Dr. Hieronymous Alloy

Problem is if you start saying there are racial differences between different subraces of humans it can . .ah. . . invite more controversy than the game perhaps needs. It's probably best if the differences are purely cosmetic.

Sure, but I am interested more in spells which make use of strength. If strength based mage wont use spells much, that will be disappointing. In PoE1 I mostly used Cipher as weapon DPS/crowd control character, so for that build resolve is not,helpful anyway. There is now choices to be made between weapon and spell damage but in my opinion it’s a boon not a problem (at least in theory). You could make a strength based mage that used buffs and CC and so forth, it'd just be the damage powers that you'd avoid; your dps would all come from autoattacking with your equipped weapon, whatever that was. Similarly you could still make the classic might/dex/per/int Cipher that dumps resolve, you'll just need to avoid all the damage powers. That's about half the power list, and includes things like Silent Scream that combine debuffs and damage, or Pain Block that combine buffs and healing. You'll be locked into a much narrower selection of powers and a much tighter role than you were before. (You'll also run into the issue that all CC powers blow right now because they take longer to cast than the expected value of their durations, but that's a separate confounding variable, the rock to this change's hard place. ). As above . . . the problem isn't so much that a given build is impossible or sucks now, as that there's less freedom in character building. You have to specialize, and to the extent that you don't specialize, you'll be penalized.

So did the changes to proficiencies and weapon styles not make it into the patch then? Are those still coming or back on the drawing board?

I don’t know how “might” was pitched originally as I wasn’t following development of PoE1, but “might” has never been tied to soul power within the game. It is clearly described as a stat describing physical strength and spellcasting capabilities + scripted interactions clearly tied this stat to physical strength. Looking forward to creating: 1) cipher 2) high strenght and low resolve mage and seeing if those characters are as useless as some people suggest. You could probably make a high Strength autoattack weapon mage and do alright. Not great, but alright. Try dual wands. Similarly, multiclass melee soulblade ciphers can still be pretty effective, mostly because you just ignore most or all of the cipher powers and just spend your focus on bonus melee damage. One big problem you'll run into with testing this change is that there's a confounding variable -- the extroardinarily long cast times on all powers and spells right now -- so testing out pure casters is just kinda blaaaaah regardless, and almost any hybrid you play right now tends to default towards weapon damage anyway. It's hard to test out a melee wizard right now for example because the summoned weapons have such long casting times that by the time you've finished summoning your weapon most of the fight is over. That said I don't want to imply that any build is "useless" even with the new changes; I'm pretty sure you could theoretically solo the first game using a character that hadn't taken any stat points at all at character creation, just left everything at ten. It'd be frustrating, but still possible. The game design is pretty flexible and lots of stuff is possible, it's just a question of how much pain is involved in making it possible. The difference between an effective character and an ineffective character isn't one of useful vs useless, it's one of how many times you feel like reloading.

Indeed, that's why I suspect it was done to please the anti-muscle wizard crowd. Nah, everything I've seen the obsidian guys post indicates it was just "we need to give Resolve something, this is something, let's try it", but they also thought it would make the D&D Heritage folks happy because so many people come into Pillars from earlier games and just expect the Strength stat to be useless for casters. Not so much "muscle wizard hatred" as "this confuses new people because it isn't what they expect." Which is actually another reason I prefer Might over Strength: it's a very clear and stark and up-front indication, to anyone who bothers to actually read the screens in character generation, that this isn't just a D&D reskin; it's a different system and you're in a different universe and nothing's gonna be quite what you're used to. In a way it's almost a branding element worked into the character creation screens, an up front advertisement that things are a little bit different. And in the long run I think that helps new players because there are a lot of different ways these games aren't Dungeons and Dragons and it's good to emphasize that right away with the very first number they see.

Without looking at the one-handers, just here, there's a reason the greatsword needs to be 25% damage superior to the other two handers -- it's got lower Penetration. Now that being down a point of Pen means a 25% damage loss, without the 25% damage penalty on all the other two handers, when you run the math really counter-intuitive results start popping out vs. various AR ranges. If you moved the Estoc down to the same damage as the quarterstaff, for example, the Estoc would be clearly superior vs almost all damage ranges because it would hit overpenetration bonus damage much faster. I think the appropriate "fix" would be to move the quarterstaff into "group three," collapsing the calculation into two "groups," and just consider "+25% base damage" one of the Greatsword's two perks. Every weapon gets two perks, for the Greatsword it's two damage types and bonus damage, for the Estoc it's double extra penetration, etc. for the quarterstaff it's reach and accuracy, etc. edit: I'd probably collapse a lot of the one-handed weapon groups too, there are just too many tiers and it's too confusing. That problem is more complicated though

When I first heard them describe the longer cast times, what I thought they were planning was, like, a new, fifth, extra-long casting time (say, instead of fast/average/slow/very slow, fast/average/slow/very slow/"ritual"), that the very top end "closer" spells would have, and then in a given fight you could try to cast a Big Epic Ritual or the enemy would try to cast a Big Ritual and you'd have to scramble to interrupt or counter it before it went off. Something like that could be amazing. Have a little popup appear above a character's head ("Llagufaeth has begun casting a Ritual!") like the current floating "No Pen" text, so it would pop out of the combat log spam. Could make combat very reactive and awesome.

Roughly speaking, yeah. Might is preferable to Str/Res; give Res something else new. That said if we're talking a little more specifically Might would need some further tweaks; MaxQuest lays out an excellent argument for what those tweaks should be here (essentially,that Might should be additive with other passive damage boosts like soul whip, and then that total boost should be multiplicative), but that's getting into more detail than we maybe need in this thread right now.

The duration on Cipher charms (whisper of treason, puppet master) are actually ten seconds, not six; when you run the math though and adjust for misses and grazes (vs. equivalent defense etc.), the expected duration does work out as less than the cast+recovery time, so you're literally taking yourself out of the fight for longer than you're taking the enemy. The math on that does change a bit once you adjust for dexterity, intelligence, etc., but it's still crap. Compare with the first game where Whisper of Treason also had a ten second base duration, but a "fast" casting time (which I believe was a little over two seconds in practice). To get a comparable ratio of cast time to duration, Whisper would need a duration of over thirty seconds in Deadfire! And then you have the problem that a thirty second charm effect is just crazy long (especially after Intelligence is figured in) and functionally game-ending (especially when cast on players). So you can't just multiply the durations out because then it's too powerful. Personally, I think a lot of cipher powers specifically need to be returned to fast/instant cast (Ciphers have to build up focus anyway; they're already casting far less than the ex-vancians in Deadfire). This is especially true for the crowd control powers (for the reasons above). Past that, for the ex-vancians. . . they need to go back through and, for each power and spell, decide whether it falls in the "this is a standard every fight bread and butter kinda thing" (i.e., minor missiles, summoned weapons, etc.) or the "this is a decisive spell that you bust out for boss fights or the big closer in a standard fight" (i.e., Ningauth's Shadowflame). First category needs to be relatively quick and easy; save the long cast durations for the second category. Oh and give people who aren't wizards a way to generate Concentration.

I'd also prefer a return to Might. I understand why the change was made (resolve needed something) but I hope they can come up with something else that's useful for Resolve. The main reason I don't like this change, though, is that it makes hybrid builds and hybrid classes a lot more difficult to play and unworkable, for reasons I've gone over in detail previously -- you only get so many stat points at creation, and the more you have to spread them out, the less effective you are at any one thing. Previously, you could make a front-line priest who was a good damage dealer and a good healer, or a mage who was good with summoned weapons and spells, or a druid who was effective as a shifter and as a caster. Now with this change everyone has to pick something to suck at; it's a lot less possible to make an effective hybrid. This is, of course, worst for ciphers, who due to the focus mechanic, have to be good at dealing physical damage in order to generate focus to do anything else; this change makes it much more difficult to make a single-class cipher that's good at anything (though part of this issue is occluded by other issues like current casting times). If you put your points in Strength, you can generate focus, but you won't have the points for Resolve and literally half (I counted) of your powers will be much less effective; if you put the points in Resolve, you'll have effective damage powers, but you'll have a much harder time generating the focus to cast them (due to low Strength). Given all that, it seems like a really weird design choice to replace Might with Str/Res at the same time that multi-classing is being added to the game and the party size is being reduced from six to five, both changes that make hybrid characters more relevant and important. It feels like the game is shunting players towards making hybrids and then simultaneously hobbling hybrid builds specifically. There are some ways they could try to force this fix to work -- for example, giving Ciphers the option to use Resolve for weapon damage instead of Might, etc. -- but they all feel like the game design equivalent of adding epicycles to the geocentric theory of the universe. Yes, with enough special extra rules and tweaks you can make it work . . . but the elegant, better solution (i.e., in this analogy, Might / Heliocentrism) is right there in front of you.

So, I've made a table of how I personally perceive (at the moment) the relative "powerfulness" of classes based on stat in comparison to each other: (for beta1, and somewhat extrapolated for beta2 based on the changelog) Yeah, I'd been looking at your chart. It's hard to disagree. I think I'd want to frame the problem differently; the main issue I see, with the Strength/Resolve change specifically, isn't so much that a Strength-based weapon mage is worse or better than a Might-based weapon mage, but that it's mechanically more difficult for hybrid builds; you have to choose between being a summoned weapon wizard and a spell wizard in a way you didn't before; Ciphers have to choose to be actively bad at something (either weapon damage & focus gain, or spell damage, or CC duration/effectiveness) where they didn't have to make that choice before. Realistically, you can max out two stats without taking a penalty somewhere else and more if you're willing to take a hit somewhere else, so the stat/class balance ends up looking more like a graphic equalizer than a clear 1 for 1 table. The broader class balance issue right now though is spell effect durations are orders of magnitutude too short to justify their casting times; for example, for a number of Cipher CC powers, after you adjust for miss/graze/crit, the expected-value durations are so low that statistically you're better off never casting them at all and just using your autoattack instead (extremely min-maxed stats can break out of this problem but only just barely). TO be honest though I need to spend a lot more time with the new beta before I have anything new to say about stat balance; right now I'm going by the numbers more than by play experience and that's dangerous. I agree with your suggestions about additive/multiplicative damage stats, I'm still pretty sure Perception needs the return to 50% criticals, Con needs something that isn't a reskinned hit point bonus. I don't like Strength/Resolve replacing Might but I understand why they did it I just hope they come up with a better answer.

Iv'e been thinking about this and I think the answer is to change the "casts per level" advancement path so that instead of 2/2/2/2/2 etc it moves in a tree, i.e., at 20th level you have 5 level one and level 2 casts, 4 level 3 and 4 casts, 3 level 5 and 6 casts, 2 level 7 and 8 casts, and 1 level 9 cast. That is, five casts that are either level 1 or level 2 powers, four casts that are either level 3 or 4, etc., for a total of 15 casts in any single fight, only one of which can be a top-level power. Set up the progression in a tree so that you've always got more low level casts and your top level casts are always a scarce once-per-fight resource. Reason being, right now top level vancian casters can simply flood the map with powers -- a level 20 wizard has 40 spell casts to play with per fight; they're functionally post-scarcity. Note that casts per level is a different thing from powers per level etc.

Technically speaking, today, you got your wish: a new Pillars of Eternity Beta Patch.

Yeah, in theory that's how it *should* work out at higher levels at least. Thing is you spend a lot of time at lower levels (especially if, like me, you reroll a lot!) and you're also buying passives, so the net effect of all the changes is a real lack of versatility vis-a-vis the first game (at least for the casters I've tried so far). It might be different once they do a balancing pass and more spell abilities are worthwhile.

That's all true and I see the upsides, but I think it's an overcorrection because it makes fights really repetitive -- there's no tactical choice of what to do in each combat, there's only the strategic choice of what to pick on level up; you're using the same abilities every time because they're the only ones you have (unless you're a wizard).

Could be but the thing is .. everyone lost a lot of flexibility due to the reduced allotment of powers. See, e.g., priests, who went from getting all spells every spell level to getting at most two. Ditto druids. Even Ciphers get one fewer power now than they used to (used to be 2/1 alternating levels, now it's 1/level). Wizard grimoires are a ticket out of that bind that right now nobody else has.

Caster trinkets are all gonna functionally need to be spellbook equivalents or else the wizards have a huge versatility advantage.

I think you only need the right hand (18-stat) columns, really. I only ever heard Sawyer phrase the design goals positively, not negatively -- i.e, "you should be able to be a smart barbarian if you want", not "you should be able to be a really dumb barbarian if you want." Looking at the columns on the right yeah Resolve and Con are the weak sisters.

As I understand the "no bad builds" goal, the idea isn't so much that anything will be workable as that you should be able to make anything work -- i.e., if you want to make something normally counter-intuitive, like a highly intelligent barbarian or a muscle-bound wizard, here should be a build for that concept. That doesn't mean that (again for example) a wizard who pumps Might but takes only CC powers should be just as good as one that takes pumps Int and CC powers; just that if you want to build a Mighty wizard you can do that (by focusing on damage powers) and if you want to build a smart wizard you can do that too (and focus on AoE & CC powers, or DoTs). Honestly I [mostly] like where most of the stats are right now (i.e., first beta build) apart from Perception and Resolve. I'll also be a bit contrarian and say that I think it's at least somewhat important for stats to stay the same game to game so people can transition characters from one game to the next fluidly. It's not a huge deal if there are minor changes but overall people want to transfer their characters so you want to keep that possible with a minimum of re-drafting.

Right now pretty much all magic isn't worthwhile, apart from a few heals and buffs, because the cast times are crazy long relative to the effect and the lack of grazing means everything misses all the time. Thing is, all that's a known issue, and they've announced a bunch of different changes in the next patch. We're basically all twiddling our thumbs right now waiting on said beta patch because there have been enough changes that it's hard to predict how they'll interact until we get our paws on it.

There's also a problem among two handed weapons; compare greatsword, morning star, and estoc. The price the Estoc pays for the penetration bonus is so huge that for most armor ranges the Greatsword is far preferable due to higher base damage, despite lack of penetration. This will be especially true next patch when they move to a more graduated penetration penalty (a good change that was needed). Almost all the weapons need a fair bit of tweaking one way or another. From what they've said in the streams there are a bunch of tweaks in the next patch so let's see what happens.

Do you have any thoughts on Perception? How does the math shift if we move back to a 50% critical instead of a 25%?

another good discussion of Deadfire ciphers here: https://forums.obsidian.net/topic/95146-state-of-ciphers%E2%80%A6/ It's one forum "up" though.

The PoE system does somewhat discourage stat dumping/spiking, in several ways. 1) By design intent at least, every stat is useful for every character, so if you take any one stat to minimum, you're creating a weakness. 2) If you math out exactly how much additional benefit you get from each additional point of dex/might/per, there are diminishing marginal returns -- you're better off, generally speaking, with more even stats than you are spiking and dumping. (MaxQuest gave an example of this above). Only real exception to both those rules is that Resolve isn't sufficiently useful for non-melee characters but that's a known issue they're working on. There's also a small issue that almost any caster does want to max out Intelligence but generally this isn't a problem because almost everyone who plays a caster wants to do so with max intelligence anyway.

See, that's what comes of crunching the numbers. Solid work. It could be a little confusing but I suspect it wouldn't matter too much because anybody sufficiently willing to dive into the numbers to the point that they started to care about things like 'is the bonus from Might additive or multiplicative?" could figure that proposed system out. Simple phrasing is "bonuses of the same type are additive, then you multiply each group of bonuses."