Jump to content

SCS : dealing with mage Invisibility


Recommended Posts

That sounds like it would be even more problematic for the AI than it is for the player, which is already substantial when you're talking about how frustrating it is to cast a high level spell that completely misses its intended target because the target randomly decided to wander off in the middle of you casting the spell and you have no way to adjust where you targeted once you've started.

Edited by Bartimaeus
Link to comment

Just to reiterate: it is quite easy for me to make SR-specific tweaks to SCS AI, but I will need an explicit and clear consensus steer from the SR-user community as to what those tweaks should be. I don’t have a sufficiently clear understanding of how SR tweaks the anti magic system to design it myself.

Link to comment
13 hours ago, Bartimaeus said:

That sounds like it would be even more problematic for the AI than it is for the player, which is already substantial when you're talking about how frustrating it is to cast a high level spell that completely misses its intended target because the target randomly decided to wander off in the middle of you casting the spell and you have no way to adjust where you targeted once you've started.

Cone of Cold/Prismatic Spray etc. have too-slow projectiles by default, but it's fixible.

It wouldn't be my preferred solution though, waiving specialist restrictions for spells of level 3 and below (making those universally accessible) is, so that Conjurers have Detect Invisibility.

8 hours ago, DavidW said:

Just to reiterate: it is quite easy for me to make SR-specific tweaks to SCS AI, but I will need an explicit and clear consensus steer from the SR-user community as to what those tweaks should be. I don’t have a sufficiently clear understanding of how SR tweaks the anti magic system to design it myself.

It's really more of a SR-problem to fix though, because it is the Conjurer who's specifically disadvantaged against Improved Invisible enemies with SR, and this is true even without SCS.

All Conjurers in an SR game now have the same problem that a solo Conjurer would, but the solo at least doesn't have to share xp with n other party members.

Link to comment
5 hours ago, polytope said:

wouldn't be my preferred solution though, waiving specialist restrictions for spells of level 3 and below (making those universally accessible) is, so that Conjurers have Detect Invisibility.

I've published that option in a mod, so players have that option now. :beer:

As far as the question of what SCS should do, I'm not sure it matters much for Conjurers. @DavidW as I understand things, SCS makes all spell attacks able to target improved invisible people. But when SR is installed, SCS only makes this change for AI versions of the spells, not for players. So with SR+SCS, players must cast:

Detect Invisible -> spell attack -> other spells

... while AI casters may do that in either order:

Detect Invisible -> spell attack     -> other spells
spell attack     -> detect invisible -> other spells

So SCS can:

  1. SCS AI should conform to the SR system - do not allow spell attacks to target invisibility, and design scripts such that, whenever they want to cast a spell attack at an improved invisible target, they will first cast Detect Invisible or True Sight. (Or Glitterdust or Detect Illusions or Oracle or Faerie Fire or Invisibility Purge.)
  2. SCS should change SR spell attacks conform to the SCS system - allow spell attacks to target invisibility for both players and AI. After all this contradicts the vanilla spells-vs.-invisibility rules, but it was a choice SCS makes for good reasons. It would not offend me if SCS makes the same choice even when SR is present. (This would allow SR Conjurers to lob abjurations into mage fights, contributing to their allies... though opinions differ as to how important this is. More broadly, the difference would be small because it does not absolve anyone of the need to cast both a divination and a spell attack; it just liberalizes the order in which those two things may occur.)
  3. Keep things as are, give the AI more options than players have. I don't hear many complaints about it.

In the abstract, #1 would be ideal - that's what I would have requested a couple years ago. BUT, we have been talking about changing SR to allow spell attacks to target invisibility. I was against it at first but @Bartimaeus convinced me of its benefits... unfortunately I can't find the discussion now. At any rate, I don't want to tell you to support spell attacks not targeting invisible for SR's sake, and then turn around and change SR to set spell attacks to penetrate invisibility.

So we (anyone invested in SR) should finalize this decision for SR independently, before giving guidance as to SCS.

Link to comment

Strongly against the "specialists can use 3rd level or lower spells of their opposition school" idea as a solution: no real argument against it, I just don't like it. I would honestly favor eliminating opposition schools altogether before that.

I don't know if anyone else has had this thought, but I really am starting to think that the whole mechanic of improved invisibility preventing direct spellcasting targeting but not indirect spellcasting targeting or ability targeting or melee targeting or ranged/missile targeting but only just direct spellcasting targeting...is flat-out kind of dumb to begin with. Honestly, if we're talking about trying to find "least bad solutions", get rid of the mechanic by setting every spell to be able to target through improved invisibility en masse. I think it would solve the following issues:

1. The Non-Detection problem, which is a whole other thing that still needs to be resolved in of itself...but all spells being able to pierce through improved invisibility would mean that you can get rid of the "can see through improved invisibility" opcode and let Non-Detection truly protect the stealth/invisible/improved invisibility states without issue, further eliminating the AI's current advantage of always being able to see through invisibility when they have Detect Invisibility/True Seeing running even when the player can't.

2. Conjurers not having a way to pierce invisibility, which you don't need to fix if their general spellcasting can always pierce improved invisibility.

3. The AI, assuming it's aware of it a la SCS, would no longer ever need to worry about whether it needs to cast (or perhaps even memorize!) Detect Invisibility or True Seeing before going into its normal spellcasting against an improved invisible target.

4. Invisibility Purge is no longer hot garbage in comparison to Detect Invisibility (and actually would be stronger, as it probably should be at 3rd level spellcasting compared to Detect Invisibility's 2nd level); also, other spells like Glitterdust probably don't feel quite as bad to memorize anymore.

However, there are four issues of varying importance that I feel it would introduce: one, non-SCS AI would presumably not be aware it can use spellcasting against improved invisible creatures, allowing the player an advantage if they have SR installed but not SCS (how well does the vanilla AI and vanilla spell memorization handle dealing with improved invisibility in the first place?); two, any spells installed after SR would be unaffected (...unless it's put in a secondary component installed much later a la the other secondary components); three, improved invisibility gets marginally weaker (but IMO not really, because SCS is always having its spellcasters memorize invisibility-piercing spells, and this mechanic doesn't affect anyone but spellcasters); four, anti-illusory divination spells get weaker (but I'm not exactly sure how much - I feel like there's still good cause to memorize both Detect Invisibility and True Seeing even if they don't allow you to directly pierce improved invisibility, but...).

Link to comment
6 hours ago, subtledoctor said:

So SCS can:

  1. SCS AI should conform to the SR system - do not allow spell attacks to target invisibility, and design scripts such that, whenever they want to cast a spell attack at an improved invisible target, they will first cast Detect Invisible or True Sight. (Or Glitterdust or Detect Illusions or Oracle or Faerie Fire or Invisibility Purge.)
  2. SCS should change SR spell attacks conform to the SCS system - allow spell attacks to target invisibility for both players and AI. After all this contradicts the vanilla spells-vs.-invisibility rules, but it was a choice SCS makes for good reasons. It would not offend me if SCS makes the same choice even when SR is present. (This would allow SR Conjurers to lob abjurations into mage fights, contributing to their allies... though opinions differ as to how important this is. More broadly, the difference would be small because it does not absolve anyone of the need to cast both a divination and a spell attack; it just liberalizes the order in which those two things may occur.)
  3. Keep things as are, give the AI more options than players have. I don't hear many complaints about it.

This is helpful.

(1) isn't going to happen: handling antimagic attacks is really complicated and deeply embedded into SCS, and I've been clear since Demivrgvs that I wasn't going to write a whole new version for SR. (I would consider it if someone actually coded it for me, I suppose.)

(2) is reasonable but more complicated than it looks. SCS actually doesn't contradict the vanilla rules, because vanilla EE also lets antimagic spells bypass invisibility! Of course that's not a coincidence - Beamdog intentionally copied SCS, and liaised with me about it. On an EE install, SR is actually overwriting the original-game behavior. But it wouldn't be difficult to cancel that overwrite. 

Do people agree with SD that this would be a good idea? If so I'll implement it.

Link to comment
7 hours ago, DavidW said:

This is helpful.

(1) isn't going to happen: handling antimagic attacks is really complicated and deeply embedded into SCS, and I've been clear since Demivrgvs that I wasn't going to write a whole new version for SR. (I would consider it if someone actually coded it for me, I suppose.)

(2) is reasonable but more complicated than it looks. SCS actually doesn't contradict the vanilla rules, because vanilla EE also lets antimagic spells bypass invisibility! Of course that's not a coincidence - Beamdog intentionally copied SCS, and liaised with me about it. On an EE install, SR is actually overwriting the original-game behavior. But it wouldn't be difficult to cancel that overwrite. 

Do people agree with SD that this would be a good idea? If so I'll implement it.

Yes, but I would personally prefer it if SR took a more definitive stance on exactly what it's trying to accomplish (preferably in a fair and cohesive manner!) with regards to the combined issues of improved invisibility/Non-Detection/anti-invisibility spells/the anti-magic spell system: it shouldn't be SCS's responsibility to try to patch up the problem that SR is needlessly creating. But...in the face of no agreed upon way forward for SR, it would at least mean that SCS players can expect fairer play between SR and SCS until/if something is decided. Over the years, there have been a number of threads like this one as well as posts within the SRR thread complaining about how these systems interact, which is why I added the option of patching the "can penetrate improved invisibility" flag back to anti-magic spells for SRR, so I do think it is important that the situation be improved in some manner.

Edited by Bartimaeus
Link to comment
11 hours ago, Bartimaeus said:

However, there are four issues of varying importance that I feel it would introduce: one, non-SCS AI would presumably not be aware it can use spellcasting against improved invisible creatures, allowing the player an advantage if they have SR installed but not SCS (how well does the vanilla AI and vanilla spell memorization handle dealing with improved invisibility in the first place?)

If you don't want to give Conjurers Detect Invisibility you could instead make Improved Invisibility (and Shadow Door, Mislead etc.) into a superior Blur, i.e. strip the secondary opcode #20 type 1, leaving only regular invisibility and apply instead opcode #66 with a value of about 170 as the fade amount along with a +4 bonus to AC vs all weapon types and whatever saving throws you think pertinent, a mild aqua opcode #9 (colour pulse) would make it look more similar to current II.

On the plus side this would make II more useful against monsters that naturally see through plain Invisibility, and why not? It's supposed to be better.

On the negative side this would probably trigger the same problem as stripping opcode #16 from Haste as discussed in the other thread, the AI may pointlessly recast II if not truly under STATE_IMPROVEDINVISIBILITY, I think it will happen less frequently though.

My earlier suggestion about expanding the low level spell choices of specialists is indeed the most boring way to do it, but also the least likely to break things and the best fit with the existing game since NPCs have really never stuck to the rules about what can be memorized. C.f. druids totally casting Hold Person, high level evil clerics casting Creeping Doom etc.

15 hours ago, subtledoctor said:

I've published that option in a mod, so players have that option now. :beer:

Yes, but some form of fix really needs to be integrated into a basic install of Spell Revisions. @Graion Dilach published another alternative on the first page. We can't assume every player who downloads SR & SCS will know about these supplemental mods, I believe the underlying SR should be updated to ensure it functions properly rather than relying on users to familiarize themselves with 3rd part additions.

Edited by polytope
Link to comment

The real problem IMO is that the systematic changes aren't explained anywhere. Diffing the spell list documentation makes it really hard and easily overlooked to explain the logic changes for players let alone other modders to know, and by the time someone realizes some changes of SR are actually detrimental and doesn't play well with the assumptions given by the other mods or the players they're already deeep within a run.

I wrote my hacks only because I think m0rpheus562 brought the changes up on Discord at one point. I couldn't comprehend the logic changes on my own even.

TBH please don't get me involved in this further. I've lost all interest in SR at this point.

Edited by Graion Dilach
Link to comment
19 hours ago, Bartimaeus said:

don't know if anyone else has had this thought, … setting every spell to be able to target through improved invisibility en masse

That’s a fascinating idea. I kind of love it. But, it very much sounds like its own mod. Way out of scope for Spell Revisions. I would write it as a mod that is independent of SR. And, on the EE engine, apply something like IWD’s Evasion for a (~50%?) chance to avoid spells, in place of the rule about not being targeted. 

However, as you noted, the big problem with it is how AI scripts would work with it.

9 hours ago, Bartimaeus said:

it shouldn't be SCS's responsibility to try to patch up the problem that SR is needlessly creating

Nobody is asking for that. And SR doesn’t needlessly create any problems. Let’s remember that, as a piece of software, SR is much older than BGEE. We have made it compatible for our own benefit but its design did not - could not - contemplate EE issues. The non-EE engine had a simple rule that you cannot target semi-invisible Predators with spells, and SR maintained that rule but (IMO) improved how it works in play. SCS changed that for some spells, for reasons, but that was a decision specifically in the context of SCS being installed. SCS was deferential to SR in this regard, perhaps overly deferential, and the only question for SCS at this moment is whether it should reconsider that deference. Personally I think it should, and it sounds like most people agree on that limited point.

(I’m not convinced the SCS/EE change is broadly for the best. Especially with BG3 coming out there are a lot of new BG(2)EE players who are not fluent in 2E/BG systems. I could see a player approaching an enemy mage with II and MGOI, successfully casting Secret Word and thus being given the false impression that other spells will also work. But, OTOH, I think this is a rare possibility and it doesn’t keep me up at night.)

8 hours ago, polytope said:

Yes, but some form of fix really needs to be integrated into a basic install of Spell Revisions. [redacted] published another alternative on the first page. We can't assume every player who downloads SR & SCS will know about these supplemental mods, I believe the underlying SR should be updated to ensure it functions properly rather than relying on users to familiarize themselves with 3rd part additions.

That suggestion, while good, is completely outside the scope of SR. It’s not a necessary fix - it’s a cool idea that players might, or might not, like to apply alongside SR. Or even without SR!

I think people are missing the forest for the trees, a bit. 99% of people using this mod don’t have problems here. They want to cast spells at invisible enemies, but the vanilla game has taught them it is not possible. So they cast “Detect Invisible” and voilà, now they can. You have seen a handful of posts (five? six?) over the course of several years with complaints, and you see that as evidence of a systemic problems. But honestly, I am not convinced. A lot of the complaints are based on straight-up incorrect information or assumptions. In this very thread the first post talks about SCS changing Breach to AoE and the second post talks about SI:Div being problematic in SR. 

I’m also still not convinced Conjurers are a problem. In character generation you read “cannot cast divination spells” so you are fairly well on notice that dealing with invisibility is going to be an issue. 

I do think some adjustments could be made to optimize things and account for more edge cases: 

1. I would have Nondetection block divination attacks, and set it to be a specific protection rather than a spell protection. 

2. I would maybe give Glitterdust a bigger AoE and maybe allow Glitterdust specifically to target invisible enemies - generally make it a more functional and effect part of the invisibility “system.”

3. I would probably enable SR spell attacks to target invisible enemies. This would conform to what SCS players may expect, and to what new EE players may expect; and while it differs from the non-EE rules, Bartimaeus has convinced me it works well there. 

Edited by subtledoctor
Link to comment
1 hour ago, subtledoctor said:

(I’m not convinced the SCS/EE change is broadly for the best. Especially with BG3 coming out there are a lot of new BG(2)EE players who are not fluent in 2E/BG systems. I could see a player approaching an enemy mage with II and MGOI, successfully casting Secret Word and thus being given the false impression that other spells will also work. But, OTOH, I think this is a rare possibility and it doesn’t keep me up at night.)

The in-game documentation is fairly explicit. SCS updates the documentation of every spell that it tweaks. (Though unmodded BGEE does not, which is probably unwise - I assume the pain of doing a multilanguage tlk update dissuaded them.)

Link to comment
On 9/17/2023 at 10:54 AM, Salk said:

But was there not a valid reason for the original concept of making it impossible to cast spells at invisible enemies?

Yes there was. And that still stands (setting aside Bart’s idea for a mod changing it). SCS and the EEs specifically exempt spells with the MagicAttack secondary type, only. (Spell Thrust, Secret Word, Pierce Magic, etc.)

On 9/17/2023 at 10:38 AM, DavidW said:

The in-game documentation is fairly explicit. SCS updates the documentation of every spell that it tweaks

Not criticizing the change or the documentation. But I’ve been skimming the BG Reddit and there seem to be a lot of new players, whose only experience is 5E D&D and who seem often lost dealing with 2E rules. People on Reddit tell them “read the 300-page manual” and they often respond with some variation of “are you feckin’ serious??” For such players, even well-reasoned inconsistencies are potentially liable to trip them up. 

All that said, I am on record as supporting the SCS rule. 

Edited by subtledoctor
Link to comment

Join the conversation

You are posting as a guest. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...