Loren Tyr

Replay value aside, even if you do end up with the same group of NPCs every time that would be by choice and it would be the NPCs that you like the most (both in terms of the characters themselves as well as the mix of classes and abilities they offer). Suppose they had only put five in PoE; there's no telling it would have actually been the five that you like playing with. And personally, even I did stick to the same five and only those five were available, that would still annoy because then they are forced upon me rather than that I choose to take them along. So I definitely would like to see more rather than fewer NPCs, and preferably also with distinctive backgrounds / races / abilities that aren't necessarily available for the player character or hirelings. Though it would be nice to have some kind of pre-built hireling as well, something in between full NPCs and regular fully modifiable hirelings (and available only from a specific inn / hiring location; possibly also in the form of random Stronghold guest, so you could actually take Korgrak along with you for example). These would have some unique features and maybe a bit of (written) background, but without extensive quest and dialogue and such. This would I think represent a nice way of adding some additional more distinctive NPC options to the game, but at a low cost in terms of development resources.

Yeah, it's the same as the Vital Essence bug: https://forums.obsidian.net/topic/87978-vital-essence-bug/ (though there it's a bonus). Essentially, the -25 gets multiplied by the class Health multiplier (eg. 5 for Fighters) and the CON Health/Endurance bonus. So your character drops -175 = -25 x 5 x 1.4 (suggesting that he has a CON of 18).

For Defiant Resolve the reason is that it is implemented differently from normal StatusEffects (it's actually four StatusEffects, but never mind that). Essentially, as far as I can tell, the Defiant Resolve effect is permanently active, given a bonus of 0 with the Duration parameter value set to 0. It also has a StatusEffectTrigger which goes off when Will has been targeted, which resets the activation timer, applies the DurationAdjustment of +10 (and the effect adjustment of +10). This DurationAdjustment is not modified by INT. Whether not applying INT here is an oversight is an open question. Normally the INT modifier is applied whenever the StatusEffect takes effect, but in this case that would presumably be when you start or load the game, way before it triggers. It should be possible to still add it when the StatusEffect.OnTrigger() function is called though. They could temporarily set the Duration parameter of the StatusEffect to the DurationAdjustment value, call the CalculateDuration function and increment the current duration by the adjusted value this gives, then reset the Duration parameter to its original value. But keep in mind that this is very general code, it's not remotely specific to Defiant Resolve. This will thus affect various other abilities as well. Which should be fine (emphasis decidedly on *should*), but these things do always have the potential to have an unintended knock-on effect somewhere down the line. As for Scroll of Confusion, duration seems fine to me. Just tried it, and for me it gives the usual 8s base adjusted by INT, same as when casting it directly from a spellbook.

It doesn't, it's specific to the paladin. I was wondering that myself so I was testing it out, which is when I discovered the suppression.

All else aside: given that the whole spells per rest thing harkens back to ye olde D&D, it's about as far away from innovative as you can get. And that origin to a degree has always been part of the problem with it, in that in the translation to cRPG there no longer was a GM to keep the narcoleptics in check when needed. I'd also say that in terms of target audience, with respect to gameplay it is probably more on the axis of action RPG to 'strategic' RPG (for lack of a better word; ie. slower-paced, pause-and-play style or possibly turn-based RPG). IE games had the resting system because D&D had it, but a certain nostalgia aside I can't imagine that many people are drawn specifically to the somnial component of it. Changing from (or reducing the emphasis on) a per-rest approach to some alternative like a resource- and/or cooldown-based structure doesn't inherently change the pace or mechanical and strategic complexity of the game, and by extension the target audience.

With regard to multiple paladins, note also that multiple Sworn Enemy's on the same target lead to all but one of them being suppressed as well.

Based on deisum's save I have been able to consistently reproduce the bug. As far as I've been able to determine, it happens when in a first combat Psychic Backlash is triggered (normally). This then leads to it triggering at the beginning of the next combat, immediately after it starts. It is not dependent on AI settings, the only requirement seems to be just the Psychic Backlash triggering in the previous combat. It is also erased by reloading. If you have a combat where Psychic Backlash triggered and save immediately afterwards, then walk into a next combat the bug will manifest. But if you then reload the save you just made and walk into the same next combat, it will not. I've uploaded a save here from which it can easily be reproduced: https://www.sendspace.com/file/j88st2. This is just deisum's save advanced slightly to get it to a convenient testing point. To produce the bug, do the following: immediately after loading the save, move up and left with the protagonist Cipher and Pallegina while stealthed. You'll want to avoid the Blight moving up and down the North-South corridor there to get to the druid to the left; but from the save, if you start moving immediately you miss the Blight and get to the druid at a spot where you can fight him without the Blight seeing it. Once you reach the druid, unstealth the Cipher and start blunderbussing (AI is off, so that's all you'll be doing); Pallegina's just there to remove any annoying Blights the druid might summon with her Abjuration ability, she doesn't need to do anything otherwise. In most cases the druid will fairly quickly get to casting Nature's Mark, triggering Psychic Backlash (normally). Finish off the druid and wait for the combat mode to end. Save the game, then move the Cipher back right and up to the Blight that's patrolling there. Right after combat mode starts now Psychic Backlash will trigger, targetting the Cipher herself. Load the save game you just made, again move the Cipher to the Blight to initiate combat again; this time the Psychic Backlash will not trigger. Note also that in some occasions I did this same thing but the druid only cast a series of damage spells before dying (still only blunderbussing, no spells cast) so the Psychic Backlash never triggered in that combat. In this case it also never triggered when starting combat with the Blight, effectively confirming the hypothesis that it is specifically the earlier Psychic Backlash trigger. I also repeated after first removing all equipment (except Lead Spitter) from the Cipher and turning off all modals on her and Pallegina, bug still happened, so it's not an interaction with some item either.

Targetting preferences also vary by enemy though, and to some extent also by ability for a given enemy. There was a thread on this a while ago actually, just can't seem to find it at the moment though. Edit: found it, it's here: https://forums.obsidian.net/topic/86852-im-a-tank-why-would-i-want-resolve/?p=1810130 So in the Shade example given there, having good Freeze DR would reduce the chances of the Shade trying to drain you, but it might well still try to hit you the old fashioned way if your Will is low (or lower than other targets). So just slapping on a bunch of armor isn't by itself going to solve the problem of enemies targeting your main character.

I have to say, I tend to play like this as well, so to a degree it is also self-regulating (or self-regulatable, at least). But given the relative ease and low cost of resting it does start to rely a bit too much on that kind of self-regulation, I think. What factors into this is is probably also the fact that they mostly removed the whole fatigue thing, which while far from perfect does help to impose a bit more constraint on resting, especially in conjunction with the limited camping supplies. As is, you can just go to whatever inn gives you the most relevant bonus, rest there and go back; the main constraint there is basically the number of loading screens that takes, but that'll generally be the same anyway (and I have very little sympathy for people complaining about this being a nuisance; they should just rest less). Similarly, the constraint on the number of camping supplies now is a bit pointless, since you can just shop for more whenever you need to. While the fatigue mechanism only partially counters that, obviously it depends greatly on distance on nearest inn, it does help impose at least a bit more cost on resting, as well as increasing the value of local inn resting bonuses.

Yeah, the 'n_Andred' part is just to designate that it's for the knife throwing game (well, the 'n' is to designate that it's numeric actually). These are global variables accessible from anywhere in the game, so they need to be sufficiently distinctive to make sure their meaning is clear and they don't clash. You also have n_Teal (and b_Teal) variables for the dice game for example. This n_Andred_games_won gets incremented by one whenever you win a game of Orlan's Head. You really just need to have at least three wins and at least one more than Andred, so eg. 3-2 will do.

That dev must've conflated the two conditions for obtaining the feat I think, or they changed it. From the px1_00_cv_andred_knife_toss.conversation file: <Parameters> <string>n_Andred_games_won</string> <string>GreaterThan</string> <string>n_Andred_games_lost</string> </Parameters> ... <Parameters> <string>n_Andred_games_won</string> <string>GreaterThanOrEqualTo</string> <string>3</string> </Parameters> Those are the two checks performed to in order to grant the Flick the Wrist talent.

I was interested in how those games worked as well, and found that there wasn't a definitive answer available that I could find. So I dug around in the game source for it, and since others might be interested as well I thought I'd revive this thread for a bit and post it here. Anyway, the Dozen's dice game really is just that. There are no stats that affect the outcome, it really is just a game of chance with a 50% chance of winning. You need to win three games in total to get the Dozen's Luck feat, so it should just be a matter of trying long enough to get there. The Orlan's Head knife throwing game is more a game of skill however. It essentially does a 1-100 hit roll and adds a bonus based on Perception and Dexterity. A roll of 20 or less is a miss, a roll of more than 100 hits whatever you were aiming for (ear/nose), and anything in between hits different parts of the head (including ear/nose), by the way) at different probabilities (again also depending a bit on what you were aiming for). Perception and Dexterity both contribute equally to the bonus (/penalty) to the roll, at +2 per point of PER/DEX in excess of 10. Andred himself has 15 DEX and 14 PER, so gets a bonus of +18. Winner is the player with the highest score from three throws, or from a tie-breaker as needed. Since I couldn't be bothered to do the math on that (multinomial distributions annoy me), I ran a quick simulation instead to compute the win probability against Andred given a particular DEX/PER bonus. These are shown in the first plot (I'm a statistician, so I'm allowed to do nerdy things like these). As this shows, like in real life you should always be aiming for the nose. This gives a significantly better chance of beating Andred in an individual game (the vertical grey line denotes Andred's bonus of +18; he always aims for the nose as well, by the way). However, to get the talent you actually have to win at least three times, and you need to have won more often than lost. I've plotted the probability of achieving this after (at most) 5, 10 and 25 games, shown in the second figure. As shown, if you're as good as Andred you have about a 50% chance of getting the feat after (at most) five games, though there is still quite a chance (about 20%) that you're still not there after 25 games. If you have just average DEX/PER you might as well not bother, you'll likely be there forever.

Wouldn't Enigma's Charm be a nice addition to that as well? If you're doing Sworn Enemy first the five second warm-up doesn't matter much anyway, and it makes it a wholly per encounter affair. Takes up a talent slot of course, but it frees up an item slot (or gives you more Charm power).

Dexterity speeds up everything, including things like casting, drinking potions, etc. Basically there are three phases when performing actions: the action phase, the recovery phase and optionally the reload phase. There's also a slight amount of idle time before starting the next action phase, but that's not affected by any bonus or penalty. The base duration of the action, recovery and reload phase is of course determined by whatever you're doing, what weapon you're wielding, etc. Slower weapons have longer action and recovery duration ('average' and 'slow' for melee have the same durations by the way, counter-intuitively). All three are modified by dexterity, reload speed bonuses obviously only improve reload time. The most common time of bonus or penalty are those for attack speed. These affect recovery, not the speed of the action itself (only dexterity affects that). They mostly affect reload as well, though not all of them (armour doesn't, I believe). These bonuses are also generally additive (there are some exceptions). For example you might have a -50% armour speed penalty and a +20% attack speed bonus from an item, this would result in a net attack speed penalty of -30%.

When using traps, the accuracy penalty (or bonus, for Chaotic Orb) in practice is always exactly twice what is listed in the description. I've been wondering for a while now whether this is a bug or intentional, and since I can't find where the doubling is even coming from I thought I'd ask here. What certainly must be a bug is the accuracy on the Poison Dart Trap. This is listed as -15 in the description so would be expected to be -30 in practice, but actually turns out to be +11 instead.

I've reproduced the Knock Down + Consecrated Ground issue, a savegame with setup is here: https://www.sendspace.com/file/z9g62j and see attached image. The timing is a bit finnicky but it's quite doable. It does seem to confirm that when the Consecrated Ground pulse coincides with the Knock Down attack, the Knock Down StatusEffect is active on the character and gets applied (as a general "on hit, perform prone attack") to any other 'attack' as well, including Consecrated Ground. This suggests that it also applies to other special attacks (which I confirmed with Crippling Strike, see attached), and probably also other pulse attacks though I didn't try that (pulse attacks that affect enemies rather than allies presumably would apply the Knock Down or whatever to enemies instead, logically). To reproduce from the attached save, I did the following. First, set Edér's Resolve to -2 to make him easier to crit and set his Constitution to 200 and do HealParty to give him plenty of health and ensure that he doesn't get Knocked Down by the Justiciar. Now attack the Justiciar in front of him and then stop attacking (AI is off as well), so he is ready to attack when I need him to. Wait until the Justiciar lands a crit to trigger Consecrated Ground, this will take a couple of hits since the ACC-DEF differential isn't enormous but he'll get there eventually. Once he does, we get to the finnicky bit. If the count is (0s / pulse) . . (1s) . . (2s) . . (3s / pulse), then you need to click to attack with Knock Down on about the first . after (2s). The attack animation needs to land the hit just about when the pulse flashes. I actually got the hang of it fairly quickly and could get the effect reliably from first or second Knock Down on most attempts.

Just a bit of follow up on the DoT front, I've been experimenting a bit with it, turns out they changed the code! (I actually have a dump of an older version I can compare it to) Based on the code itself and a bit of experimenting, how it seems to work now is as follows: DamageOnTick DoT (which is most DoTs, eg. Deep Wounds, Necrotic Lance, etc.): damage is fixed per tick (gets Might bonus), ticks on 0, 3, 6, ..., etc. Still does fractional ticks at the end if you have less than 3 seconds left (with damage proportional to the fraction), but what's changed is that it now does this regardless of how small the fraction is (it had to be at least 1/3 before). If duration is an exact multiple of 3 the last tick does full damage, eg. Deep Wounds of 9 seconds will do base 3 damage on 0, 3, 6, and 9 (not sure whether the last tick in this case did or didn't happen before). DamageOverTime DoT (eg. Wounding Shot, Enduring Flames): total (intended) damage is fixed and gets spread over the ticks. Ticks are on 3, 6, ..., etc., but not on 0. Damage per tick is 3 x (total damage) / (total duration), and this now also gets a fractional tick at the end (it didn't before, if I recall correctly). Interestingly, if the duration is an exact multiple of 3 it triggers twice on the final tick for full tick damage. So a Wounding Shot of 6 seconds that is supposed to do 18 damage will end up doing 9 damage per tick for a total of 27; but on 5 or 7 seconds it will do the intended 18. So for my Thick Paladin Gunner build I'll need low duration but in a multiple of 6. An addendum: the double trigger on the final tick doesn't actually happen reliably. It happens much more often on a graze than a hit or crit. But it does start to happen quite consistently when you're running in slow mode (this must be linked to the DeltaTime, the size of the discrete time increments on which the game operates).

By the way, Death's Usher triggers at 25% rather than 10% Endurance, so that can still be reasonably helpful. Not sure it's necessarily worth losing a head item slot or an alternate racial ability, but it's not too bad either. Bloody Slaughter still kinda blows though. Might help a bit against very high DR targets and it helps you crit with status effects as well, but that really does tend to fall in the "overkill" category. Mostly for a theme build, I'd say.

I did a quick check of the source files for Necrotic Lance, which gives a partial answer: the Skeletal Wizard Necrotic Lance (which has a separate implementation) actually is implemented with an AttackBeam, whereas the normal Necrotic Lance uses AttackRanged (the WM1 Substantial Phantom and Lagufaeth versions also use AttackRanged, though the minor spellbinding of Necrotic Lance is again an AttackBeam; not sure which item actually uses that though, if any). So that answers why it's happening, although it leaves me a bit puzzled as to why it's different. If any passing dev reads this: any confirmation on whether this is actually intentional?

Where? Because as I recall they only retracted the earlier statement that it would be the final patch. That hardly constitutes a clear confirmation that there will be another one. So although there might be a 3.04, we hardly have any clear indication that there actually will be. Let alone any information on what would be in it or what the time frame on it would be. And considering that they seemingly didn't do anything at all on 3.03 in the last couple of weeks that was in beta either, I'm hardly filled with a warm glow of confidence that there actually will be any further updates. And given their poor recent track record in terms of communication I don't see that changing anytime soon. Which is unfortunate, because I expect that if they did do a better job (and had done to begin with during the 3.03 beta) of communicating, they likely could (and could have) elicit a clear contribution of (some of) the forum members here to that process.

Looking at the screenshot Edér's actually hitting himself as well, and so is your Cipher with his melee attack (isn't it also supposed to be raw damage, by the way?). Did this happen before or since you took this screenshot, or was it a one-off issue? Sometimes there seem to be glitches like these that resolve themselves when you save and load; more like the game state got a bit muddled at that particular time rather than it being a consistent bug in the ability itself.

There are also less murderous (well, less unprovoked anyway) ways of getting the Stag helmet, there's one in Dyrford Crossing for example. You can sometimes get weird glitches if game files are off actually, though usually they aren't really corrupted as such but just the wrong version or something (ie. it didn't update properly when it should have). I agree it's not a likely scenario in any case, though. And for the Consecrated Ground + Knock Down I'm reasonably certain that's just those two interacting strangely, it's been reported before. It was supposed to be fixed now though: https://forums.obsidian.net/topic/87360-shod-in-faith/?hl=%2Bconsecrated+%2Bground. It's mostly the Necrotic Lance acting like a beam that I can't place. Anyway, like I said I'll see if I can reproduce either one, try to find out why it's happening.

Mechanics generally, really. I see plenty of posts on these boards of people asking about how stuff is computed (or implicitly showing that they have misunderstood it). For attack speed, but for example also just simply how damage is computed with all the modifiers and lashes and DoTs, that sort of thing. And building on top of the more purely mathematical/mechanical side of things, also more on the practical implications of that (eg. the merits of dual-wielding vs one-handed vs two-handed style in terms of damage output). This is the sort of thing that comes up frequently and there is a lot of info in older threads on it, but it would be nice to have it all together in a nice, accessible format. But yeah, it's a lot of work to compile that sort of thing and probably much more so in the wiki format, as you mention; the more applied part is perhaps also more of a guide than pure description, so probably not entirely appropriate in the wiki anyway. So if I would do something like this, I'd probably be inclined to put it on the forum here and get it stickied in any case.

Darn. INT of 1 puts the duration on a hit at 5.5s at the lowest, that's no good. Then I'd be better off with 6.5s (unless DR is more than 17% higher than total (intended) damage), it's almost like it doesn't pay to be a gibbering idiot! I don't suppose there are any (non-weapon/shield) items that give you a general duration penalty? I vaguely recall there being something like that, but I don't know for sure.

The wiki is fairly inactive as far as I'm aware. So if you have something to contribute to it, I'd say go for it. It's a wiki after all. It could actually use a complete overhaul I think, with all the changes that have been made since the original release (in particular with regard to the gritty details of how things work; eg. attack speed mechanics and modifiers). Have been contemplating doing that myself, but the scope of the task has been stopping me so far.