38 réponses à ce sujet

[corey_russell]

This thread is about some wonky save issues we've seen with Blackraven and Grond0 for 1, but perhaps there are others.

 

I did some tests:

I cast more than 100 spells at Keldorn when his saves for spells was -1, which was exactly he needed to save for finger of death and web - he always saved. This was consistent.

 

I cast about 50 spells or so at Keldorn when his saves was -3 for spells - they were either symbol: stun or symbol: fear - in either case, -3 was exactly what he needed to save. In my tests at least, this ALWAYS worked - that is Keldorn never failed his save.

 

My test environment was vanilla BG 2, ToB + official ToB patch, no mods.

 

So far, I can't recreate the issues that happened to Grond0 and Blackraven.

 

I did my tests in my pocket plane (Ruzzel).

[Alesia_BH]

Interesting, Corey. So in your install everything is behaving as it should so far.

 

Can you try a spell without a save modifier? Grim Jim reported difference across modifier types.

 

Best,

 

A.

[Blackraven]

Thanks Corey for creating the thread. 

I've had Jan cast Sunfires next to Dagny. First I tested with a save vs spells of 0 for Dagny. She always saved, even at rolls of 1. I then decided to try with a save vs spells of 1, and again a roll of 1 sufficed.

 

I also tested a bit more with Glorydd. Strangely the combat log mentions saves vs spells while NearInfinity says resisting Mummies' Aura of Fear requires a save vs wands. When her save vs spell was 0 and her save vs wand -1, she saved this time (unlike earlier today) at a save roll of 1. When her save vs spell was 1 and her save vs wand was 0 (i.e. after unequipping the Helm of Balduran), she did get paralyzed by fear. 
The only thing that may have happened earlier today is what Grond0 mentioned, the game failing to register correctly Glorydd's saves scores for a brief moment. I checked immediately after she got held, but at that moment her saves were correctly displayed in her character record screen.

 

Would be nice if we could get the numbers clear. I don't mind accepting it if a save vs spell of 0 or even -1 is required to consistently save vs spells with no modifiers, but I do think the numbers ought to be clear. Numbers are important in this game. It's what people build strategies on. 
 

[ussnorway]

ime what you need to be clear on is what game you test/ play and what mods are involved... the basic game is solid but it's very easy to bugger up a tutu or ww game by adding your mods in the wrong order.

[Grond0]

I also tested a bit more with Glorydd. Strangely the combat log mentions saves vs spells while NearInfinity says resisting Mummies' Aura of Fear requires a save vs wands. When her save vs spell was 0 and her save vs wand -1, she saved this time (unlike earlier today) at a save roll of 1. When her save vs spell was 1 and her save vs wand was 0 (i.e. after unequipping the Helm of Balduran), she did get paralyzed by fear. 

The only thing that may have happened earlier today is what Grond0 mentioned, the game failing to register correctly Glorydd's saves scores for a brief moment. I checked immediately after she got held, but at that moment her saves were correctly displayed in her character record screen.

 

Would be nice if we could get the numbers clear. I don't mind accepting it if a save vs spell of 0 or even -1 is required to consistently save vs spells with no modifiers, but I do think the numbers ought to be clear. Numbers are important in this game. It's what people build strategies on. 

Interesting.  Did you confirm in NI that there is definitely no save penalty associated with the aura?

 

I don't think this is a mod issue - I played with vanilla for a long time and saw this type of problem a number of times.  However, I suspect you may not be able to recreate what is a sporadic problem with this type of test.  Not only is it a rare occurrence, but I think it's associated with the game not correctly adding up all save bonuses.  I'm sure you've all seen instances where behaviour is different depending on the order in which items are equipped (things like haste and magic resistance are regularly affected by this) and I think the saves issues is akin to this where the game 'loses track' of an item briefly.  That might be due to inventory changes or possibly just the nature of an attack / opponent.

[Alesia_BH]

Makes sense, Grond0. If these are black swan events then it may be difficult to pin them down in testing.

 

I can't comment on the plausibility of the proposed mechanism.

 

I'd be interested in hearing more from Grim Jim regarding his tests, and the conditions under which they were conducted..

 

Best,

 

A.

[corey_russell]

Interesting, Corey. So in your install everything is behaving as it should so far.

 

Can you try a spell without a save modifier? Grim Jim reported difference across modifier types.

 

Best,

 

A.

I have done the requested test that Alesia requested - once again used pocket plane, and Ruzzel save (his save is perfect with the one target with OK saves (Keldorn) plus rest mages).

 

I am pleased to report I have duplicated GrimJim's results perfectly:

 

1) Keldorn save vs spells at 1: with 50 casts of blindness, he eventually failed his save

2) Keldorn save vs spells at 0: with 75 casts of blindness, eventually failed his save

3) Keldorn save vs spells at -1: with 75 casts of blindness, he always saved.

 

While this could be coincidence, in both failures where you would expect success, the failed saves were near the end of the series.

 

I think from a practical point of view, players that are solo especially, should only expect automatic save vs. unmodified save penalties at -1 - if not they risk failing a save.

[Blackraven]

Interesting.  Did you confirm in NI that there is definitely no save penalty associated with the aura?

No save penalty, unless the Mummies have the Aura of Fear that belongs to Greater Mummies. There are two Aura of Fear spl files, one with no save penalty (according to aTweaks' readme intended for normal mummies) and one with a -3 penalty (for greater mummies).

[Alesia_BH]

 

I then modified umber hulk confusion to have a -4 save penalty and repeated the first test. As Corey predicted, -5 appears to be the magic number to guarantee success.

 

Walk me through the logic on this one. In Corey's test -3 was sufficient against -4s. 

[Alesia_BH]

I have done the requested test that Alesia requested - once again used pocket plane, and Ruzzel save (his save is perfect with the one target with OK saves (Keldorn) plus rest mages).

 

I am pleased to report I have duplicated GrimJim's results perfectly:

 

1) Keldorn save vs spells at 1: with 50 casts of blindness, he eventually failed his save

2) Keldorn save vs spells at 0: with 75 casts of blindness, eventually failed his save

3) Keldorn save vs spells at -1: with 75 casts of blindness, he always saved.

 

 

Interesting. Thanks Corey.

 

I'm not getting how this happens, but I'm starting to accept that it does happen. I'll try to replicate myself.

 

Best,

 

A.

[Alesia_BH]

Oops, I was thinking of his comment in the no reload thread rather than this one. In my tests against the modified umber hulk confusion, saves vs. spells of -3 and -4 were not enough.

 

Strange.

[corey_russell]

Strange.

Remember though, I only did about 50 castings - that might not be enough (e.g. maybe need 75 castings to get enough sample size). But I could only cast 6 spells with -4 modifier, so had to rest after every 6 tries - was a bit slow.

[Alesia_BH]

Remember though, I only did about 30 castings - that might not be enough (e.g. maybe need 75 castings to get enough sample size).

 

Yes. I agree that your sample size may not have been sufficient.

 

When I said "strange" I meant the entire phenomena, not just the -4s.

 

The wonky roll theory is funky and counter intuitive, yet consistent with empirical observations and controlled tests. In those cases, one goes with the data, once enough data have amassed. It seems like we're getting there.

 

I really need to replicate for myself before I'm 100% sold. I'll try soon.

 

Best,

 

A.

[corey_russell]

Alesia, how much longer until you do a test yourself? Or did you change your mind?

[Alesia_BH]

Alesia, how much longer until you do a test yourself? Or did you change your mind?

 

Oops! Sorry! I didn't realize people were waiting for me.

 

I'll do it soon. So you know, I believe you. I do. I've been assuming that your results (Grim/C_R) are accurate in my play through. I just want to be sure it works the same way in my game.

 

I'd encourage everyone else to err on the side of caution, too. By now we have combat evidence, a controlled test, and a replication. That's pretty compelling.

 

Best,

 

A.

[corey_russell]

I just thought of another real-life example. Not too long ago, I had a fighter with save vs. spells at 0 and he got stunned by the basic mindflayer in the Beholder layer in the Underdark - yet as far as I was aware, there is no save penalty for the basic flayer stun so that should not have happened by D&D rules. But now I know I need -1 to guarantee a save, based on both mine and Grims' test.

 

And I wasn't really waiting on you - was just wondering if you still intended on following through or did you forget.

[Alesia_BH]

I still intend to follow through. I just haven't found the time. I guess I don't relish the idea of casting Blindness 100 times. 

 

Best,

 

A.

[corey_russell]

Sequencers? Spell triggers? You might be able to cast quite a few blindnesses that way.

[Alesia_BH]

I'll probably just create a ring that grants, like, 200 L1s and permanent Improved Alacrity. 

 

Best,

 

A.

[Alesia_BH]

Btw. I did try to replicate. I wasn't able to observe an anomaly in 100 trials. It's possible that I need a larger same size. It's also possible that we have uncontrolled variables across our tests, including but not limited to the manner of save reduction.

 

Nonetheless, I'm inclined to trust the findings of others here. We have multiple replications and multiple sightings in the wild. That's enough for me to play it safe going forward.

[Alesia_BH]

Interesting. You're using the Mac version of the game, correct? I'm not sure if that would make a difference, but then again very little about this issue makes sense to me.

 

Yes. And I agree that the sheer strangeness of the phenomenon makes it hard to know what variables might be relevant.

 

I should collect a larger sample size before we conclude that there are any differences. If I still don't observe an anomaly, we can talk about test conditions in greater detail.

 

Best,

 

A.

[corey_russell]

Alesia, did you check the rolls? E.g., did you ever get the lowest roll? Maybe you got extremly unlucky and kept making your saving throw...is that how your characters make their saving throws, the game throws them a lucky rabbit foot? : )