FoH / Holy Bolt and -enemy magic resistance

atari wrote: 5 days ago

I've tested this in singleplayer first with hero editor items and now again with regular game items (Void,

Renewed Black Cleft,

Guardian's Light and

Entropy Locket) against demons and undeads. I can assure you that +Magic Skill Damage is working with

Holy Bolt. Maybe the numbers aren't correct but I needed 8-9 hits without those items and 5-6 with them against a

Zombie in

Blood Moor (and yes, I made sure that the skill level was the same in both cases). That sounds right for a 50% damage increase. Please test for yourself if you don't believe me.

Honestly, I don't know what to tell you. Your test provided a false positive, and repeating this just reinforces your belief that MSD was the reason. An Undead

Zombie in Bloor Moor with extremely low HP is the single worst target you could have chosen for this test.

If you test a Demon who actually has magic resistance and calculates magic then you'll be better off, but like I said it won't change a thing with +MSD.

Holy Bolt can double‑hit undead targets (FOH HB cannot), and that alone can create a stable “5–6 vs 8–9” Pattern even when MSD does nothing. The damage from HB is also not flat - it has a min/max as well, which will impact every attempt.

In the tgEXcalibur video - he originally had +MSD and -EMR gear on, and now after hours upon hours of testing and editing the video you'll see in the side-by-side comparison that the panel he created now only includes -EMR after he removed the old footage. Why do you think that is? There's no way a build that experiences a +50% boost to be completely abandoned and reverted back to standard gear after seeing no real change in damage. It just can't happen.. It would automatically be the BiS and be undeniably better in the eye test alone - yet his conclusions were it was much worse than the standard setup. The only thing that was incorrect was EMR, which does nothing for half of the targets but impacts the other half - and losing a plus skill for this when an extra skill could scale damage for both demons/undead would be more optimal in theory.

I've also tested it with

Fallen in

Blood Moor and the result was the same. +msd is working for

Holy Bolt (and doesn't work for the bolts from foh).

atari wrote: 4 days ago

I've also tested it with

Fallen in

Blood Moor and the result was the same. +msd is working for

Holy Bolt (and doesn't work for the bolts from foh).

Thanks for taking the time to actually test this and providing your results.

Knappogue wrote: 4 days ago

atari wrote: 4 days ago

I've also tested it with

Fallen in

Blood Moor and the result was the same. +msd is working for

Holy Bolt (and doesn't work for the bolts from foh).

Thanks for taking the time to actually test this and providing your results.

Thanks indeed @atari for all the testing! And to @D369 for the insights about the various mechanisms in play.

My problem with tgExcalibur's testing was that they compared the FoH

Bolts at different skill levels and thus base damage numbers, useful for comparing actual builds but less so when investigating mechanisms.

I can't say I've followed every turn in the discussion here lately. But currently it seems to me as if the regular HB (but not FoH HB) was updated to include MR/MSD modifiers for RotW, while still keeping the regular MR piercing ("as if MR=0") for undeads. (The latter is exemplified by the fact that

Achmel the Cursed and his minions, who are magic immune, melt to my Pally's

Holy Bolt s from main skill or FoH alike without any sunder or minus res gear - I just tested to re-confirm this.) This would also imply that the interpretation here is actually correct:

Knappogue wrote: 1 week ago

...
In in incredibly oversimplified process:

My original interpretation:

• Step 1: Look at monster Magic Resistance and treat it like it is 0% (regardless of actual value)
• Step 2: See that Magic Dmg is being applied by player
• Step 3: Apply xx -EMR (enemy monster res is now 0-xx).
• Step 4: Apply total damage

...

...but only since RotW. If my reasoning is correct, it might also be that the devs simply forgot FoH when making the adjustments to

Holy Bolt.

Knappogue wrote: 1 week ago

I always really do enjoy having conversations like this and having my mind changed.
Cheers guys!

Indeed!

atari wrote: 4 days ago

I've also tested it with

Fallen in

Blood Moor and the result was the same. +msd is working for

Holy Bolt (and doesn't work for the bolts from foh).

Is there any way you can variably change the amount of MSD you have equipped so you can repeat the test with different amounts per run and see any variance in life chunking? Perhaps make 4x Small charms and remove one per run for example.

Also, do you have any EMR equipped in those previous tests? If so, removing would be ideal.

If you can choose a target with high HP, so nothing in Act 1. Act 5 Frenzytaurs/Blood Lords or even Act 4 Venom Lords would be a second option.

I think game have same logic as for druid

Poison Creeper: damage is not affected by sunder or -enemy poison resist. Those pads/roots are summoned by Creeper , not by player. Similar to

Bolts from FoH which are not casted by Palladin personally.
By the way, Sunders and new Stone do work for ravens and Grizzly in my limited testing.

NaglFarfar wrote: 2 days ago

My problem with tgExcalibur's testing was that they compared the FoH

Bolts at different skill levels and thus base damage numbers, useful for comparing actual builds but less so when investigating mechanisms.

I think it's actually a positive tho that the skill levels were different? - as the EMR setup had lower FOH skills yet chunked the Demon targets more than the other side- so the results would have been even more dramatic with equal skills.

NaglFarfar wrote: 2 days ago

I can't say I've followed every turn in the discussion here lately. But currently it seems to me as if the regular HB (but not FoH HB) was updated to include MR/MSD modifiers for RotW, while still keeping the regular MR piercing ("as if MR=0") for undeads. (The latter is exemplified by the fact that

Achmel the Cursed and his minions, who are magic immune, melt to my Pally's

Holy Bolt s from main skill or FoH alike without any sunder or minus res gear - I just tested to re-confirm this.) This would also imply that the interpretation here is actually correct:

Undead will always have their Magic Resistance ignored from all Holy

Bolts no matter what source, and your test proved that. The Sunder wasn't required for you.

Knappogue wrote: 1 week ago

...
In in incredibly oversimplified process:

My original interpretation:

• Step 1: Look at monster Magic Resistance and treat it like it is 0% (regardless of actual value)
• Step 2: See that Magic Dmg is being applied by player
• Step 3: Apply xx -EMR (enemy monster res is now 0-xx).
• Step 4: Apply total damage

...

There are two individual damage pipelines for Demons and Undead. So, they'll have different steps before applying the final damage.

Step 1. will do the 0 for Undead only. EMR only modifies Magic Resist values that are checked, and Undead = Not Checked > 0%. Demons MR is checked and EMR can therefore enter the formula at Step 2.

As Demons don't get automatically reduced to zero like Undead do, you'll notice those Demons in Ubers are immune/highly resistant and are really hard to kill with HB while Achmel who's fully immune (Undead) just gets torn up in

Baal Waves. If there was no Undead bypass you'd be fighting him at like 90% resist or higher just like those Uber Demons.

NaglFarfar wrote: 2 days ago

...but only since RotW. If my reasoning is correct, it might also be that the devs simply forgot FoH when making the adjustments to

Holy Bolt.

Direct

Holy Bolt as a skill was created first and has its own unique flag on top, while FOH HB got added later and didn't get the same flag but they both bypass undead still - just through different mechanisms. So, it's fair to consider whether these differences might contribute to some weird behavior since they aren't identical at the system level.

But yeah there is some suss stuff going on with Paladin Magic. We're cracking the nucleus here tho and we'll have our answers locked, so thanks to every single poster all for your efforts here!

D369 wrote: 2 days ago

NaglFarfar wrote: 2 days ago

My problem with tgExcalibur's testing was that they compared the FoH

Bolts at different skill levels and thus base damage numbers, useful for comparing actual builds but less so when investigating mechanisms.

I think it's actually a positive tho that the skill levels were different? - as the EMR setup had lower FOH skills yet chunked the Demon targets more than the other side- so the results would have been even more dramatic with equal skills.

Right, I don't think I fully understood your interpretation from his testing - are you saying that demon targets are actually affected by -EMR/MSD modifiers from FoH-generated holy

Bolts? This seems too to go against @atari 's results. Or does it only affect the target of FoH, due to this target getting a "normal"

Holy Bolt in addition to the lightning? I found the footage too messy to really see a difference, but maybe it is there. If the results in tgExcalibur's testing are actually different I would agree that there is no problem (as long as we look for a difference only, and do not try to estimate its magnitude after the direction of the difference has been established).

Given a positive result (one variant turning out stronger than the other) the confounding of skill- and EMR/MSD related differences is not really a problem, because the different hypotheses give rise to different predictions as you say. Depending on which variant wins out in the test we can support one or the other hypothesis, but not both.

From tgExcalibur's interpretation "no difference for FoH" it is a bit of a problem though, because the numbers predict a difference regardless of whether -EMR/MSD works or not, and he sees no difference. So is it the case that EMR/MSD does not work and the skill level difference is to weak to provide a reliable signal, or is it the case that -EMR/MSD actually works but the signal is still too weak after subtracting the lost damage due to a lower skill level? This was my main concern with his testing.

NaglFarfar wrote: 2 days ago

I can't say I've followed every turn in the discussion here lately. But currently it seems to me as if the regular HB (but not FoH HB) was updated to include MR/MSD modifiers for RotW, while still keeping the regular MR piercing ("as if MR=0") for undeads. (The latter is exemplified by the fact that

Achmel the Cursed and his minions, who are magic immune, melt to my Pally's

Holy Bolt s from main skill or FoH alike without any sunder or minus res gear - I just tested to re-confirm this.) This would also imply that the interpretation here is actually correct:

Undead will always have their Magic Resistance ignored from all Holy

Bolts no matter what source, and your test proved that. The Sunder wasn't required for you.

Knappogue wrote: 1 week ago

...
In in incredibly oversimplified process:

My original interpretation:

• Step 1: Look at monster Magic Resistance and treat it like it is 0% (regardless of actual value)
• Step 2: See that Magic Dmg is being applied by player
• Step 3: Apply xx -EMR (enemy monster res is now 0-xx).
• Step 4: Apply total damage

...

There are two individual damage pipelines for Demons and Undead. So, they'll have different steps before applying the final damage.

Step 1. will do the 0 for Undead only. EMR only modifies Magic Resist values that are checked, and Undead = Not Checked > 0%. Demons MR is checked and EMR can therefore enter the formula at Step 2.

So if @atari 's results are correct there must be a hybrid going on for regular HBs, implying that undeads get reduced to 0 first but then also affected by -EMR (and +MSD). In line with @Knappogue 's first interpretation.

D369 wrote: 5 days ago

atari wrote: 5 days ago

I've tested this in singleplayer first with hero editor items and now again with regular game items (Void,

Renewed Black Cleft,

Guardian's Light and

Entropy Locket) against demons and undeads. I can assure you that +Magic Skill Damage is working with

Holy Bolt. Maybe the numbers aren't correct but I needed 8-9 hits without those items and 5-6 with them against a

Zombie in

Blood Moor (and yes, I made sure that the skill level was the same in both cases). That sounds right for a 50% damage increase. Please test for yourself if you don't believe me.

Honestly, I don't know what to tell you. Your test provided a false positive, and repeating this just reinforces your belief that MSD was the reason. An Undead

Zombie in Bloor Moor with extremely low HP is the single worst target you could have chosen for this test.
...

If testing by killing only one

Zombie once I agree it would be a risk for a false positive. But I don't think you need to kill that many to know whether -EMR or MSD makes a difference or not. Statistical power goes up very quickly with more attempts, everything else held equal, especially when going from a low number to begin with. I'm assuming @atari tested with several zombies and not only one?

D369 wrote: 2 days ago

But yeah there is some suss stuff going on with Paladin Magic. We're cracking the nucleus here tho and we'll have our answers locked, so thanks to every single poster all for your efforts here!

I agree, there are lots of things to disentangle here apart from the actual damage output!

dionisy_cn wrote: 2 days ago

I think game have same logic as for druid

Poison Creeper: damage is not affected by sunder or -enemy poison resist. Those pads/roots are summoned by Creeper , not by player. Similar to

Bolts from FoH which are not casted by Palladin personally.
By the way, Sunders and new Stone do work for ravens and Grizzly in my limited testing.

Agreed, this is a possbility regardless of whether it was intended or not in this case. One could imagine that the game originally was made so that the FoH-related

Holy Bolt s emanate from the target of the FoH rather than the player, and thus would have applied the target's own damage modifiers - which were of course irrelevant before -EMR and MSD was introduced into the game.

Lore-wise I also think one could make a case for

Holy Bolt s, regardless of source, being a separate type of "Divine" magic that is unrelated to other types of magic altogether (similar to how it is conceptualized in the DnD universe). Like if the

Holy Bolt s actually were cast by a divine force called upon by the paladin, and not the paladin themselves. Then I would have been totally OK with just keeping all

Holy Bolt variants outside the rest of the "magic" (non-elemental) damage framework, and not make -EMR or MSD apply at all. Also it is not like the FoH pally needs any help, even in comparison to the Warlock.

NaglFarfar wrote: 17 hours ago

Right, I don't think I fully understood your interpretation from his testing - are you saying that demon targets are actually affected by -EMR/MSD modifiers from FoH-generated holy

Bolts?

Demons are the ONLY things affected by EMR.

NaglFarfar wrote: 17 hours ago

Or does it only affect the target of FoH, due to this target getting a "normal"

Holy Bolt in addition to the lightning?

Holy

Bolts from FOH still hit the main target.

NaglFarfar wrote: 17 hours ago

From tgExcalibur's interpretation "no difference for FoH" it is a bit of a problem though

tgExcalbur didn't know the Demons were the only things affected by EMR and was focused on overall damage for all targets. After seeing nothing happened most of the time (because of Undead) he thought there was no discernable difference. What he doesn't notice is that there is massive chunking to the last wave in WSK Throne room compared to the non-EMR side. It lines up perfectly for about 40-45% more.

NaglFarfar wrote: 2 days ago

So if @atari 's results are correct there must be a hybrid going on for regular HBs, implying that undeads get reduced to 0 first but then also affected by -EMR (and +MSD). In line with @Knappogue 's first interpretation.

Undeads can never get reduced below 0.

If EMR was working on Undead it would have to compete with the rule that forces Undead targets to always read 0. It would never win this. The only thing we're looking at here is MSD having some odd behavior against them under Direct

Holy Bolt.

For every other target and FOH Holy

Bolts - MSD isn't active. In the original tgExcalibur video it was so blatantly obvious that it was doing absolutely nothing, and this is why it was removed after I mentioned MSD doesn't work.

NaglFarfar wrote: 2 days ago

If testing by killing only one

Zombie once I agree it would be a risk for a false positive. But I don't think you need to kill that many to know whether -EMR or MSD makes a difference or not. Statistical power goes up very quickly with more attempts, everything else held equal, especially when going from a low number to begin with. I'm assuming @atari tested with several zombies and not only one?

We need to stop talking about Zombies. These are the worst targets imaginable for a test like this. I've written some suggestions already.

NaglFarfar wrote: 17 hours ago

I'm assuming @atari tested with several zombies and not only one?

I've tested

Holy Bolt without anything, with +msd and with -emr against multiple undead and demon enemies in

Blood Moor and

Arcane Sanctuary with players 8 setting in singleplayer. For that I first used a hero editor paladin with modified small charms with +100%msd and -100%emr. Later I confirmed the results with a pala I made in maxroll's d2planner, which I gave uniques and rune-words with +msd and -emr.
In my testing both +msd and -emr worked with

Holy Bolt (the skill). Since

Holy Bolt ignore the magic resistance of undead I assume that the magic resistance of undead is set to 0 and then -emr is applied but I'm not 100% sure if that is how the game handles it. I can only say taht you are doing more damage with -emr than without.
For the bolts from

Fist of the Heavens I couldn't see any difference with either +msd or -emr.

But feel free to test it yourself. The more people confirm those results the better. It is not hard to make a char with items in hero editor or d2planner and test it in singleplayer.