thelee

Deadfire pathfinding is mostly better than PoE1 except for one very specific "optimization" they made in Deadfire: ignoring perceived spammy clicks. What this means is that if you tell a character to move to the same "spot" (whatever the unit is for pathfinding), they will re-use their existing route if possible without re-processing anything. Why is this annoying? Because if they take a stupid route (because of some temporary obstacle or something) and you want to tell them to stop being stupid, you can't just re-click the same location and expect them to find the better route. You have to completely give them a different destination or press X to cancel all action, and then click for them to find the smarter route.

has this already been reported as a bug? re: the original topic, finishing blow is less useful than you think (at least in my experience having rolled an assassin/priest of skaen a long time ago) because it competes at that level with shining beacon and divine terror, both of which i think are good candidates for assassinate. pillar of holy fire is obviously a good pickup late game and synergizes well with any +fire advantage gear you might have. i forget if searing seal works correctly with assassinate, but for AL5 that's pretty much reserved for two "free" uses of shadowing beyond.

yet another z-ordering UI issue, this time the glowing eyes on the figure at the bottom of the screen floating above the looting box:

edit - this post has been edited to clarify the nature of this bug. For any devs, please do skim this thread, angry posts aside, due to the many examples being given. there are many interactions with afflictions that appear to be just fundamentally broken now, such as counters, suppression, resistance interaction, or weakness interaction. I do provide a dropbox link and output_save for a specific such case in this post. edit 2 - a more complete summary of the rest of this thread i posted on page three, but i'll re-quote here: original post: I thought this was just limited to Suppress Affliction, but it's not. Just now, Tekehu got hit by a debuff (Putrid Blast) that both sickened and immobilized him. That's fine, his AI script is set up to use his shapeshift (which grants dexterity immunity) when he has a dexterity affliction. It dispelled the immobilized. But he still can't move. In fact, when I try to move, the game erroneously claims that the remaining Sickened affliction is the reason why he can't move: Given that I'm also a software engineer, I try to be patient and understanding about bugs that crop up because I know they're a fact of life, but seriously, what the hell is going on that these kinds of regressions are occurring. This is especially alarming given that the game is pretty much approaching end-of-life for support if not with the next patch to bring turn-based mode out of beta, then soon after that. Dropbox link to output_log and save: https://www.dropbox.com/sh/kbocdfa6ag9vejf/AAAqIQSwvXVwQ6ktHld7Qzcia?dl=0

One thing you can bear in mind - you can use it to self-damage yourself at the start of combat to trigger Llengarth's Safeguard as well as potentially to get your health low enough to use Potion of Final Stand, both of which you can then extend (along with all other buffs) with Wall of Draining. The trouble with Blood Sacrifice is that if you have cast a few spells from different levels, it's pretty random on which ones it will give back, and you can easily kill yourself spamming it too much unless you have a no-death effect (like Potion of Final Stand). it's also a good synergy with human or death godlike racial passives. multiclassing with priest or having a priest in your party solves the accidental self-knockout problem because unlike potion of final stand you can use barring death's door without needing to be near death first (and if you want to go for this angle, you can follow up with a potion of final stand after you've safely been dropped to near death). mentioned this in messages with steampunkgears but i'll reiterate it here just for anyone else coming to the discussion - blood sacrifice is powerful because you can control what spells you get back, and I suspect the only reason why the entire forum community hasn't come to an agreement that blood mage is like the most powerful subclass ever is because people haven't actually tried to play the blood mage like this, instead just using it as a slower, random form of self-empower for spell cast restore. it's worth "inefficiently" trying to use blood sacrifice at times, just so that it helps restore spells you do want later down the line (like if you opened with a few self-buffs, even though you don't need to restore spells yet you can use blood sacrifice to fill up again so that when you do want to restore spell casts you don't end up restoring spell levels you didn't want). it skews you heavily to AL1-3 and AL4-6 (to a lesser extent) spells, but it's worth it because spells in poe1/deadifre are generally better designed to be good from beginning to end regardless of spell level (unlike BG where frequently lower level spells were just not worth the time late game, hellloooo larloch's minor drain) @thelee I understand the (+7 Accuracy,+15% Damage) near death. but i feel like the +1might and +1res are a waste here. unless you go 9 might as a base stat and put that one point somewhere else like con, same with res i guess. it's actually when bloodied or worse, so it's fairly easy to trigger (death godlike is pretty hard to trigger without barring death's door or potion of the final stand). in the right set up the human racial can way trump anything you get from a more optimal stat distribution. key phrase being "in the right set up"

One thing you can bear in mind - you can use it to self-damage yourself at the start of combat to trigger Llengarth's Safeguard as well as potentially to get your health low enough to use Potion of Final Stand, both of which you can then extend (along with all other buffs) with Wall of Draining. The trouble with Blood Sacrifice is that if you have cast a few spells from different levels, it's pretty random on which ones it will give back, and you can easily kill yourself spamming it too much unless you have a no-death effect (like Potion of Final Stand). it's also a good synergy with human or death godlike racial passives. multiclassing with priest or having a priest in your party solves the accidental self-knockout problem because unlike potion of final stand you can use barring death's door without needing to be near death first (and if you want to go for this angle, you can follow up with a potion of final stand after you've safely been dropped to near death). mentioned this in messages with steampunkgears but i'll reiterate it here just for anyone else coming to the discussion - blood sacrifice is powerful because you can control what spells you get back, and I suspect the only reason why the entire forum community hasn't come to an agreement that blood mage is like the most powerful subclass ever is because people haven't actually tried to play the blood mage like this, instead just using it as a slower, random form of self-empower for spell cast restore. it's worth "inefficiently" trying to use blood sacrifice at times, just so that it helps restore spells you do want later down the line (like if you opened with a few self-buffs, even though you don't need to restore spells yet you can use blood sacrifice to fill up again so that when you do want to restore spell casts you don't end up restoring spell levels you didn't want). it skews you heavily to AL1-3 and AL4-6 (to a lesser extent) spells, but it's worth it because spells in poe1/deadifre are generally better designed to be good from beginning to end regardless of spell level (unlike BG where frequently lower level spells were just not worth the time late game, hellloooo larloch's minor drain)

FYI - do people remember baldurdash? there were tons of fan-made bug fixes for BG and BG2. (there was also a fan-made patch for Fallout 2, which is extra hard considering that I think that required actually hex/binary editing of .dlls since it wasn't as open to modding as the infinity engine.) games are definitely more complex these days and likely more prone to bugs, but it's not like the 90s-early2000s were some gloriously bug-free zone of gaming. I just really hope that Obsidian has a few more patch fix cycles after TB-mode gets out of beta, because these regressions are frustrating.

you know what, what really doesn't make sense though is that in PoE1 traps were super lame (and extremely bugged, even now) but then they added new traps in White March I/II that were every bit as powerful as the best traps in BG2. so i literally have no idea what went on in obsidian's balancing brain for PoE1. I was just happy that traps weren't altogether bugged in Deadfire like they were in PoE1. I just wish they scaled a bit better into late game - traps mostly just become useful for their debuffs and you still need lots of mechanics.

For some reason I thought this got fed through the same double inversion as everything else. Of course it doesn't, and 3% is what you want in that case. Ugh, what bad design for the stat to have increasing returns on both effects. Resolve does get fed into the double-inversion system, but probably not in the way you were anticipating. If you have a -% bonus on hostile durations due to having resolve greater than 10, that gets inverted to cancel out much more +% duration effects (so a resolve of 20 would get you -30% hostile durations, which when inverted can cancel out up to +42.8% worth of duration bonuses, e.g. from intellect). You might have been thinking that each point of resolve gives you a separate -3% bonus, which when inverted can make it impossible to get up to -100% or more duration (which would be absurd), but no stat works like this, they all sum up their effects first before being tossed into the inversion grinder. and yes, it is unfortunate that it has increasing returns on both effects. at least increasing returns means that high-end resolve can be extremely powerful, whereas Con is just forever kind of a lame stat right now.

i feel like obsidian must have PTSD from how stupidly powerful traps were in BG2 and Throne of Bhaal. You could beat the demogorgon in a split second with enough spike traps.

Hard not to feel really frustrated at these seemingly glaringly simple UI errors being introduced almost a year after release. To reproduce: 1. Select save game from menu 2. Click on an empty save or existing save to rename and create a selection box. 3. Scroll away 4. Notice that the selection box alone continues to scroll, instead of being clipped within the bounds of the save/load menu. This is pretty low priority UI polish, but I can't help but feel that it is a "code smell" of other z-ordering or UI problems (including the floating text that persists between screens that is still an issue). There's no dropbox link or output_log or save, because this issue is not linked to a game state.

check combat log (hover/shift over some aoe damage to see all modifiers)? the right-click menu on abilities is both more comprehensive (notably damage PL scaling) but also less comprehensive (conditional effects, etc.) when it comes to modifiers than the combat log.

This is my dropbox folder for just deadfire bugs. And this is after I culled it for all the issues I reported pre 1.2 and issues that had been clearly fixed and weren't getting any further views. edit - also, even if nethack is complex, there's a difference in types of complexity. netflix is mechanically complex, sure. but it's at its heart a very simple program. you don't have to worry about shaders or threading in nethack coding. edit 2 - I had to cull in the first place because I was hitting the Dropbox free storage limit. Keep in mind that I only use Dropbox to store but reports for PoE1 and Deadfire.

it's been a while since I tried this, but i think it's like PoE1 where you can have one of both, but putting on on top of/too close to the other will deactivate it.

unfortunately this got nerfed a bit - sparkcrackers used to trigger two times (for two attack rolls to try to connect) but now (as of like 4.1) only trigger once. still really fun (and probably the most powerful application of sparkcrackers) i think the possible problem with trying to extend sparkcrackers "distract" (not distract) duration is that it would require a re-work of the stealth system. if you keep making noise in the same place, i think they might just end up walking into the area of effect and start combat.

for a blood mage i personally find constitution to be more important than virtually any other builds since health becomes a way to regenerate spells, so the extra health cushion can be useful. (But i am playing on potd + skaen/ondra/wael/berath challenge, which means i have no information on my specific health (wael) and also knockouts become lethal after 10 seconds (berath), so the convenience of the extra health cushion for me might be specific to that set of challenges.) might is particularly bad for blood mage since IIRC, might will increase the raw self-damage. as a part rogue you will already have lots of damage boosts for martial sneak attacks (and eventually deathblows) anyway, so balancing that out with perception is more important than having a wee bit more might. move might points elsewhere (but don't go below 10). again keeping in mind that my recommendation for con might be influenced by my challenges, i would personally recommend moving might and possibly even some perception into con (a blood mage can spam eldritch aim as needed if accuracy is a problem). for a simple intro to how a blood mage can play, i would pick up chill fog, eldritch aim, and then at AL2 concelhaut's siphon. overlapping chill fogs to ensure constant blinds (for sneak attack and general debuffing utility), eldritch aim to max accuracy for both sneak attacks and chill fogs, and then concelhaut's siphon spamming to keep your health up should give you a nice preview of how powerful a blood mage can be. If you want an easy time for early potd fights, get slicken instead of chill fog (and just use chill fog when you get a grimoire) and just repeat spam slicken (plus optionally eldritch aim to help land) until everyone who's not a floating creature is constantly tripping and you can just pick them off at your leisure.

I mean, wasn't this the case with PnP or even classic IE? A dagger did 1d4 and you didn't get to go any more often than someone wielding a 2d4 bastard sword. I'm not saying that this was a good thing (I only understood it as a role-playing choice in BG/BG2 to be deliberately suboptimal), but how do newer TB or editions of D&D handle this? In 5E D&D, weapon damage follows a fairly predictable formula. First, there are two tiers of weapons, Simple and Martial; Simple weapons start at 1d6 damage, Martial weapons at 1d8. In addition to damage, weapons have various properties, which can upgrade or downgrade their damage die. While not every weapon has a property, they typically get one beneficial property for free. So longswords (bastard swords are no longer separate) have the Versatile property, meaning they can be used either one or two handed; they're still 1d8 weapons in one hand, but they go up to 1d10 two-handed. The Light property, which allows weapons to be dual wielded by default, and the Finesse property, which allows attackers to use Dexterity instead of Strength for their attack and damage rolls, are particularly relevant to this discussion; these almost always comes with a damage downgrade (the only two Light Martial weapons are also Finesse, and do 1d6 damage instead of 1d8, which is consistent with getting one property for free and losing a die size for the second; one is Finesse only and keeps the 1d8 die size; one has Finesse and Reach and does only 1d4. Simple weapons include four Light weapons that take a damage downgrade to 1d4, including the Sickle and Club which have no other properties, but also the Handaxe, which has an additional property but keeps 1d6 damage). Finesse and Light are both fairly desirable properties in this system in spite of the damage die tradeoff: Dexterity is a better stat than Strength outside of melee offense, so sacrificing a bit of damage on your weapon to be able to focus on a better attribute is a worthwhile tradeoff. Dual wielding is good in action economy terms: it's one of the easiest ways to get a damage dealing Bonus Action (you get one Action and one Bonus Action per turn; non-martial classes never get more than one attack per Action and only high level Fighters get more than two attacks per action, so a Bonus Action attack is a big boost to DPR). Daggers specifically are still bad, but this is mostly because they're Simple weapons and any character that frequently attacks with a weapon will find a way to get a Martial weapon. As Simple weapons go they're arguably above the curve, with three beneficial properties, including Finesse and Light, but only one die size downgrade from the Simple weapon base (though things don't really go below 1d4, so they're at the floor). interesting. in many ways that sounds like how deadfire structures its weapons (e.g. avg 12 dmg for "fast" weapons, 16 dmg for "slow" weapons, with one free weapon trait, and then damage/PEN downgrades as you add on other stuff), just over-optimized for RTwP. would be curious to see what changes (hopefully some) come down the line for TB mode.

i think one of the biggest things you could do to improve traps would actually just be to make it so that you could sustainably keep a supply of higher-level ones going (vendors don't stock too many and they tend to stock lame ones). with sufficient scaling, a party of five mid-high level traps is passable mayhem. but definitely not a sustainable practice.

while they don't combo anymore, woedica's fists are still among the best weapons in the game. legendary-scaled monk fists by level 17 with a +31% raw lash (for perfect disposition) is even better than what a monk might be able to accomplish (at least without stacking bonus PL).

I've seen similar lines of argument over at reddit before, but I feel the evidence doesn't support the case. The first game itself would have suffered over time sales-wise if that was the case, but to my understanding it held up in the sales department well after the first few months in release. Moreover the audience reception seems very positive across all platforms if reviews and aggregate ratings are anything to go by, be it GOG, Steam, Amazon, metacritic, etc. Moreover the current completion rate for the first game is at 12.6%, which may not seem like a lot but is higher than Divinity: Original Sin's which is at 8.8% (as per Steam achievements). In theory there could be an abnormally disproportinate opinion in the silent majority but I think it's something of a stretch. by comparison, the Deadfire completion achievement "The End of the Beginning" is at a whopping 19.4%.

I mean, wasn't this the case with PnP or even classic IE? A dagger did 1d4 and you didn't get to go any more often than someone wielding a 2d4 bastard sword. I'm not saying that this was a good thing (I only understood it as a role-playing choice in BG/BG2 to be deliberately suboptimal), but how do newer TB or editions of D&D handle this?

Also, re: moon godlike scaling: Yeah, that's why I want to push for making things more integrated systematically into the game, instead of creating parallel effects. I think Bellower subclass showed just how poorly chanters are integrated into PL scaling to their detriment. If a Moon Godlike can improve their healing with improved might, Prestige, or Potion of Ascension or heck, the Heart-Chime Amulet (+1 PL for moon godlikes at night, seems like a perfect match), that seems to me a better scenario than either their current fixed 10 healing or a separate character-level scaling. Like I said, ideally also chanter chants would also use some sort of PL scaling, so that PL scaling is more impactful for them (as befits a caster) and also doesn't create a weird situation where multiclass chanters can chant just as well as single-class chanters. I'll probably add more responses throughout the day or two, but those are the big ones I wanted to mention real fast.

@MaxQuest (boy this thread is getting hard to quote/parse, so here's some specific responses): Riposte - currently it's 25% on miss, right? I propose something like a 15% chance on miss or graze. I think my original idea of having a separate chance for miss and graze is unnecessarily complicated. But maybe it's unavoidable considering TB's very huge graze range. Sabre/Stiletto - I also strongly disagree with these changes. I also disagree with the assessment that the +50% recovery time penalty is not worth it for damage dealers. It is definitely worth it in most underpenetration changes. But I mostly disagree because not suffering a +50% recovery time penalty was something that was special to swords, and you're taking that away. Perplexing Sap - in case you don't know, the reason why this ability is broken is because "confused" means that you don't distinguish between ally/enemy... including for the perplexing sap ability. So you actually end up basically confusing/hobbling the entire battlefield (yourself included) from this. It's awful and one should never upgrade Sap. That's why I'm proposing leaving Sap alone and re-designing Perplexing Sap to be something less stupid, because even Obsidian has acknowledged that Perplexing Sap is an ability they wanted to revisit. Prone - I mentioned this in passing, but I strongly disagree with lengthening the actual prone-on-ground duration. Prone is not hard CC and we shouldn't blur the line with it. It's a stronger interrupt (+50% interrupt time on a perfect interrupt, can be conditionally much better against fast enemies). The only thing I would change about prone is to make the actual interrupt effect +3s instead of +2s like a standard interrupt (so you're proned for the same duration but the recovery time is now 3s instead of 2s) because it seems like an unnecessary complication/murk about interrupt vs prone. Seals - I really don't think they need adjustment, I think they need bug fixing. By "hack" I mean "a fragile workaround". If hazard effects ever get fully fixed, I think seals become gangbusters (except for repulsing seal) and they definitely don't need a 3s cast time. I could get down with repulsing seal having a 3s cast time though, since the out-of-combat cast is pretty negligible, and the distinction between it and pillar of faith at AL2 is very murky in combat (raise your hands if you knew that repulsing seal prones on graze vs pillar's prone on hit, and hands up if you knew that repulsing seal gets +2 acc/PL instead of +1 acc/PL for pillar of faith) and so it would help differentiate it without being notably that much more powerful. (Note that everything I say about seals also applies to other hazard effects, which include some wizard walls and I think a druid effect or two). Kalakoth - I actually don't think -10 acc and -10 reflex make it that much more usable. Again I reiterate that a stronger single acc debuff (-15 or -20) is more usable for AL1. edit - for that matter priest Barb's is only barely useable and has longer range and is -10 all defenses and wizards really should be better at this kind of effect than priests at this level I think. Resource generation - given that there exists a paladin talent that unconditionally grants +1 resource and that costs an ability point, conditional resource generation seems gamely targeted at 0-2 resources for a typical fight with minor metagaming (hence me sayign so for Virtuous Triumph in a different comment). I would think that the current fighter talent is close to this range--albeit requires particular metagaming so you don't have an anti-synergy going on, and has higher potential upside than Virtuous Triumph. If there's one thing I would fix, it'd be the aforementioned anti-synergy - a Barbarian gets special "on-being-crit" effect because they actually are expected to get lower deflection on top of having low inherent deflection, whereas fighters have inherently high deflection and have a lot of passive boosters. The only thing I can think of off the top of my head is a moon godlike-type effect where you get +1 discipline the first time you are dropped a health category. Maybe that's too good, but seems to me you limit the upside (like paladin Virtuous Triumph) and it's a little bit less anti-synergy and also a bit less murky (low probability odds for impactful events strikes me as murky).

Then it's clearly a bug, becose all these abilities have proper keywords (Punishment and Fire/Electricity) in both abilities.gamedatabundle and attacks.gamedatabundle. I guess something in the code prevents game from reading/using these keywords. U can create a topic in Technical support and write about it. If someone haven't done it already... I've reported hazard issues sooooooo many times. And keyword interaction is just one aspect. All hazard effects do less damage than stated in tool-tip (which is odd considering tool-tips are auto-generated), they don't get PL scaling for damage, they don't get empower bonus for damage, etc. Hazards were also extremely weird and buggy in PoE1, you'd think they'd have learned their lesson for Deadfire, but... (From what someone else mentioned in one of my bug reports, it's not that the ability is missing stuff, it's that the ability creates a separate hazard effect, and it's that effect that is missing keywords or stat interactions). yeah, I could be down with something like that. if Obsidian feels like doing some new content...

Something I've learned is that low completion rates is not even close to abnormal. Lots of players just never finish the games they pick up, regardless of the game. It doesn't really have much connection to satisfaction rates. That's the whole reason why Berath's Blessings were introduced in the way they were instead of like an actual "new game+" mode, because they didn't want to gate it behind completing the game.