MaxQuest

@thelee, Yeah, it indeed took quite the time) And now, it's been quite interesting to see your point of view. Afterall, the sub-mission was to check how much my view matches that of other players. And to find the alternative perspectives in case I missed something. I don't quite understand though why you disagree with the overall mission, since the idea is to come up with suggestions that do satisfy almost everyone. And if the final list is empty, so hey, what's the problem P.S. it seems that a proper reply will take some time to put together. So I will elaborate on afflictions, racials, priest and general stuff later.

It was actually discussed and suggested at least twice before

Pillars of Eternity 1 has achieved a pretty good state by v3.0+. And several bugs aside, it looks that Deadfire is slowly headed in the same direction. But at the moment, there are still a range of spells, abilities and effects that are what we can call: trash (or trap) options. And also there a few things that are either a little too good, or allow to bypass certain design limitations. And it's understandable that comparative power of talents/abilities/effects/etc in practice is not always the same as initially expected, due to overall complexity and changes in mechanics, plus emergent synergies, opportunity costs, variety of encounters, asymmetry in enemy defenses, armor threshold deadzones, max health 'policy' and so on. So the idea is: to come up with a collective set of fine-tuning adjustments to cut-off the least impactful ones and to end up with a list of community suggested changes. So if Obsidian decides to do any balance changes for v5.0.0, they could use this as some sort of survey data. And of course there is also hope that some of these could be modded. The workflow is gonna be as follows:There will be a preliminary list of suggested changes. And you can: a). suggest to add a change; b). suggest to remove an already added suggestion; or provide an alternative adjustment. Just don't forget to argumentate your opinion. After awhile, I'll make a poll based on these. When the poll is ended, a final list will be made. And it will include changes that have passed a certain threshold of unanimity. The preliminary list of suggestions (to start the discussion) is: Edit: Based on the first round of gathered feedback, here's the rev2 of the preliminary list: Edit: Here's the rev3 version of the preliminary list: Edit: The poll is up and in a separate thread: here P.S. Big thanks to Phenomenum for assistance and theorycrafting on the initial list. P.P.S. Hopefully I didn't forget/miss much)

There is a big difference. When casting from scroll: - you don't benefit from MIG, INT (which is quite important for DoTs and periodic damage) (although this is slightly compensated by scrolls being considered PL0, and thus higher relative_power_level_bonus) - you don't benefit from PER - half of your arcana is used instead of your class power level - but scroll's base cast/recovery time is often faster than that of a spell. E.g: > Pull of Eora: 6s cast and 2s recovery. > Scroll of Pull of Eora: 1.6s cast and either 3.0s or 4.0s recovery (I don't remember exactly). Here's an example of Shining Beacon vs Scroll of Shining Beacon: Special attention to accuracy: - scroll accuracy gets: only half of arcana, while - spell accuracy gets: perception_bonus + ability_level_bonus + relative_power_level_bonus where: ability_level_bonus = 2 * (ability_rank - 1) relative_power_level_bonus = power_level - ability_rank So generally we could say that: - spell accuracy gets: perception_bonus + power_level + ability_rank - 2 And a caster with 9 PL, casting a rank 9 spell, with 20 perception and a perception inspiration, already gets +31 accuracy. Meanwhile you would need 62 arcana to get such accuracy bonus.

It's a bit weird indeed. But you can't leave that Elemental Keyword on attack, otherwise pierce damage won't damage enemies immune to that keyword. But if you want for AA to not get the accuracy penalty with Essence Interrupter when Pierce damage is dealt (because it was more advantageous)... then it's better to tweak the code behind Arcane Archer's Passive itself, e.g: if either primary or alternative damage type is elemental - don't apply the penalty. - first you check your damage types and enemy "keyword immunities" vs those. And discard the options that enemy is immune to. - second you add the penetration bonuses. - third you check enemy "damage immunities" and "damage AR"; and select the advantageous damage type.

Confounding Blind is not totally useless because those -3 deflection stack up to 10 times (which is good vs bosses; or vs enemies you multi-hit or focus-fire) But yeah, I do agree that it is unpleasant that it doesn't stack with -10 Deflection from Flanked because... I would think that Flanked is a passive. Turns out it's not. P.S. I like your general idea of speeding-up the deflection penalty accumulation. I.e. going from -3 Deflection per hit (for up to 10 times) to at least -5 Deflection per hit (for up to 6 times)

It is the same problem as described: here Some enemies are not just immune to damage of type "x"; but have also immunity vs specific keyword. And Essence Interrupter's attack has "shock" keyword. Ideally, weapon attacks perhaps shouldn't have their elemental keywords persistently stored, but rather applied dynamically after. E.g: - you shoot the bow - the game checks that there is Pierce DamageType and Shock AlternateDamageType, and checks if enemy has Pierce KEYWORD immunity or Shock KEYWORD immunity, and discards unmatching damage type. - if enemy is not immune to neither of these keywords, apply PEN bonuses (like Heart of the Storm), compare to enemy AR values, and select the most advantageous one. - if it turned out to be elemental, attach related keyword. E.g: Shock damage type? -> add Shock keyword This way: - You will be able to deal pierce damage to enemies that are immune to shock keyword. - You won't be healing greater blights when pierce damage is dealt. - Arcane Archers won't get acc. penalty when Shock damage is dealt. But they will, when Pierce damage is dealt. But it would require extra handling, and might be inconvenient to implement for the programmer...

Not sure about this... +25% is added in "BonusDamage" section - should be some reason behind this. I hope) But it's not displayed in his damage breakdown tooltip: And the overall calculation is correct:

Very good question. But must say I only knew of upcoming Deadfire, because I often check this forums and occasionally /r/projecteternity. And after Deadfire release, I've only noticed it: - on Twitch - and briefly mentioned on Little Bit News (which is a gaming related feed) I was also unaware of PoE1 until 2016. And got into it only because I was replaying through NWN2, and decided to search if it's creators finally made any new RPG in magical setting. (specifically was searching for "Obsidian" and "Josh Sawyer", hehe) Generally the best ways to reach to a player like me are: - wikipedia (if I've liked some previous games of the studio, I might check it) - twitch - promo/info windows that appear when one opens steam. - and once I heard of the game: gamepedia, where I would check for some introduction on story/classes/mechanics

That first +25% seems to be overriden by the +50% bonus that you get no matter what. As for the conditional damage bonus: you get it only for the enemies that are below 50%. I.e. it is not progressing from enemy hp 100%->0, but from 50%->0. Specifically for Devastating Blow, there are these values (already posted by Phenomenum) (edit: and by error3 above): "BaseDamageMultiplier": 1.5, "TargetHealthRatio": 0.5, "BonusDamageMultiplierPerPercent": 0.06So when enemy gets below 0.5 of max hp, you get a bonus for each percent missing below that threshold. Initially I was under impression, that when you hit an enemy at "zero hp" you would get 1.5 bonus and 3.0 bonus. Or in other words that you get two separate bonuses: +50% and +200%. But no. Looking at: It seems that they are added together, for 1.5 + 3.0 = 4.5 bonus. Or in other words you can get up to +350%. And that explains why OP got those +312%.

I have took a look, and probably this is caused by the following: - Neriscyrlas: has [immunity vs freeze damage] - LAX02_CHA_CRE_Shade_Frost: has [immunity vs freeze damage] AND [immunity vs frost keyword] What does this mean is, that this shade will ignore anything that has that keyword, and this includes Frostseeker. Also I would also like to add that there also are: - CHA_CRE_Shade_Frost: which has no frost related immunities at all - CHA_CRE_Blight_Ice_Greater: which has no frost related immunities; but! it heals from frost damage AND has frost AR as it's lowest AR: 9 slash, 9 pierce, 7 crush, 5 frost

Small update. I have been going through this again. And there is a thing I have missed. CanThreaten() method besides IsWithinPotentialThreatRange() also checks for IsWithinThreatRange(). And while the first check takes into account at least 4m distance... the last one is basically melee range (distance <= your.mover.radius + enemy.mover.radius). Also I have checked from where is ThreatenEnemy() invoked. And it is called from two places: - when you engage an enemy - and during UpdateEngagement routine, for currentEnemy that is attacked in melee So in practice threatenedEnemies is almost the same as engagedEnemies, with the following exceptions: - you can threaten your currentTarget in melee, even if your engagement limit is zero - you can threaten your currentTarget in melee, even if you cannot engage it due to failing minimumLevelThatCanEngageMe check - e.g. when attacking a higher-level barbarian with Threatening Presence. Thus: - "Each nearby enemy reduced fighter's recovery time between attacks" - is incorrect - "Each nearby enemy engaged by fighter, plus fighter's current target (if not already engaged), reduce fighter's recovery time between attacks" - is closer to what really happens So: - it would be nice to either adjust the description to match the effect. - or adjust the effect to match the description.

Absolutely The first thing that CanEngageEnemy() method checks is: if (!this.CanThreatenEnemy(enemy, attack)) { return false; } And yeap, a threatened enemy is, generally speaking, an enemy that could potentially be engaged by you.

Yeah, tooltip is indeed not clear enough. Fyi there are 3 'related' status effect types: - EngagedBy: Number of enemies currently engaging the target. - EngagedTargets: Number of currently engaged targets. - ThreatenedTargets: Number of currently threatened targets. Mob Stance uses the last one. But what is a "Threatened Enemy"?: it is a nearby enemy that AllowsIncomingThreat while you AllowOutgoingThreat and CanThreatenEnemy. Sounds strange... but I don't want to invent new terms, and use those from Game/AIController.cs So: CanThreatenEnemy is if you satisfy the following conditions: - enemy is within PotentialThreatRange, is targetable, is hostile and is not a pet - you are not in stealth/invisible, and (you are not moving, unless the enemy is your current target) - your attack is melee attack (although my current understanding is: if you engage an enemy1 in melee... even if you have only 1 engagement limit, this allows you to pass CanThreatenEnemy on adjacent enemies when you made that melee attack even if it was vs enemy1) - EDIT: there is one more check (at the end of CanThreaten()): enemy must be within ThreatRange. And the difference between ThreatRange and PotentialThreatRange is that the first is 3m shorter. AllowOutgoingThreat is if you: - are not stunned, pushed back, paralyzed, terrified, knocked down, dead, unconscious, launched, grabbed are not currently in the process of using an object and are not in "InUseState" (whatever it means). AllowsIncomingThreat is if enemy: - is not pushed back, dead, unconscious, grabbed Enemy is considered within PotentialThreatRange, if: - distance_to_enemy <= 3 + your.mover.radius + enemy.mover.radius - no idea what a mover is, but fallback value is 0.5f, and in characters.gamedatabundle you can see radius values usually being between 0.5f and 1.0f for human sized enemies (large dragons have 3.0f, and kraken looks to have the biggest: 6.0f). Enemy is considered within ThreatRange, if: - distance_to_enemy <= your.mover.radius + enemy.mover.radius Now... put this is in tooltip P.S. TL.DR: you can consider the following approximation for "threatened enemies": think of them as a list of nearby enemies in melee range. The list starts empty and gets refreshed when you engage / use a melee attack. Additionally enemies get removed from the list when they disengage, die, get pushed, or when you get invisible or hard-cc'ed. The difference between threatened and engaged enemies - is that you can threaten an enemy without engaging him. For example you are attacking a higher-level barbarian (with Threatening Presense); you can't engage him, but are still threatening. P.P.S. But that's theory, based on gamedatabundles and a piece of decompiled code. Better check in practice, if I didn't miss something.

Maybe just make it a longer prone, instead of knock up?

^ Do agree with Wormerine here. It would be nice if there was a setting that would control if players see chance "to-at-least-graze" or "to-at-least-hit". P.S. Personally... I don't even read those chance numbers on mouseover. I only care for color... e.g: green - means I will land that cc for sure. Because usually I check enemy defenses/AR anyway; plus plan my party before starting the run.

^ Should we... ask if that's intended?)

I was thinking about using it for the proc on a wayfarer/tactician moon godlike. He drops low for a moment due to Sacred Immolation, and quickly heals back. But the thing is: the switch has to be automated... because if I spend extra time on micro, this voulge-trick doesn't decrease the real-time spent on combat much. It's definitely a thing, but there is a mod for that Just keep yourself from unlocking last level too early) --------------- Btw, regarding Clear Out: did you notice that it makes a Primary Attack first, and then follows with a cone attack? So it's kinda hitting the main target twice.

I have also seen that many enemies, are actually not knocked-up. For example: Rathun.

Btw, when you are below 50%, can you swap to Darryn's Voulge, auto-trigger the storm, and swap back to whatever weapons you had? Or you need to get below 50% while wielding that pollaxe?

There is also z-index related problem with combat log tooltips, plus with their auto-xy-ordering: This bug is here since beta.

True that) Although the focus has shifted a bit... In PoE1, the player would often debuff enemy fort/will in order to land a stun/paralyze/petrify -> and thus lower enemy deflection and reflex. Otherwise many physical attacks would end up in misses and grazes... and that damage that goes through, would end up eaten by DR. In Deadfire though... targeting one defense in order to lower the other is occurring less often. Partially because of resistances/immunities, partially because stun and paralyze aren't decreasing defenses as strongly as they did. And partially because we have: - spells that lower enemy WILL (Miasma of Dull-Mindedness, Arkemyr's Wondrous Torment, Psychovampiric Shield, Borrowed Instinct) target Will. - stuff that lowers enemy DEFLECTION (Confounding Blind, Crushing Depths, Pure Class, Marking, Brittle Frost, Pike modal)... also usually targets Deflection. Unless you use those weapons with Knockdown or Clear Out, which target Fortitude... which is usually even higher than Deflection.

Looks fine to me. And json is valid. Try restarting the game. And removing / re-adding them to the party.

I haven't played through SSS and FS yet. So take the further insights as out-of-touch from practice ^^ The Matriarch has resistance vs MIG and CON afflictions (which are upgraded to immunities on Hard and PotD). It also has Frightened and Charm immunity. Although looking at gamedataobjects, it is unclear if these immunities target whole INT and RES, or only the mentioned frightened and charm. So I would at least try to check if Dominate and Terrify go through. In either case it leaves at least: - PER afflictions - DEX afflictions (hello paralyze/petrify) - CON afflictions of tier2+. Because the croc has only 15 CON, applying Sickened, will reduce her max hp by 25%, and that is nice. - MIG afflictions of tier2+. If you want to reduce it's damage by ~15%... or daze for -4 PEN. Also Matriarch has relatively low stats: lvl: 20 hp: 3788 ar: 13 (base) 11 (pierce), 9 (freeze), 15 (corrode) defenses: 76/96/79/91 stats: 17/15/14/12/13/14 Especially pay attention to low INT and RES here. Because you can let your cipher/wizard debuff them even lower,.. and your further effects will start sticking for quite a longer time. At the same time matriarch's effects will elapse faster. Also it's worth to note that Matriarch has: > -8 all defense penalty vs frost attacks, and > -15 all defense penalty vs shock attacks Plus, on Story, Easy and Normal difficulties Matriarch doesn't have: Bloodlust and Deep Wounds. Running out of resources could be a problem, if your party doesn't have a lasting potential. But you could optimize that a bit by lowering the amount of hp you have to deal with. E.g: matriarch has 3788hp. Sicken it, and it will become 3030 Bring it to Near Death, and finish with Marux Amanth or Death Ring... and you will have to deal 2272 only. As for damage itself... you can focus on: - the already mentioned Death of 1000 Cuts + frequent shred damage (e.g: AP field + Mind Blades) - dots (including battle axe modals) and Cleansing Flames - mass Ninagauth's Freezing Pillar (if you can place Brilliant on your wizard) and don't forget Recall Agony. P.S. And regarding fortitude: there are MIG/CON afflictions, morning star modal and -10 all defenses from Shining Beacon. -55 is not much, but vs Matriarch should be enough.

Ah, yes! Now I've remembered... When I noticed that it behaves unreliably, I've read the description, got surprised a bit, and... tbh haven't used it since.