand that this is a simplistic approach. But the point is to rewrite and rewrite until you arrive from expectations to mathematical model. In any case, now designer could take a look at other x/y splits than 19/1 and make tables of relative 'power'. I.e. he needs to decide and fill-in: - to which spell rank every x/y feels natural to have access too - to how much spell uses, every x/y feels natural to have access too Taking into account how damage is going to be calculated; plus having these tables that represent dependency of [spell_rank_access], [spell_uses], [power_damage_coefficient] from character and class levels, he can get the final scalling formula. And in theory it should hold even if the level cap is increased, provided that it doesn't have breaking points. Immersing into a watcher and cipher was the easiest part)) I had problems with: - might and resolve attributes. - inability to enforce quarantine in Heritage Hill or at least inform that people should not go there, until watcher returns and properly inspects the machine. - sometimes I was presented with A, B or C reply options, but as my character I would select neither. - plus my eye couldn't help but notice that all "+x% attack speed" talents/enchants don't affect attack phase at all; (which brought me from immersed state and made me test stuff; cose I was thinking that something is deeply bugged; turned out to be wrong tooltip wording).

I was just taking a note of: - the fights that I've found the hardest - average DR of most common trash mobs And optimized my weapons/stats/party_composition/roles for those, before even starting a new run) P.S. I suppose that indeed nothing is set in stone yet. PoE1 mechanics (judging by some old threads, and variable names from the code) were changed a lot too.

Hmm. I didn't even consider the system from RP point of view. I agree that such approach would help immersion.But, how to put it lightly... PoE1 had enough immersion breakers for me; to shift my personal source of pleasure into mechanics department. So I am more concerned with those) Hell, I have spent twice more time testing stuff, than actual playing) Yes, there will be problems. But all solvable. Even class-related stats, can be increased according to special conversion formulas or tables; or simply in NWN2 style (like they did with hit die). Well, yeah. According to that ^ you can have:- A 20/0 ranger who has access to rank 10 spells/abilities | i.e. power_level 10 - Or 10/10 ranger/x who has access to same spells/abilities a 15/0 ranger would | i.e. power_level 8 (and same with 2nd class) Although according to the presented conversion table: A 20/0 character would have 60 power_source_points | which corresponds to power_level 10. A 10/10 character would have 40/40 power_source_points | which corresponds to power_level 7/7. (i.e. not eight) I am not sure I understand your example.Do you compare 16/0 ranger with 12/4 ranger/barbarian? Or 20/0 ranger with 16/4 ranger/barbarian?

Because: - it's a way to brainstorm the balance of presented system (basing on the information we currently have). It is sort of feedback for the devs to gauge our approach for finding weak, breaking-points in the presented mechanics. It might help them fix some balance issues they have potentially out-looked, before PoE2 goes live in the worst case; or even before encounters were designed in the best case. - it's also a pleasure in itself) He told that multiclasses will be at 75-85% power level of single class. But we don't know if he referred to all multi-class options or only to equal splits 8/8 (10/10). From his power_source_points to power_level progression, it was seen that 17/1 and 16/2 can achieve power_level 9 of a 18/0 character.

You had to see Torment's Reach, before it was fixed in v3.04. Because it was recursive)

Hmm, do we really want control and be forced to use weapons_x vs monster_x; weapons_y vs monster_y exclusively? I think it's great when there is more variation in viable options, each being with their pros and cons, which you also have to figure out. Instead of just checking damage_type_symbol shown by the game itself. What talent are you referring to? Tbh I liked that calculation around damage ranges. Mig had additional diminishing returns due to being diluted by other damage coefficients. On the other hand relative gain from +1 Mig was increasing when fighting enemies with higher DR. Making tables to see what happens, and finding the sweet spot between Mig and Dex was an interesting task. The proposed system (provided the info we have atm) feels a bit too straight-forward. Like someone holds your hand and tells/forces you to switch weapons. (and because the difference between multiplicative 0.3 and 1.0 is quite big; it's not something we would be able to ignore). So all optimization will become: knowing enemy AR, and aiming either for Pen = AR; or Pen = 2AR. Aaand that's all. P.S. thanks for other answers. They seem to be the likest thing to happen.

Fereed did it again Transcript: link P.S. It a bit worries me that there are so few (almost none) mechanics related answers. PoE1 has achieved a marvelous level of balance between Dex, Mig and Per, partially because of the way encounters were designed, allowing for DR and immunities to balance out the relative benefits in action speed increase (from Dex) and attack resolution shifting (from Acc). I.e. it looks like encounters have to be designed after all mech basis has settled down. Edit: ushas you ninja)

At least it doesn't have strange breaking points (in dmg output function) when your dps suddenly increases by x3.33, once you add just one more point of penetration. Tbh I have a strong feeling that someone was really inspired by the armor penetration from Tyranny, and really wants to make it right this time) Sincerelly speaking proper penetration system is both interesting and perfectly balanceable. And although I am no expert at weapons realism, I have a feeling that quick and light stilettos, along with spears, pikes, warhammers and really heavy great axes can indeed have some nice penetration. And balance wise: - for quick weapons, penetration could compensate their low base damage; - while for heavy ones, it could compensate their attack/swing duration. In Tyranny that approach didn't work through. But it wasn't because of penetration system itself, but due to lack of play-testing, their attack-speed system, abilities/talents like Hundred Fists and Riposte, multi-strike property and haste spell (which coupled together made the slowest of 2H weapons the one and only optimal choice)

There is indeed a big flaw, if the system works in the way it was presented in Josh's post. I hope through he meant a bit another thing. And that DmgRoll and Penetration are added together before the [downgrade] check is done. Will try to explain it via examples. So let's say we have 3 enemies and 3 characters: Enemy X has 1 pierce AR. Enemy Y has 3 pierce AR. Enemy Z has 9 pierce AR. Character A has 1 pierce Pen. Character B has 2 pierce Pen. Character C has 9 pierce Pen. A hits X [dmgRoll = 5] | dmgRoll + Pen = 6 vs 1 AR | result: 100% of 5 damage goes through A hits Y [dmgRoll = 5] | dmgRoll + Pen = 6 vs 3 AR | result: 100% of 3 damage goes through A hits Z [dmgRoll = 5] | dmgRoll + Pen = 6 vs 9 AR | result: 30% of 5 damage goes through B hits X [dmgRoll = 5] | dmgRoll + Pen = 7 vs 1 AR | result: 130% of 5 damage goes through (because Pen >= 2 * AR) B hits Y [dmgRoll = 5] | dmgRoll + Pen = 7 vs 3 AR | result: 100% of 4 damage goes through B hits Z [dmgRoll = 5] | dmgRoll + Pen = 7 vs 9 AR | result: 30% of 5 damage goes through C hits X [dmgRoll = 5] | dmgRoll + Pen = 14 vs 1 AR | result: 130% of 5 damage goes through (because Pen >= 2 * AR) C hits Y [dmgRoll = 5] | dmgRoll + Pen = 14 vs 3 AR | result: 130% of 5 damage goes through (because Pen >= 2 * AR) C hits Z [dmgRoll = 5] | dmgRoll + Pen = 14 vs 9 AR | result: 100% of 5 damage goes through This will be basically the good old [20% of damage always goes through DR]; But now the threshold will be 30%; and there is also penetration at play (to give you a chance to deal bonus damage i.e. those 130% if your Pen is double their AR) TL.DR: it would be nice if: - for damage [upgrade to 130%] you would need to have Pen > 2 * AR - for damage [downgrade to 30%] you would need to have Pen + DmgRoll < AR

That's interesting... and it rises so many questions... - does that mean that [Arquebuss_1 with 9 Pen on 100 dmg roll] and [Arquebuss_2 with 11 Pen on 30 dmg_roll] will deal same damage vs enemy with 10 AR (in respective damage type)? - how does [Pen vs AR] multiplier {0.3, 1, 1.3} (or {-0.7, 0, 0.3}) stack with [graze/hit/crit] and other damage coefficients? Multiplicativelly or additively? - is that [Pen vs AR] added in addition to [20% of damage going through DR/AR no matter what]? Or instead of it? - are there any means to increase your Pen? Or will we have to deal reduced damage vs dragons and other high AR creatures? - does that also mean that all (non-raw) damaging spells will have Pen rating attached to them? How about DoTs?

I have probably missed the question. But if you are asking for personal preference, I would say: party. I like fiddling with character builds such that they perfectly complement each other. When the party acts as a well assembled single unit. Plus there are bonus interactions, banter and it is also less tedious (in PoE at least; in Tyranny for example it's less tedious to solo). But solo'ing also has pros of it's own. Greater challenge. Curiosity. But mostly it's fun to check and see if you can solo that <x> boss.

- Paladin: I always had a feeling that paladins (aside from their double FoD alpha-strike) lack steady dps. An extra full attack ability or a talent that would add extra FoD charge/charges could be quite handy. Also there could be some talents that would augment 2h weapons, because as it is now, melee FoD paladins are forced exclusively into dual-wielding. - Rogue: this class feels weaker partially because there are more enemies on PotD. And rogue lacks both AoE and hard-cc. Reducing party limit to 5 will ameliorate this problem through. - Fighter: decent class, which unfortunately never finds a spot in my party. Mainly because he has no solid passive buffs / auras / synergies to provide to the rest of the group. Giving him an ability to rally his teammates or some passive 'bolstering presence' talent would be great.

It's indeed premature But, with the information we have, I am inclined to try: - Aloth: wizard + cipher (especially if blast/minor-blights generates focus); or wizard + barb (citzal style) - Eder: rogue + monk (deathblows torment's reach); rogue + priest (deathblows shining beacons); or fighter + monk (if take the hit generates wounds) - Pallegina: paladin + fighter (lady of the pain style); could be paladin/chanter if there will be a way to combine Immolation with Dragon Thrashed - Maia Rua: ranger + cipher (twin arrows + driving fight focus generation) - And for MC: cipher + monk (if there will be no hp/endurance split and Torment's Reach generates focus)

Iirc you only need 14 Per to spot the ring. 10 Dex to take it and 10 Mig to drag her out. Edit: check this post.

Hmm, looks like I'll have to roll with a male MC for a change...

Could be interesting. But it depends how exactly split is done. I remember in DA:O there was a split at Battle for Denerim. I didn't care much for companions that I didn't travel with. So I had my party of MC + Alistair + Leliana + Wynne in BiS gear and with fine-tuned tactics going for Archdemon; while the rest, half-naked were fighting the hordes at the gate. It was fun,.. but I'd prefer if there was a temporary split of current party. For example due to a collapse in some dungeon. At the same time Heijoushin made a good point about potential imbalance of such a split.

And my banana of treason... ^^

Hehe. I'd just prefer there would be an option with silent charge, as if character was focused, yet with a sly smile on her face. The only combat sound could be just a slight heh, once the last of enemies falls down. I.e. something, something rogue style.

Imagine your group as members of some fantasy SWAT. They will use weapons and tactics that lead to minimal loses It's the developers task to provide an array of different, yet viable tactics to use. It's a player's task to find the most optimal ones

That won't be a problem as headbands were suggested as head gear that is available to everyone. The balancing is done via their lower stats. For example at the time you find Maegfolk Skull (with 4 Mig and Unbending), you also can discover a +2 Mig Headband. True. And technically they all scale at the moment: - (Fire) Battle Forged: as already mentioned - (Moon) Silver Tide: base_healing = 10 Endurance +3/level after level 1 - (Death) Death's Usher: indirectly and weakly but a bit scales, because as enemy hp gets bigger, so does their 0-20% hp range. - (Nature) Wellspring of Life: +3 Mig, +2 Con, +2 Dex - are all percentage modifiers. The problem is that godlikes might be balanced at some moment; but increasing lvl cap and adding stronger headgear options can destabilize their situation. So if devs will want to add more powerful hats later on, they will be more restricted in their design space, as they will have to take godlikes scaling into account as well. Tbh headbands don't solve this problem completely, but at least reduce it by a factor of 2.

Yeah. That'd be great) Or if it's too costly to look up for extra voice-actors, (for the preset lines); it would be cool to at least use some sort of voice modulator to do the task (i.e. new presets from the existing ones) Edit: And it would be absolutely awesome, if we could be able to mod/add such presets too. (something, something like portraits)

^ I.e. do chants and invocations suffer from accuracy malus of medium and large shields? Yes. Also it's worth mentioning, that wielding [1h+nothing in offhand] grants a +12 acc bonus to some of chanter's abilities (and not just his auto-attacks). For example it works with Dragon Thrashed. Thanks to Boeroer for noticing.