You would get the same results if you tested Imoen vs. Jan Jansen.
Confirmed: Pitting Imoen against Jan, Imoen rolls between 3 and 22, Jan rolls between -1 and 18.
Best,
A.
@Alesia_BH exactly which version of Imoen were you using for this test? As I mentioned before, there is one version of Imoen that has the MAGESCHOOL_GENERALIST kit, that being the one from Chateau Irenicus, the one with Imoen's Belt. That would explain the results of your above test.
I think the cause and effect here is backwards. The reason I started investigating this in the first place was that my gnome illusionist/thief was failing saves that he shouldn't have and I/Ts are properly given the illusionist kit. The issue here is that Imoen and Anomen have NONE as their kit. You would get the same results if you tested Imoen vs. Jan Jansen.
@GrimJim I am not seeing this as the case. I tested with Jan and Imoen. I double checked to make sure Jan had the illusionist kit and Imoen had none kit. In this case Jan saves against Celestial Fury AND Blindness as I would expect him to. He rolls 1-20 against CF and 3-22 against Blindness. Not only that, but you are being extremely vague in your refutation of my results. You say that your I/T was failing saves they should have made, but you do not give any details on what and who they were trying to save against. Without proper explanation or evidence or details of your own experimentation I have a great deal of trouble accepting this, especially when it is contrary to the results I am getting from my experimentation. Indeed, I really have no clue what you are talking about... What exactly do you think is the problem with the NONE kit?
@Alesia_BH exactly which version of Imoen were you using for this test? As I mentioned before, there is one version of Imoen that has the MAGESCHOOL_GENERALIST kit, that being the one from Chateau Irenicus, the one with Imoen's Belt. That would explain the results of your above test.
I used the ToB Imoen, summoned via the Fate Spirit- the same Imoen used in the test with Nalia.
The Imoen/Nalia results were the same as the Imoen/Jan results, as per Grim Jim's hypothesis.
Odd... hmm... Perhaps this is due to our different versions... I believe based on some of your screenshots in the no-reload thread that you are using v1.3 while I have the latest version, 2.x (correct me if I am wrong).
I will try summoning Jan and Imoen with the fate spirits and see what happens.
Do try. The Imoen, Nalia, and Jan were all summoned through the Fate Spirit.
Their save v spells were all set to 1. Nalia and Jan both failed against Imoen's darts of stunning. Imoen rolled 22's against Nalia and Jan's darts of stunning.
If we posit that none provides a comparative advantage, non-mage kits and wild mages behave like none, and player created unkitted characters are granted the generalist kit, not none, we could, I believe, explain all the results we've observed so far, including the previously perplexing findings from monsters.
Best,
A.
(I'll note that thus far we have observed no differences between v1.3 and v2.X with respect to this behavior. I've been able to replicate everything others have observed in V2.X, and others have been able to replicate what I've observed in v1.3. Of course, further study may uncover difference.)
You say that your I/T was failing saves they should have made, but you do not give any details on what and who they were trying to save against.
I initially tested against umber hulks (udumber.cre) and their confusion gaze. My I/T's save vs. spells was 1 and confusion gaze has no penalty, yet he was occasionally failing the save. In my install, umber hulks have NONE as their kit and confusion gaze has no school. In any case, the MAGESCHOOL_GENERALIST kit was not involved at all.
What exactly do you think is the problem with the NONE kit?
I think the issue is that certain kits, including NONE, are considered to be specialists in school-less spells and effects, including on-hit weapon effects.
I double checked to make sure Jan had the illusionist kit and Imoen had none kit. In this case Jan saves against Celestial Fury AND Blindness as I would expect him to. He rolls 1-20 against CF and 3-22 against Blindness.
I don't have a good explanation for this, but everything else I've seen in discussions about this topic is consistent with my hypothesis, so I suspect that it's some subtle difference in installs that we're not picking up on.
We should seek replication of Tresset's Jan/Imoen results.
We should try to ascertain whether there is, in fact, a difference between Tesset's install and others. If there is, we should try to delimit the scope of those differences.
My tests were done on jan12.cre and imoen10.cre on a clean install, using the console to spawn them in and Ctrl+Q to add them to the party. I confirmed in NearInfinity that that version of Jan is an illusionist and that version of Imoen is a NONE. My results were the same as Alesia's where Jan needed a -1 to avoid being stunned by Imoen's Celestial Fury.
Just tried Imoen vs Jan via Fate Spirit in unmodded v2.3.67.3: - Imoen casting blindness at Jan results in saves of 3-22 as expected (Imoen gets no bonus for attacking, but Jan gets specialist bonus for defending). - Imoen attacking with Celestial Fury results in saves of -1-18, which is consistent with my earlier results (though not Tresset's). In this case Imoen is getting the bonus apparently linked to the 'None' field for kits for school-less weapons, but Jan is not gettting a defensive bonus due to his kit.
The other way round is consistent with that: - Jan casting blindness results in saves of -1-18 (note I had to add blindness to Jan's spells - some illusionist he is ). - Jan attacking with Celestial Fury results in saves of 3-22.
I'm not great at saying thank yous to people, but I'll make the point of doing so here to everyone helping to look into this. I appreciate finding the source of something a bit obscure can be irritating, but think of the warm glow of satisfaction when a long-standing problem is finally nailed down!
I double checked to make sure Jan had the illusionist kit and Imoen had none kit. In this case Jan saves against Celestial Fury AND Blindness as I would expect him to. He rolls 1-20 against CF and 3-22 against Blindness.
I don't have a good explanation for this, but everything else I've seen in discussions about this topic is consistent with my hypothesis, so I suspect that it's some subtle difference in installs that we're not picking up on.
Well, I found one... I re-ran this test and I appear to had been in error earlier. He is rolling a -1 to 18 in reality. Sorry about the misleading info.
There is some weirdness going on here, for sure, and it is really frustrating me. I am beginning to wonder if specialists receive a penalty to their saves against any school other than their specialty. My testing seems to indicate something to this effect, but I am getting too tired to think anymore.
By the way @GrimJim, I appreciate the details. It has helped shed some light on this.
There is some weirdness going on here, for sure, and it is really frustrating me. I am beginning to wonder if specialists receive a penalty to their saves against any school other than their specialty.
I don't think so. Using Imoen to throw chromatic orbs against Jan there should be no specialist effects applying to either, so with the +6 bonus spell saves should vary from 7-26 - and they do.
If we posit that none provides a comparative advantage, non-mage kits and wild mages behave like none, and player created unkitted characters are granted the generalist kit, not none, we could, I believe, explain all the results we've observed so far, including the previously perplexing findings from monsters.
I think the issue is that certain kits, including NONE, are considered to be specialists in school-less spells and effects, including on-hit weapon effects.
The specialist bonus was implemented by looking for matches between the spell effect applied and the specialty school of the relevant character without excluding what we'll call the null case, a match between none and no school. Later, when the non-mage kits were added, and later still when the wild mage kit was added, they were coded in such a way that they behave like none for specialty school checks: they all gain a bonus when using unschooled weapons or spells. In contrast, when a specialist mage or character coded generalist uses unschooled weapons or spells, they do not gain a bonus: there is no advantageous match between their school and the schoolless effect. This leads to the following behavior with unschooled spells and weapons:
1) None v Generalist, advantage None 2) None v Mage Kit other than Wild Mage, advantage None 3) None v Non-Mage kit or Wild Mage, even 4) Mage Kit other than Wild Mage kit v Generalist, even 5) Mage Kit other than Wild Mage kit v Non-Mage kit or Wild Mage, advantage Non-Mage kit or Wild Mage 6) Non-Mage kit or Wild Mage v Generalist, advantage Non-Mage kit or Wild Mage
Letting "Schoolless"= None + Wild Mage + Non-Mage Kit, and "Schooled"=Generalist + Mage Kit other than Wild Mage, then we can reduce to:
Schoolless v Schoolled, advantage Schoolless
Put another way, if it's not one of the original mage kits or a generalist, the engine treats it as a schoolless specialist, and confers bonuses with unschooled weapons and spells
Note that your observation concerning the distinction between "generalist" and "none" was absolutely essential in teasing this out. It was the behavior of my unkitted fighter paired with her berserker twin that had led us to abandon the null case theory of the anomaly. We now understand that the behavior of my fighter and her twin is consistent with the null case theory: my fighter was flagged generalist, not none, and berserkers behave like none, not a specialist mage.
In theory, it should be possible to write a WeiDU routine that patches all CREs using "None" to use "Generalist" instead--as I reckon that would solve the issue for NPCs and monsters alike?
Granted, an engine fix would be ideal but the WeiDU fix would help kill the wait
In theory, it should be possible to write a WeiDU routine that patches all CREs using "None" to use "Generalist" instead--as I reckon that would solve the issue for NPCs and monsters alike?
The problem with that solution is that the other members of the "Schoolless" category, notably kitted warriors and such, would retain their comparative advantage. In fact, it would be extended, since they'd face more generalists
I'd rather see ToBex and EE deal with this on an engine level.
Agreed; this takes an engine fix (although I'm not sure Asc64 is still around to add fixes to ToBEx?). Hopefully a fix will be implemented in an upcoming patch for the EEs. Looks like a temporary WeiDU fix would be more awkward than I had thought.
So, if I'm understanding the null case bug discussion correctly, in the case of applying save-or-else weapons, an unkitted Bard would do better with bolts of polymorphing than a Jester of the same level would?
Others have said what it actually is. The actual term came from a UK TV program called Jim'll Fix it! The host was a man called Jimmy Saville who fixed problems for people who needed help. Unfortunately he mis-used his position to satisfy his own desires with kids. It only became public after his death due to cover-ups.
@Serg_BlackStrider So a couple of your save vs speels are very high still (for Mazzy and Viconia), does this get problematic? I feel like that would make me nervous in a no reload setting.
What do you do for enemies that require specific protections like vampires, beholders or mindflayers? NPP lasts 4 rounds -- do you just keep unprotected party members far away?
Also, you're far braver than I am, I'd be terrified to do the watcher's keep maze -- the dead magic zone to Tanaari fight has caused me more headaches than most other fights in the game.
As a fighter Mazzy could boost her saves significantly with Potions of Invulnerability if needed (those aren't too rare) and Viconia has 80 Magic Resistance unbuffed so I see no problem there.
Aye, in general, I prefer to fight at range and only have one or two well protected characters at close range and those aren't a pure fighters. Vs Vampires we have Helmets of Charm Protection, Amulet of Power and Improved Mace of Disruption. Beholders are more tricky. I don't use the Shield of Balduran (with SCS it's not much sense with it anyway) so mostly I sent one mage character polymorphed into Mustard Jelly and blast them with wands. Mindflayers need some kind of protection from their stun attacks but overall keeping at range or having good enough AC and a stock of Potions of Genius works fine. In my setup I use a Polytweak mod and with it their attacks don't drain Int anymore (draining 25% of health with each successful hit instead) but that's another story.
Wow, you guys and gals have the Infinity Engine literally down to a science. I feel like I'm reading a discussion by the research faculty of a "Department of Baldur's Gate" in a university.
Put another way, if it's not one of the original mage kits or a generalist, the engine treats it as a schoolless specialist, and confers bonuses with unschooled weapons and spells
Here's a way to think of it. Picture the devs, back in the BG1 days, seeking to implement a specialist bonus, but over-looking the null case. Unaware of the error, they then compounded the problem every-time they added a kit to the game.
Wow, you guys and gals have the Infinity Engine literally down to a science. I feel like I'm reading a discussion by the research faculty of a "Department of Baldur's Gate" in a university.
I think this is just another example of the cultural differences between our communities.
Here's how we've historically seen this and related issues.
If a behavior is unintended, but well know and long-standing, we prefer to treat it as a mechanic, rather than a bug, for bug reload purposes. We do this in the interest of minimizing avenues for players to justify reloads and avoiding immersion breaking experiences. Take the save anomaly issue, for example. It's unintended -unequivocally- but I would never consider allowing a reload if it ended my game. I can think of many cases where we've lost characters to behavior that we knew was unintended, and preferred to accept the dynamic as is, rather than invoke the bug reload rule. For us, invocation of the rule has been -and should be- extremely rare.
The SCS component is titled Make Party Members Less Likely to Die Irreversibly. It achieves that, to a certain extent, in my experience, at least with regards to physical damage within a certain range (ie- somewhat beyond the original -10). We accept the dynamic as it is, while acknowledging that the component may achieve less than it was intended to, or than we might like it to. I don't think any of the Bioware regulars would even consider reloading if it failed to protect one of our characters in a particular encounter.
If we were unhappy with the way the component behaves, we'd fix it or mod it (call it what you will) -preferably before starting a run- and declare the fix or mod in our introduction post. Negotiating around the component's limitations by invoking the bug reload rule would not be a viable option, in our view.
Reasonable people can disagree on this. We've long known that the Bioware community has maintained different, stricter standards on these issues.
You should play as you see fit. This is not the Bioware community exclusively. It is our community- legacy Bioware and Beamdog alike. We can accommodate a diversity of perspectives, I'm sure.
Comments
The Imoen/Nalia results were the same as the Imoen/Jan results, as per Grim Jim's hypothesis.
Best,
A.
I will try summoning Jan and Imoen with the fate spirits and see what happens.
Their save v spells were all set to 1. Nalia and Jan both failed against Imoen's darts of stunning. Imoen rolled 22's against Nalia and Jan's darts of stunning.
If we posit that none provides a comparative advantage, non-mage kits and wild mages behave like none, and player created unkitted characters are granted the generalist kit, not none, we could, I believe, explain all the results we've observed so far, including the previously perplexing findings from monsters.
Best,
A.
(I'll note that thus far we have observed no differences between v1.3 and v2.X with respect to this behavior. I've been able to replicate everything others have observed in V2.X, and others have been able to replicate what I've observed in v1.3. Of course, further study may uncover difference.)
We should try to ascertain whether there is, in fact, a difference between Tesset's install and others. If there is, we should try to delimit the scope of those differences.
Best,
A.
- Imoen casting blindness at Jan results in saves of 3-22 as expected (Imoen gets no bonus for attacking, but Jan gets specialist bonus for defending).
- Imoen attacking with Celestial Fury results in saves of -1-18, which is consistent with my earlier results (though not Tresset's). In this case Imoen is getting the bonus apparently linked to the 'None' field for kits for school-less weapons, but Jan is not gettting a defensive bonus due to his kit.
The other way round is consistent with that:
- Jan casting blindness results in saves of -1-18 (note I had to add blindness to Jan's spells - some illusionist he is
- Jan attacking with Celestial Fury results in saves of 3-22.
I'm not great at saying thank yous to people, but I'll make the point of doing so here
The question now is whether Tresset can replicate our findings, upon further study.
If so, it seems to me that we have an internally consistent, generalizable working theory.
Best,
A.
There is some weirdness going on here, for sure, and it is really frustrating me. I am beginning to wonder if specialists receive a penalty to their saves against any school other than their specialty. My testing seems to indicate something to this effect, but I am getting too tired to think anymore.
By the way @GrimJim, I appreciate the details. It has helped shed some light on this.
I will get to the bottom of this... Somehow...
1) None v Generalist, advantage None
2) None v Mage Kit other than Wild Mage, advantage None
3) None v Non-Mage kit or Wild Mage, even
4) Mage Kit other than Wild Mage kit v Generalist, even
5) Mage Kit other than Wild Mage kit v Non-Mage kit or Wild Mage, advantage Non-Mage kit or Wild Mage
6) Non-Mage kit or Wild Mage v Generalist, advantage Non-Mage kit or Wild Mage
Letting "Schoolless"= None + Wild Mage + Non-Mage Kit, and "Schooled"=Generalist + Mage Kit other than Wild Mage, then we can reduce to:
Schoolless v Schoolled, advantage Schoolless
Put another way, if it's not one of the original mage kits or a generalist, the engine treats it as a schoolless specialist, and confers bonuses with unschooled weapons and spells
Note that your observation concerning the distinction between "generalist" and "none" was absolutely essential in teasing this out. It was the behavior of my unkitted fighter paired with her berserker twin that had led us to abandon the null case theory of the anomaly. We now understand that the behavior of my fighter and her twin is consistent with the null case theory: my fighter was flagged generalist, not none, and berserkers behave like none, not a specialist mage.
Best,
A.
Granted, an engine fix would be ideal but the WeiDU fix would help kill the wait
1) The SoA cap in place
2) Invisibility detection and a movement rate bonus for Jonicus
Best,
A.
I'd rather see ToBex and EE deal with this on an engine level.
Best,
A.
The host was a man called Jimmy Saville who fixed problems for people who needed help.
Unfortunately he mis-used his position to satisfy his own desires with kids.
It only became public after his death due to cover-ups.
Aye, in general, I prefer to fight at range and only have one or two well protected characters at close range and those aren't a pure fighters. Vs Vampires we have Helmets of Charm Protection, Amulet of Power and Improved Mace of Disruption. Beholders are more tricky. I don't use the Shield of Balduran (with SCS it's not much sense with it anyway) so mostly I sent one mage character polymorphed into Mustard Jelly and blast them with wands. Mindflayers need some kind of protection from their stun attacks but overall keeping at range or having good enough AC and a stock of Potions of Genius works fine. In my setup I use a Polytweak mod and with it their attacks don't drain Int anymore (draining 25% of health with each successful hit instead) but that's another story.
It all boils down to this, Pokota.
Put another way, if it's not one of the original mage kits or a generalist, the engine treats it as a schoolless specialist, and confers bonuses with unschooled weapons and spells
Here's a way to think of it. Picture the devs, back in the BG1 days, seeking to implement a specialist bonus, but over-looking the null case. Unaware of the error, they then compounded the problem every-time they added a kit to the game.
It effects None and all the non-original kits.
Best,
A.
The mod name, JimFix, is derived from the name of the author, GrimJim.
Best,
A.
I think this is just another example of the cultural differences between our communities.
Here's how we've historically seen this and related issues.
If a behavior is unintended, but well know and long-standing, we prefer to treat it as a mechanic, rather than a bug, for bug reload purposes. We do this in the interest of minimizing avenues for players to justify reloads and avoiding immersion breaking experiences. Take the save anomaly issue, for example. It's unintended -unequivocally- but I would never consider allowing a reload if it ended my game. I can think of many cases where we've lost characters to behavior that we knew was unintended, and preferred to accept the dynamic as is, rather than invoke the bug reload rule. For us, invocation of the rule has been -and should be- extremely rare.
The SCS component is titled Make Party Members Less Likely to Die Irreversibly. It achieves that, to a certain extent, in my experience, at least with regards to physical damage within a certain range (ie- somewhat beyond the original -10). We accept the dynamic as it is, while acknowledging that the component may achieve less than it was intended to, or than we might like it to. I don't think any of the Bioware regulars would even consider reloading if it failed to protect one of our characters in a particular encounter.
If we were unhappy with the way the component behaves, we'd fix it or mod it (call it what you will) -preferably before starting a run- and declare the fix or mod in our introduction post. Negotiating around the component's limitations by invoking the bug reload rule would not be a viable option, in our view.
Reasonable people can disagree on this. We've long known that the Bioware community has maintained different, stricter standards on these issues.
You should play as you see fit. This is not the Bioware community exclusively. It is our community- legacy Bioware and Beamdog alike. We can accommodate a diversity of perspectives, I'm sure.
Best,
A.