Dorothy_Dorothy_

13 hours ago, polytope said:

There are actually three separate things here.

In the majority of maps there are rest-interrupt spawns coded into the .are file, often more than one possible monster, and with variable chance of it turning up, but never dependent on your character's xp. There are also spawned groups from SPAWNGRP.2da encountered at various points in most dungeon/outdoor maps, again, a variety of possibilities, i.e. the same spawn point could have RDUndead (mummies and ghasts) or RDUndea2 (shadows and wraiths), you won't know until you get close.

However there are also "trap" areas in many maps, which function much like an undetectable and undisarmable floor trap and run a script which does check your characters xp (not your level, so multiclass characters get equally difficult spawns) when deciding to add creatures, those scripts like SPWNDEAD.bcs are the scaled ones that can confront you with liches etc., however these are only triggered when advancing through the map and only once for each such scripted spawning, unlike other spawnpoints which may reset over time.

mmmmm.

I think there's a fight right before Dace in Spellhold that strongly depends on your level, because if you do all the dungeons available before going to Spellhold you'll fight a lich and some mummies with some accompanying trash undead there. If you go early, it's a much easier fight. It's not triggered by rest, but when you come down that hallway they kind of pop up there right around the corner. Might be because they're crowded up into the corner so you can't see them until you're properly in the room.

I don't remember if SCS adjusted the spawns in either of these methods, but I'm too lazy to go look through the changelog over there at the moment! heheh

As far as THAC0 goes, there's definitely a jump from shadows to mummy lords. It's more of a pain inherent in "open world" environments. Hard to fine tune encounters when the player's THAC0 could swing 10 points while only gaining like four levels.

On 3/28/2024 at 6:00 AM, RoyalProtector said:

In the third edition, they changed poison so it progressively damages your stats. I don't think it can kill you, at least in NWN/NWN2, it just can make you extremely weak. I had mixed feelings about this, because I'm used to poison being an added source of HP damage, though I can see why they changed it.

The 3.5 / Pathfinder 1e method of applying Ability Score Damage does have its problems. Mostly just that if it doesn't do Con damage it can't kill you (if your Con hits zero it kills you), and because Lesser Restoration is a fairly trivial spell and it almost completely negates the impact of most poisons.

That's why I run my PF1e games using the alternative Afflictions rules that Paizo put out in Unchained. It'd be a real PITA to implement in the Infinity Engine, but probably possible by putting delays with separate saves on the periodic effects, each failed save stacking up some penalty. You'd need to add some functionality to various restorative spells to cleanse those penalties if the spell could have cured poison, though.

On 3/27/2024 at 10:31 PM, polytope said:

Planar prison also, but like the planar sphere (1st floor) that's actually a safe place to rest. So is most of Spellhold/underwater city and there's a gnome inn in the underdark. Mostly you're missing out on temples for that portion of the game which behooves you to keep a cleric alive at all times unless you've found a Rod of Resurrection (let's face it, most parties have). It varies by area, but I'm pretty sure the most dangerous creature you can see as a rest spawn in BG2 is a (single) gauth, most others being nuisances. Worse things might spawn in Watcher's Keep but in every level except the 3rd one it's simple to backtrack and find an exit. Similarly backtracking through the map along path's you've already taken is usually safe because the devs didn't implement wandering monsters (there are a few exceptions like the Firewine ruins in BG1).

Well, that would make it harder as the entirety of Spellhold and I guess parts of Siege of Dragonspear would need to be done without rest. SCS currently has an option to disable resting in the Illithid city of the underdark.

mmmmm I think wraiths and liches can spawn on random rest encounters on harder modes in the shadow temple ruins, for example. It's also level-dependent I think, but it's been a long time since I looked into how that's done. IIRC it's a table that compares your current level and current chapter of the plot. If you're high level because you've cleared a lot of content before finally going to Spellhold, or if you're medium level and you're coming back from the Underdark, the encounters can get pretty nasty (like liches).

And don't forget that they're extra dangerous because you're resting - so you're probably injured or running low on spells or both.

3 hours ago, polytope said:

In most instances, throughout this series, it was simply easier for the designers to overpower the encounters than to restrict the player character's frequency of resting in a way that's plausible and not immersion breaking. This means both random monster groupings and set encounters in IE games are scaled on the assumption of encountering a party near or at full strength in terms of their hp pool, memorized spells, x-uses per day items. Very few areas that you have to complete without leaving or resting.

I used to play on super hard modes because the random encounters triggered by resting in a dungeon were hard enough to be extremely dangerous especially if you needed the rest. But that was obnoxious because what it really did was just make me run back to town to sleep in an inn and then spend ten seconds walking back through the dungeon to get to where I was (the Planar Sphere and Spellhold to Underdark sequence is really the only time you can't do this), but with the added tedium of dealing with companions getting chunked sometimes and having to quickload if it chunks the Bhaalspawn. So then I just started playing on Core and just... stocking up on potions and scrolls because by then I had modded scrolls to use the user's caster level and things like that. It really flowed a lot better in some ways but it's not strictly a fair analysis since a veteran player knows the whole dungeon and just plans resources accordingly.

Isn't there a mod somewhere that actually removes resting outside taverns?

2nd edition D&D kind of has "gotcha" and "puzzle monster" mechanics as a core conceit. Bosses with low enough THAC0 to guarantee they hit against AC every time aren't really any more problematic than player characters who can become invincible for a turn or two at a time. Or dozens of Drow and Mindflayers who have like 90% chance of ignoring any spell cast at them. Or players who've played the game so many times and have access to Near Infinity so they know exactly what's around that corner already.

While learning Near Infinity, I've spent hundreds of hours and many playthroughs manually tweaking .cre files and whatnot to make BG2 conform closer to more modern design philosophy specifically to remove those kinds of obnoxious agency / counterplay -denial mechanics. But more seriously to the point, the design of puzzle bosses and gotcha gimmicks being problematic for gameplay / "fun" is more an argument for modifying those creatures first before rescaling the way ability scores work.

Personally I don't think a +4 in any particular number is gonna break anything. So I think the 3e style of Ability Scores is a superior design overall because it standardizes them. And one of the major benefits of standardized ability scores is exactly that it forms a more manageable, stable basis of comparison against which to balance encounters.

3 hours ago, subtledoctor said:

THAC0 is an amazingly useful shorthand and easy to understand and you will never convince me otherwise. 

As for AC, I used to concede to you children of summer that pre-3E AC was weird. But no longer! I recently learned the origin of it and it’s fascinating. Apparently the system stems from an old naval combat simulation where the best ship armor you could have was 1st-class, and the next-best armor was 2nd-class, etc., getting worse as you get further from 1st class. (Think about airline seats or cruise ship accommodations - we still use a system where lower is better and nobody complains that it is “hard to understand.”) The took that naval defense mechanic and plugged it into the d20 roll-over to-hit roll, and ended up with a working fantasy combat system! At a time when there was almost nothing to compare it to. 

I just can’t bring myself to be critical of that. 

If it helps, just remember that it's algebraically the same as +to-hit and +AC where higher is better. The children of summer are still learning their gradeschool math, though, so maybe we should keep the THAC0 just for the educational value... ;D

And because it'd be a huuuuuuge PITA to change the engine now!

15 hours ago, DavidW said:

I think you're misunderstanding how that system has evolved. The odd numbers aren't useless in 3e/3.5e - feat qualification requirements are usually odd-numbered, and it functions as a mechanic to ration ability score boosts - but they're certainly much less significant. But the logical conclusion isn't 'we need to invent something for odd-numbered values to do': it's 'we need to drop odd-numbered ability scores entirely'. You see this most clearly in recent versions of Pathfinder - what matters are the bonuses, with the base scores kept around only for nostalgia value, and indeed dropped entirely in the most recent version (one just talks about having Strength +4 or Intelligence -1).

Of course, there's a gap between abstract RPG design and what makes sense internal to the IE and its modding community. (It would be technically feasible to move to a pure + system, given UI editing, but it feels like a bridge too far. Though it might be interesting to code, just as an exercise.)

So giving all the evens and odds "something to do" could still be done just by writing a looped script to "squish" all the ability scores in .cre files. This could prevent both the issue of "needing an 18 to function" and also help to reduce numerical inflation at all stages of the game, while also rewarding every single increment of your ability scores. If everyone's in a rush to 18 just to feel like they can do their job, suddenly the hard ceiling of 25 feels very low and rigid. Like without any way of adjusting or customizing your ability scores in play (apart from like one or two items in the entire game that give you a +1 somewhere), it's a lost opportunity for players to feel like they have agency over their character's development. This way, with squashed stats, with every +1 you feel like it's meaningful to get an item that gives you that +1 but it also opens the possibility of modders to add in nifty trinkets and consumables that give a stat boost as one more way to expand content.

At that point, the only lingering problem would be bonus spell slots for spellcasters based on their ability scores. Since spell levels are on integers 1 - 9, you wouldn't be able to "squash" the table without just giving +1 slots to SL1 and SL2, then +1 to SL3 and SL4 and so forth. But all the mages are having their ability scores squashed along with everyone else, they shouldn't complain. When you get to the end of the loop, just double back to give an extra slot in the lowest two spell levels again or something.

On 3/24/2024 at 5:06 PM, subtledoctor said:

The obvious problem with 3E ability scores (and 4E, and 5E, and Pathfinder, etc.) is that you can boost an ability score and get no benefit from doing so. This seems so shockingly bad to me, it boggles my mind that WotC has stuck with the system all this time. The answer seems obvious: have every ability score do two things, and give a bonus to each one on alternating ability scores. So for STR, a damage bonus for every even score starting at 12, and a to-hit bonus on every odd score starting at 13. For DEX, an AC bonus for every even score starting at 12, and a ranged thac0 bonus for every odd score starting at 13. Et cetera.

My own mod's allocation of bonuses is quite idiosyncratic, due largely to the particulars of this game's campaign. But surely there is a way to be more systematic with stat bonuses without rendering every other point increase useless...

I've shared your complaints about the "only evens count" issue of 3.x's architecture for decades, but I quite like your solution to it. It helps to keep numerical inflation under control while also elegantly resolving the "dead spot" problem. For years now I've used a mod I found somewhere that used a constantly-refreshing effect + spell to give various bonuses based on current ability scores. I think you may have been the author! It's been my favorite "gameplay systems" mod for a while.

If it's true that David can implement a customizable formula option for users to tweak during / before install, I'm gonna use your scale for my first test (waiting on beta 9! ;D)

Area of Effect is defined by the projectile.

The projectile in SPWI507 (HOLD.pro) hits every (enemy) creature within (vanilla is 4 feet, don't know SR's value).

The effects of SPWI507 (Cast Spell) is applied to every creature hit by the projectile, casting SPWI507D on each target individually.

This particular spell has no primary/secondary targets, everyone is equal.

 

For the purpose of Spell Deflection, if the save were on the parent spell, the target would get to save against the spell before spell deflection was checked, saving them deflection charges if they were successful, which would be inconsistent with how single target spells work.

Oh! That solves a mystery. I suspected that the area of effect information must be encoded in the projectile, but I didn't know for sure - thanks for clearing that up!

As for Spell Deflection, I think the way that family of spells works in NWN1/2 is that the benefactor of the Spell Deflection wasn't supposed to be able to avoid area spells at all if he wasn't the direct specific target of the spell. Which I'm not even sure I would want to try doing with how limited my knowledge is of BG2EE's workings. I can see how you would want Hold Person to eat SD charges, but I'm not sure that something like Fireball or Cloudkill should. For that matter, does each "tick" of Cloudkill and Icestorm eat a charge? Or just once per casting? Because I think both Cloudkill and Icestorm (and for that matter other spells like Stinking Cloud and Call Lightning) cast a child spell once each tick within the area...

But I don't really understand what that has to do with where the spell's save is invoked. You mean that if the save were on the parent effect that it would be called before the projectile lands at its target location?

 

Are you sure that's how it would work? I thought it cast 507D on every target within the area?

That's at least within rules, as targeting individual creatures with similated area of effect effects is definitely NOT in the rules.

 

I'm not sure I really understand what you mean.

Your proposed solution to move saving throws out to the parent spell would not work well, I don't think. Consider your example of Hold Monster: it targets an individual but everyone near that individual is meant to save against the effect. You don't want the target's nearby associates to avoid the spell just because the target made his save, do you?

Are you sure that's how it would work? I thought it cast 507D on every target within the area?

Is 507D the .spl with the area defined?

For that matter, I didn't change the targeting field, so it looks like my version SR has either altered Hold Monster to only affect one creature regardless, or there's something else that's defining the area and number of targets of both spells.

Ugh... so it is exactly what I feared, then? Do you mean that the RNG only provides a single roll for a given frame or tick, and the designers are just sort of assuming that it will all happen so fast that it won't matter much?

I have another question if you can help:

I've noticed that the druid Avenger kit gets a few spells that other druids don't get. I think SR has swapped out Chaos (which my .tlk says they're supposed to get at spell level 5, but I think SR has removed it in favor of the higher level sphere of chaos -style spell) for Cone of Cold. And anyway, I wonder if there's an easier way to propagate the changes I make to the wizard version of Cone Of Cold. Can you call a wizard spell from within a priest/druid spell? Since both versions call a child spell which does the damage, can I alter the wizard version of the child spell and then have the druid parent spell just call that wizard child spell instead of its own? Will it properly detect the druid's caster level? Or will it screw up because caster levels are "typed"? Are they typed? The druid and wizard versions of cone of cold both reference the same .tlk description entry...

SPDR501 calls SPDR501D normally
SPWI503 calls both SPWI503D and SPWI312D normally

Have SPDR501 call SPWI503D and SPWI512D without creating new .spl files and tagging them as "druid" spells?

So, I was noodling around in DLTCEP the other day, correcting a lot of errors in various files that my latest mod list accumulated (just a few mods couldn't recognize others mods' items and spells and so their changes obviously didn't affect each other, but that's not important).

And I noticed something odd about how Spell Revisions (v4 beta 14) does its spells.

Area spells are usually formatted so that they just have an effect with opcode [146] (Cast Spell At Target), basically affecting all targets within the area with a second spell linked to within that effect.

This is where the potentially huge problem comes in: the save to avoid the effects of the spell is not here at this Cast Spell At Target effect, SRv4b14 apparently has put the save under the individual effects of the child spell called in the parent spell!

Here's an example of what I mean. Importantly, for this particular spell (SPWI507 which is still "Hold Monster" in this mod), I have already altered the effects in this screenshot so that the parent effect of SPWI507 allows the save and I've removed the save from the child spell's Hold effect (opcode 175). Whereas the original Spell Revisions version of the spell had the save placed within the child, rather than the parent.



This seems really really odd for a number of reasons:

For example with the Slow spell, each of its individual component effects allows a save... which means that a target could attempt a save against each effect - only suffering those effects which it failed to save against, right? So then, virtually all of SR's area spells aren't actually written correctly because targets within those areas are likely saving against at least some of the effect-lines of each spell! This can become quite important in some cases that self-cure on a delay.

A good example of this was SR's version of Shadow Door: with SR's Shadow Door, adjacent creatures are supposed to be allowed a save to avoid being sucked into a "mini-Maze" effect. The problem here is that a separate effect on a time-delay was allowing a save when it shouldn't.

/*
If you'll recall, Maze has a duration dependent upon the target's Intelligence score, with a range of a few d4 rounds. Shadow Door being a lower level spell with other effects it provides as its main benefit was supposed to auto-liberate Maze victims on the 4th round if their intelligence didn't let them escape sooner.
*/

So the targets which were supposed to be liberated on the 4th round (poor 9-Int Korgan!) were failing their saves and staying in the maze for longer than they should have (as if by the original much higher level spell Maze).

I have to wonder if there's something I'm missing here. Is it supposed to be like this for some reason? For example, does the engine auto-magically understand that all the effects within a header are supposed to share the same saving throw result?