Dr. Hieronymous Alloy

How much deflection do you get from a large shield? From memory, a small shield gives around 25 or so deflection, which is fairly big; three AR is also pretty big though -- that's, what, padded to scale? I can't check exact figs right now. Overall I feel like the increased power of healing due to the health changes is probably going to be a bigger factor than anything. Right now I can see three ways to make a functionally invulnerable character in Deadfire: stack healing effects and just facetank things, stack AR (against all damage types), stack deflection (and other defenses). I suspect the real issue will be characters who manage to stack all three of those on top of each other, not the relative value of one of them vs the other.

It's hard for me to feel like this matters all that much. It's a difference of degree not of kind and deflection got kindof overpowered in the first game anyway -- after about midway through the game Eder became nigh-invulnerable just due to stacked deflection. As long as it's still possible to make an effective high-deflection character, I'm not sure another stat making it redundant is that big a deal. Like, you can still make a high-deflection low-armor character, and it'll still work, you'll just be vulnerable to rare lucky hits overpenetrating. Plus, we don't know what monsters and stuff will be out there. Just a few bosses or monster types with high Pen but low Acc could make deflection really strong. If there are lots of ways to buff AR that could help too. This feels like a fine-detail issue to be tweaked later after the big fixes are all in place, possibly even a post-release kind of issue.

Ok yeah that changes the math dramatically to the point that we need to know what is intended before we make any other suggestions. If it's just five over for overpenetration then the Estoc becomes much stronger for example.

Yeah, it'd still have a use. Just a much narrower range of uses =(

I hope someone mods in wind speed and direction into the world map some day, so you have to tack and so forth. I mean, it would be too prohibitive and awkward in the base game, and Obsidian shouldn't waste time on it. But I'd love it as a challenge mode.

This seems to be the simplest, most straightforward solution. You could also, say, give people (10+/-[stat]) points of concentration or interruption at the start of each fight. Even with a system where Per/Res give interrupt/concentration, you have to have concentration do something so that melee people care about it and not just casters. Added penetration for extra concentration seems like the obvious answer, then interrupt takes that away. Lot of things would have to be playtested though -- can Concentration go negative, or does it have a floor? What rate of gain/loss "feels" right? How is Interrupt "spent" ? etc etc. A system like that could work and could even provide an interesting added complexity layer on top of the current combat mechanics, as you try to keep opponents off balance, and it would give casters a way to boost spell Penetration, which they sorely need and don't really have right now. Could get overpowered though if it led to everyone overpenetrating all the time or trivialized other Penetration bonuses (why have an estoc if you can just concentrate a bit?). Seems like it could be worth exploring though and we are still in beta so there's some time.

Looks good overall. I'd make a few minor quibbles -- i.e., I'd probably substitute the word "addressed" for "solved" sort of thing Maybe mention that healers face the same division-of-roles issue that hybrids do, especially for traditionally "melee" classes like druids and priests. My hope is that Obsidian will be able to think of something we haven't, honestly, some new mechanic that's outside the box we're thinking in.

In one of the earlier streams, pre-beta, the overland map ship was very clearly a miniature, with a base, not an actual ship sailing around. I get why they changed that but I kinda preferred it. I'd like to see an animation of your isometric-view character marching through the wilderness but there may be engine limitations, dunno.

I don't want the NPCs chain-casting instant AoE disintegration at me, no! I want the game to be fun and challenging. A six second "Big Spell Incoming! Siren! Alert! Alert!" type mechanic with a little popup and suddenly I have a limited amount of time to interrupt the enemy caster's Big Incoming Nuke could be a lot of fun. Vice versa too if I have to defend my casters so they can get their Big Nukes off without the enemy interrupting them. I don't want it to take eight seconds to cast a ten second charm, though. That's not fun it's just lame.

I think there is definitely a place for long-cast (6+ seconds) "ritual" type spells. But they need to refine the concentrate/interrupt mechanic first, those spells need to be pretty potent to the point of being potentially decisive, and they should be high level powers not base level bread and butter stuff.

Even six seconds seems a bit long to me! I really like the idea of giving Ciphers and Chanters very short cast times and even shorter recovery times, so that they can wear armor with minimal penalties; both of them really suffer compared to the ex-vancians and they need something to balance out the fact that a Wizard can dump their whole spell book before a Cipher or Chanter gets two casts out.

Pretty much every change they've made in this beta has been either in response to player complaints, part of the attempt to incorporate balanced multiclassing, or both. (Josh Sawyer posts on Somethingawful, both from a few weeks back now so a bit outdated). I think what's happening with the Might --> Str changes is that 1) Resolve needed a boost (true), 2) it's easier to move something from a different stat than to come up with a whole new thing for Res (true), and 3) a lot of players were vocally complaining because Might helping Wizards just doesn't fit with their AD&D character template so it's wrong and needs to change (hence why Sawyer commented "all your dreams are coming true!" when he announced the change). I suspect the issue with a lot of these changes is that the people who liked things as-is (i.e., Might) were quiet and content, while the people who wanted change (Str) were very vocal.

Even with multiplicative crits, at 25% crit bonus instead of the 50% it gave last game, Perception is a dump stat right now relative to Might and Dex, giving around a 20% damage bonus if you stack it to 20, whereas Might and Dex each give 30%. To even them out, Per would have to give a 50% (multiplicative) crit bonus.

Yeah, as I understand it, in Deadfire most spell characteristics -- cast time, area of effect, etc. -- seem to be defined by overarching categories in the game data. For example, most spells have either "small" "medium" or "large" areas of effect, which are all standardized. Cast times I'm not sure of the names of the categories but it's basically four -- instant (.5) /fast (3) /average (6) /slow (9). I don't have a particular problem with that in the abstract if it makes the game easier to balance by using defined categories. I worry a little that it will lead to broad-strokes rather than nuanced balancing but I worry about almost everything. Note, however, that this means most Cipher powers -- Whisper of Treason, Mental Binding, etc., -- haven't just had their cast times lengthen as a result of the general move to longer cast times; their cast times are also in longer "categories" than in the last game. For example, Whisper of Treason was a "fast" cast in the last game, which meant 2.8 seconds cast, .2 seconds recovery; now it's an "average" cast, which means six seconds of cast time and 2 seconds of recovery.

Pretty much any fantasy series and especially any fantasy game series, eventually hits the god tier. I was thinking more Athkatla than ToB, though. The ToB stuff can be saved for PoE 3.

I've always kinda viewed PoE 1 as the equivalent to BG1, and hoped that Deadfire would be the equivalent to BG 2 in terms of larger scale and scope.

Mathematically, what would a 30% dual weilding bonus rather than a 50% bonus do to the numbers above? Would it be enough to bring dual scepters in line with other weapons?

I almost feel like balancing the modals is a whole separate question by itself because they're so all over the place -- some are great, some are useless, most need their numbers tweaked dramatically. With the arquebus modal specifically -- yeah, it's basically locking the arquebus into "opening sniper shot weapon", and it can do that just fine, but that's the only thing it can do; if you want to fire twice, it's inferior. Which is great for the people who want to make that specific build but not so great for anyone else who wants to use guns. If you just want to be a "gun ranger", there's no two-hander option for you.

https://pillarsofeternity.gamepedia.com/Timeline might help

I think that's a symptom of the 50% bonus to dual weilding that carried over from the last game, where it was needed to counter the effects of DR; DR converting to AR means that now it gives a disproportionate advantage. Knock that back down to something more reasonable and the scepters and wands drop out and everything else falls within about a 10% variance range (except arquebus). I mean you're right dual wield scepters and wands are clearly out of line with everything else, also, yeah. And I agree that the difference between the arquebus and the arbalest isn't enough to "break the game" -the scepter/wand inbalance is the bigger problem -- but it's enough that anyone running the numbers will always choose something other than the arquebus; it's too much worse -- it doesn't break the game, but it breaks the arquebus (at least, for most builds). One of the things that really got irritating about the first game was that (after about level 10, once you stacked action speed bonuses), bows were always mathematically superior to guns, so pretty much every build shifted over, and guns we're heavily deprecated, especially in WM content. 10% range is (at least for me personally) within the "ok, it's roleplay" range, especially if there are situational advantages (alpha strike etc); more than 15% starts getting painful and you're crippling the character to no purpose. Part of this may be a long-standing raw nerve on my part re: guns due to the aforesaid first-game balance issue.

You really should be taking penetration into account. There's a 4 point discrepancy between the hunting bow and arquebus and that's a world of difference. I'd argue that in this specific case I was taking it into account, because it isn't the only 9 penetration weapon, and I included another 9 Pen weapon, the arbalest, in the comparison.. The Arbalest has a Penetration of 9, but does roughly (3.43/3.81 =~.9) 10% less DPS than the Arquebus. Meanwhile the Arbalest is only (3.81/4.1=.92) about 8% less DPS vs the hunting bow. The gap between the Arbalest and Hunting Bow is fine and balanced by things like penetration etc. It's the additional difference between the arquebus and the arbalest that seems like an outlier that needs fixing. Raise the Arquebus to 3.7 DPS and it's still the worst DPS ranged two hander and still inferior DPS wise to the equal-penetration Arbalest (but better for special "full attack" abilities and alpha strike shots). The point where I'll admit there's some fuzziness to my numbers is whether or not I should be using the "with -5 accuracy" and "with +5 accuracy" modifiers in calculating these numbers, but, all else being equal, they're a real shift, so I think including them is valid for purposes of the comparison (even if in practicality all else is never equal).

"sub-Optimal" in this instance being "a 10% to 17% reduction in average DPS" depending on what weapon you're comparing against, holding all else & all unknowns to be otherwise equal (yes, there will be magical arquebi, but there will also be magical hunting bows, etc.). And yeah there will be a few niche builds that can use the arquebus but it's a clear outlier; at 3.43, the next-worst-dps weapon is the Arbalest at 3.81, while the most effective two-handed ranged dps option is the hunting bow at 4.1. The average DPS for all two-handed ranged options is (3.81+3.43+3.85+4.1+3.9+3.9)/6 = 3.83; the median is 3.875. So the Arquebus is 12% below median damage, while the hunting bow is only 5% above median damage. If we compare Arquebus vs. Hunting Bow, it's a 17% reduction in DPS. That's huge. Even if we're just comparing against the next best alpha strike weapon, can you really justify a 10% damage penalty vs. the Arbalest? (I'm comparing between two-handers only because there are some obvious balance issues with the one-handers, scepters and wands in particular). I'd suggest adjusting arquebus DPS by reducing the reload time such that it has a DPS only 5% or so below median / 10% below hunting bow -- i.e., around 3.7 dps instead of 3.43 (and I'd also suggest giving it crushing damage type, for the other reasons stated above).

I suspect implementing MaxQuest's proposed recast of additive/multiplicative Might might help a lot with melee overpower, without need for many additional changes to melee past that. I could be wrong because this is based more on what I've read than on personal playtesting, but I get a sense that the things overbalancing melee combat right now are mostly 1) Multiplicative might stacking with other multiplicative bonuses, so that a few specific class combos (barbarians with Frenzy) end up really really powerful, and 2) Dual-weild still having a 50% speed bonus despite the changes to the DR/AR/Penetration system such that dual weilding no longer needs a huge bonus to be useful (as it's no longer negatively impacted in the same way by DR). Then combine those with the problems casters are facing (long cast times, dramatically limited power selection, etc.) and that's the difference. Melee's effectively been lightly buffed while casters have effectively been hugely nerfed, net effect is a relatively huge preference for melee.

This is . . . actually a really good point. For single player games this is usually exactly how you want the balance to go -- in something like Skyrim, it makes sense to have players balance versatility vs. specialization in a 1 for 1 tradeoff. The difference I'd point to is that PoE is a party-based game, and in a party based game specialization is inherently rewarded -- all else being equal, it's usually better to have one dedicated healer and one dedicated tank than to have two paladins. So you actually want to give hybrids a slight boost, relatively speaking. Compare with medium armor -- Sawyer has commented in the streams that the game gives medium armors a slight stat boost, because normally people either go lightest-possible or heaviest-possible, so you have to give the medium armors a boost "above curve" to get anyone to use them at all. Another good example is the "power level" mechanic -- dual class characters don't have a 50% power level, it's more like 75% to 90%or so (I think? somewhere in there).