cittran Posted September 24, 2020 Share Posted September 24, 2020 Apologies in advance for the ugly formatting with the nested spoilers. So I was messing around with the Corvas in the Simulacrum -- because I was testing out a build idea -- and I noticed that, despite being modded for a significant bonus of Corrosive and Heat damage, nearly all of the status procs I was seeing were Impact. The Corvas has two different damage types depending on the game mode, unlike every other Archgun. If used as a weapon in Archwing missions, Corvas deals a mixture of Impact and Heat damage, and this is reflected accurately in the arsenal stats for the Archgun section. However, if equipped in the Heavy Weapon slot, Corvas deals a mixture of Impact, Puncture, and Slash damage, but no Heat damage. (This is also reflected in the arsenal stats in that section, which shows the weighting as 80% Impact, 10% Puncture, 10% Slash.) Additionally, the firing pattern changes. (Which is not explicitly mentioned or explained in-game, but is mentioned on the Wiki page for Corvas.) As an Archgun, Corvas has a projectile similar to the Arca Plasmor -- an energy-blast projectile shotgun that has an AoE. As a Heavy Weapon, Corvas instead fires a spread of pellets (11, according to the arsenal), which, when I tested in the Simulacrum, seemed to be an accurate count. Now, being a player who wants to get to the bottom of stuff that is impeding my build, I did some further testing. Quite a bit, actually, as I also tested several other Archguns. Corvas is not unique in seeming to have an improper damage weighting being used for calculating status procs. (I can't actually tell the exact damage values in-mission for each sub-type of damage, so I don't know if this is because the arsenal doesn't reflect the actual damage values for the weapons, which are then being used in a correctly-weighted status proc formula, or if it's because they're using incorrect values only for calculating the status proc distribution weighting formula. Or possibly using incorrect values for both.) I tested all the archguns I could as Heavy Weapons via the Simulacrum (Ballroom Simulacrum, if that makes a difference), and tested them against a group of 20 lvl 170 Corrupted Heavy Gunners. I also tested using 1 Corrupted Ancient, because they protect allies from status procs by 'absorbing' them. It was very obvious that the procs they were being affected by were biased in a way that didn't match the ratio of the damage types. I was unable to test the following Archguns in Heavy Weapon mode, so I can't verify their proc weighting via the Simulacrum. Dual Decurion (no Gravimag installed) Imperator (no Gravimag installed) Phaedra (no Gravimag installed) Cortege (I don't have it) Kuva Ayanga (I don't have it) (Incidentally, being unable to use Archwings in the Simulacrum is a bit inconvenient. Is there any way we could get a second mode for the Simulacrum that lets us test AW builds in the future?) As for the ones I did test, here were the results: These Archguns seem to have correct weighting, and I didn't bother testing them outside the Simulacrum Cyngas appears to have the right proc weighting, though I'd recommend actual detailed testing to be sure about that. Grattler may have an incorrect weighting, but it's rather difficult to tell, because the AoE is so large, and the status chance is fairly low Imperator Vandal seems fine, as the procs trended towards the modded damage ratios. Larkspur seems fine, but I admit that one is difficult to be sure on b/c the beam goes everywhere and the launcher mode just yeets enemies, and the ragdoll makes reading their procs almost impossible. On a semi-related note, the "+ x% status" mods for primaries/secondaries/melee were all buffed from their previous gain of (I think +15%?) to +90%, but the Archgun and Archmelee ones are still at their previous values, +60%. Was this intentional, or did the arch-weapon mods just get missed? Mausolon seems fine These Archguns appear to have incorrect weighting, so I tested them on both Mariana, Earth (as a Heavy Weapon) and Montes, Venus (as an Archgun) Corvas is heavily biased towards Impact procs (almost exclusively so), despite having been modded as follows, with the listed modded damage values Spoiler Slot order is left-to-right, top-to-bottom Slot 1 - Primed Rubedo-Lined Barrel (max) Slot 2 - Hollowed Bullets (max) Slot 3 - Parallax Scope (max) Slot 4 - Corvas Toxi-lexitio -- +96.4% Electricity, +102.2% Toxin, +2.2 Punch Through (max) Slot 5 - Critical Focus (r4) Slot 6 - Empty Slot 7 - Empty Slot 8 - Combustion Rounds (max) Total Damage bonuses +187% Base Damage +198.6% Corrosive Damage +120% Heat Damage Modded Damage Values Archgun mode (base damage split between Impact and Heat; 45.45% (repeating) Impact, and 54.54% (repeating) Heat) Spoiler Base Damage is 400/880 Impact, and 480/880 Heat for a Charged Shot; half those values for a Quick Shot Modded Charged Damage Impact - 1148.0 Heat - 4408.3 Corrosive - 5016.9 Total - 10573.3 damage (10.8% Impact | 41.7% Heat | 47.5% Corrosive) Off by +0.1 damage, probably from rounding Modded Quick Shot Damage Impact - 574.0 Heat - 2204.2 Corrosive - 2508.5 Total - 5286.6 damage (same ratios) Off by -0.1 damage, probably from rounding Heavy Weapon mode Spoiler Base damage per pellet is split between Impact, Puncture, and Slash, with no innate Heat damage Impact - 64 Puncture - 8 Slash - 8 Total - 80 damage per pellet (80% Impact | 10% Puncture | 10% Slash) Half those values for a Quick Shot Base Damage per Charged shot is 880, split among 11 pellets Modded Charged Damage/Projectile Impact - 183.7 Puncture - 23.0 Slash - 23.0 Heat - 275.5 Corrosive - 456.1 Total per pellet - 961.3 damage (19% Impact | 2.4% Puncture | 2.4% Slash | 28.7% Heat | 47.4% Corrosive) Total - 10573.3 (same ratios) Modded Quick Shot Damage/Projectile Impact - 91.8 Puncture - 11.5 Slash - 11.5 Heat - 137.8 Corrosive - 228.0 Total per pellet - 480.6 damage (same ratios) Total - 5286.6 (same ratios) Fluctus is also visibly biased towards Impact procs, which is definitely not what I was expecting. If it were Slash procs, I'd be less surprised. I'd still call it a bug, but it would be one that made sense. Fluctus has a damage ratio of 20% Impact, 10% Puncture, and 70% Slash, so this one is pretty weird to me. Spoiler Slot order is left-to-right, top-to-bottom Slot 1 - Venomous Clip (max) Slot 2 - Contamination Casing (max) Slot 3 - Charged Bullets (max) Slot 4 - Magma Chamber (max) Slot 5 - Dual Rounds (max) Slot 6 - Primed Rubedo-Lined Barrel (max) Slot 7 - Electrified Barrel (max) Slot 8 - Zodiac Shred (max) Total Damage bonuses +187% Base Damage +360.0% Corrosive Damage +60% Heat Damage +90% Slash Damage Spoiler Modded Damage Values Archgun mode stats are identical to Heavy Weapon mode stats Heavy Weapon mode stats Base Damage per shot is 50/250 Impact, 25/250 Puncture, and 175/250 Slash Modded Damage/Projectile Impact - 143.5 Puncture - 71.8 Slash - 954.3 Heat - 430.5 Corrosive - 2583.0 Total - 4183.0 (3.4% Impact | 1.7% Puncture | 22.8% Slash | 10.3% Heat | 61.7% Corrosive) Total without Dual Rounds Total - 6692.8 (same ratios) Total with Dual Rounds Velocitus is also visibly biased towards Impact procs, which is also odd, though not as much as the apparent weighting for Fluctus. Yes, Velocitus deals Impact damage, but it also deals 3 other types of damage, and all 4 of them are exactly the same values. The proc weighting should be pretty close to even, though it being exactly even is unlikely because RNG. Instead, in my testing in the Simulacrum, and the two mission nodes, I saw very few Slash procs, a scant number of Puncture procs, a lot of Impact procs, and almost no Magnetic procs. Spoiler I actually tested Velocitus with two mod configs, because I wanted to see what would happen if I didn't modify the damage types at all -- the result was the same, with Impact procs being much more frequent Spoiler No change to damage types Slot order is left-to-right, top-to-bottom Slot 1 - Empty Slot 2 - Empty Slot 3 - Hollowed Bullets (max) Slot 4 - Shell Rush (max) Slot 5 - Empty Slot 6 - Parallax Scope (max) Slot 7 - Critical Focus (r4) Slot 8 - Primed Rubedo-Lined Barrel (max) Total Damage bonuses +187% Base Damage Spoiler Modded Damage Values Archgun mode stats are identical to Heavy Weapon mode stats Heavy Weapon mode Spoiler Base damage per shot is evenly split between Impact, Puncture, Slash, and Magnetic Spoiler Charged Shot Impact - 400 Puncture - 400 Slash - 400 Magnetic - 400 Total - 1600 damage per shot (25% Impact | 25% Puncture | 25% Slash | 25% Magnetic) Spoiler Quick Shot damage is only 37.5% of Charged Shot damage Impact - 150 Puncture - 150 Slash - 150 Magnetic - 150 Total - 600 damage per shot (same ratios) Spoiler Modded Damage per shot Spoiler Charged Damage per shot Impact - 1148 Puncture - 1148 Slash - 1148 Magnetic - 1148 Total - 4592 damage per shot (same ratios) Spoiler Quick Shot Damage per shot Impact - 430.5 Puncture - 430.5 Slash - 430.5 Magnetic - 430.5 Total - 1722 damage per shot (same ratios) Spoiler Modded for additional damage types Slot order is left-to-right, top-to-bottom Slot 1 - Primed Rubedo-Lined Barrel (max) Slot 2 - Dual Rounds (max) Slot 3 - Charged Bullets (max) Slot 4 - Hollowed Bullets (max) Slot 5 - Venomous Clip (max) Slot 6 - Parallax Scope (max) Slot 7 - Shell Rush (max) Slot 8 - Critical Focus (r4) Total Damage bonuses +187% Base Damage +180% Corrosive Damage Spoiler Modded Damage Values Archgun mode stats are identical to Heavy Weapon mode stats Heavy Weapon mode Modded Damage per shot Spoiler Charged Damage per shot Impact - 1148 Puncture - 1148 Slash - 1148 Magnetic - 1148 Corrosive - 8265.6 Total - 12857.6 damage per shot (8.9% Impact | 8.9% Puncture | 8.9% Slash | 8.9% Magnetic | 64.3% Corrosive) Without Dual Rounds Total - 20572.2 damage per shot (same ratios) With Dual Rounds Spoiler Quick Shot Damage per shot Impact - 430.5 Puncture - 430.5 Slash - 430.5 Magnetic - 430.5 Corrosive - 3099.6 Total - 4821.6 damage per shot (same ratios) Without Dual Rounds Total - 7714.6 damage per shot (same ratios) With Dual Rounds As it stands, I have no clue what might be causing the incorrect weighting on Corvas, Fluctus, and Velocitus, but that's the testing I can do with the tools I have available. Link to comment Share on other sites More sharing options...

KitMeHarder Posted September 24, 2020 Share Posted September 24, 2020 Link to comment Share on other sites More sharing options...

cittran Posted September 24, 2020 Author Share Posted September 24, 2020 I'll grant it's possible that Fluctus is behaving oddly due to the forced proc, and I'll also grant it's possible that both Corvas and Velocitus have forced procs due to their natures of being a giant pellet shotgun and literally a mass driver, but if those are intentional, they should be explicitly mentioned, and they are not. Additionally, those procs should be calculated separately from other status procs, and should be applied in addition to any procs from the status chance of the shot itself, rather than potentially being an either-or situation. I've not seen any damage numbers in my UI that have multiple proc symbols next to them, so I can't say whether or not this is the case. Link to comment Share on other sites More sharing options...

KitMeHarder Posted September 24, 2020 Share Posted September 24, 2020 31 minutes ago, cittran said: but if those are intentional, they should be explicitly mentioned, and they are not. Good luck asking a game developer for this. That's what the wiki is for, and if it's not there you add it (or at least comment it), as that's just how crowd-sourcing works. 33 minutes ago, cittran said: Additionally, those procs should be calculated separately from other status procs, and should be applied in addition to any procs from the status chance of the shot itself They are, but I'm not booting up the game to take screenshots. So either believe me, don't, test yourself, ect... Link to comment Share on other sites More sharing options...

cittran Posted September 25, 2020 Author Share Posted September 25, 2020 I *have* been testing myself, and while using less than 100% status chance, they didn't seem to be applied correctly. I will thank you for pointing out the forced procs, because it's allowed me to test with that in mind, and thus the ability to modify how many procs I should expect for each given shot. However, while testing using *only* Shell Rush, Modified Munitions, and Market Target, things still seemed odd. Velocitus has a base status chance of 25%, so with Modified Munitions (+60% status chance, so +15% total), every shot should have a 40% status chance -- and with Marked Target, all shots done while aiming should have a status chance of 70% (+120% status chance while aiming, so +30% total). This includes both charged shots, and uncharged shots. (Or it should, going by the wording of the mod, but I'm technically making an assumption here.) So, on unaimed, uncharged shots, not very many were procing, which is expected (I counted 14/30, which is probably within the margin of error at 46%). On aimed, uncharged shots, the procing was probably within the margin of error (I counted 45/70, about 64%). On unaimed, charged shots, the procing was probably within the margin of error (I counted 25/70, about 36%). On aimed, charged shots, it got funky. Then I tested it by adding Dual Rounds, which adds +60% multishot, and should allow the following scenarios: Uncharged, unaimed -- up to 2 procs from 2 rounds Only 40% chance for each round to proc, and only a 60% chance for 2 rounds, which I think works out to be a 9.6% chance of having 2 procs (please correct me if I'm wrong) I counted 6/70 proccing 2 status effects, which is 8.5% (margin of error) Uncharged, aimed -- again, up to 2 procs from 2 rounds Has a 70% chance for each round to proc, and a 60% chance for 2 rounds, which I think works out to be a 29.4% chance of having 2 procs I counted 20/70 proccing 2 status effects, which is 28.5% (margin of error) Charged, unaimed -- up to 4 procs from 2 rounds, 1 impact proc per round guaranteed Has a 100% chance for one impact proc from the first round, a 40% chance for round 1 to proc an additional status effect, a 60% chance for a second round (with a 100% chance for the 2nd round to trigger another impact proc), and a 40% chance for round 2 to proc an additional status effect. I'm not quite sure how to calculate the overall %'s for each proc type here, but I think the overall status chance for both rounds to proc their own additional status on the same shot should still be about 9.6% I counted 8/85 proccing 4 status effects, which is 9.4% (margin of error, assuming my math is correct) Charged, aimed -- up to 4 procs from 2 rounds, 1 impact proc per round guaranteed Has a 100% chance for one impact proc from the first round, a 70% chance for round 1 to proc an additional status effect, a 60% chance for a second round (with a 100% chance for the 2nd round to trigger another impact proc), and a 70% chance for round 2 to proc an additional status effect. I'm honestly not certain how to calculate the overall %'s for each proc type here, but I don't think the overall status chance for both rounds to proc their own additional status effects on the same shot should should be any different than the uncharged & aimed shots I made before. All that fully-charging the shot does is guarantee a forced impact proc on a round that hits. Going by that, the chance for 4 procs should still be 29.4% I counted 10/85 proccing 4 status effects, which is 11.7% (assuming my math is correct, this is not within the margin of error) Did 4 additional rounds of testing with this method, got the following counts Round 2 -- 6/85 (7.06%) Round 3 -- 10/85 (11.7%) Round 4 -- 6/85 (7.06%) Round 5 -- 9/85 (10.5%) In total, 41/425 shots procced 4 status effects, which is 9.6% (I'm sincerely hoping my math is wrong here, because if it's not that's definitely a bug) With the knowledge from later tests, I'm pretty sure this is the area of the bug I was noticing -- it doesn't seem to be taking Marked Target into account for Charged shots that are Aimed. The only one that seems like it could be somehow messed up from that set of tests would be the charged & aimed shots while both Dual Rounds and Marked Target are equipped, where my expected chance for the maximum number of procs on a shot was quite a bit higher than the actual number I got. Then, because that was odd, I did more testing by adding the 4x 60/60 dual-stat mods (without Dual Rounds this time) With the combination of 4x +60% status chance from the dualstats, another +60% from Modified Munitions, and +120% while aiming from Marked Target, my final chances should be 25% status chance (base) * (1 + (5*0.6)) = 25*4 = 100% status chance without aiming theoretical 130% status chance while aiming Tested without Dual Rounds Uncharged, unaimed -- exactly 1 proc from 1 round A 100% status chance means a guaranteed proc I counted 85/85 1-proc rounds (100%, matches expected) Uncharged, aimed -- up to 2 procs from 1 round Has a theoretical 130% chance for each round to proc which means a 70% chance for one proc, and a 30% chance for two procs I counted 22/85 2-proc rounds (25.9%, margin of error) Charged, unaimed -- exactly 2 procs from 1 round Has a 100% chance for one impact proc from the first round, and a 100% chance for an additional proc I counted 85/85 2-proc rounds (100%, matches expected) Charged, aimed -- 2-3 procs from 1 round, 1 impact proc per round guaranteed Has a 100% chance for one impact proc from the first round, a 100% chance for an additional proc, and a theoretical 30% chance for a 3rd proc I counted 0/85 3-proc rounds, and 85/85 2-proc rounds This means that Marked Target, statistically, did apply to the uncharged & aimed shots, but did not apply to the charged & aimed shots The fact that it's applied unevenly is strange. It's possible that Marked Target doesn't apply to Charged Shots at all, but I'll test that more with the next steps. Tested again with Dual Rounds (with modified expectations for aiming possibly not giving any benefit beyond 100% status chance) Uncharged, unaimed -- 1-2 procs from 1-2 rounds A 100% status chance means a guaranteed proc for each round, but there is only a 60% chance of an additional round, so there should be a 40% chance of 1 proc, and a 60% chance of 2 procs I counted 46/85 2-proc rounds (54%, margin of error) Uncharged, aimed --1-4 procs from 1-2 rounds (assuming >100% status chance, otherwise exactly the same as the previous Uncharged & Unaimed test) Has a theoretical 130% chance for each round to proc which means a 70% chance for only one proc, and a 30% chance for two procs, for each round 100% chance for a proc on 1st round, 30% chance for a 2nd proc on 1st round 60% chance for a 2nd round 100% chance for a proc on 2nd round, 30% chance for a 2nd proc on 2nd round If actually has only 100% status chance per round, then should be split roughly 40/60 between 1-proc rounds and 2-proc rounds I think this works out to be A 0.0% chance for 0-proc rounds A 40% chance for only 1 round, which has a guaranteed 1-proc, and a 30% chance for a 2-proc, so thus a 70% chance for the 1-proc A 28% for only 1 proc (0.4 * 0.7) A 12% chance for 2 procs from 1 round (0.4 * 0.3) A 60% chance for 2 rounds, both of which have a guaranteed 1-proc, and a 30% chance for a 2-proc, so thus a 70% chance for the 1-procs A 29.4% for 2 procs from 2 procs from 2 rounds (0.6 * 0.7 * 0.7) A 25.2% for 3 procs from 2 rounds (2 * (0.6 * 0.3 * 0.7)) A 12.6% for 3 procs from 2 rounds (1st round double-proc) (0.6 * 0.3 * 0.7) A 12.6% for 3 procs from 2 rounds (2nd round double-proc) (0.6 * 0.7 * 0.3) A 5.4% chance for 4 procs (assuming 130% status chance) Chance would be (0.6 * 0.3 * 0.3) Final totals = 100% 0.0% chance for 0-procs 28% chance for 1 proc 41.4% chance for 2-procs 12% + 29.4% 25.2% chance for 3 procs 5.4% chance for 4 procs 0.0% + 28% + 41.4% + 25.2% + 5.4% = exactly 100% If it only has a 100% status chance, this should be the same as the previous test, with a 40% chance of 1 proc, and a 60% chance of 2 procs Tests Round 1 0/85 0-proc rounds (0.0%) 21/85 1-proc rounds (24.7%) (close to what I calculated; about 11.8% less, relative to the estimated 28%) 40/85 2-proc rounds (47.05%) (close to what I calculated; about 13.6% more, relative to the estimated 41.4%) 23/85 3-proc rounds (27.05%) (close to what I calculated; about 7.3% more, relative to the estimated 25.2%) 1/85 4-proc round (1.18%) (much less than what I calculated; about 78.1% less, relative to the estimated 5.4%) Also, this might have actually be a 2-proc or 3-proc with lingering status effect(s) already on that target, but I'm not sure Round 2 0/85 0-proc rounds (0.0%) 28/85 1-proc rounds (32.9%) 27/85 2-proc rounds (31.7%) 21/85 3-proc rounds (24.7%) 9/85 4-proc rounds (10.6%) Round 3 0/85 0-proc rounds (0.0%) 23/85 1-proc rounds (27.05%) 34/85 2-proc rounds (40.0%) 22/85 3-proc rounds (25.9%) 6/85 4-proc rounds (7.05%) Round 4 0/85 0-proc rounds (0.0%) 22/85 1-proc rounds (25.9%) 32/85 2-proc rounds (37.6%) 24/85 3-proc rounds (28.2%) 7/85 4-proc rounds (8.2%) Round 5 0/85 0-proc rounds (0.0%) 17/85 1-proc rounds (20.0%) 34/85 2-proc rounds (40.0%) 27/85 3-proc rounds (31.7%) 7/85 4-proc rounds (8.2%) Average of 5 tests 0/425 0-proc rounds (0.0%) 111/425 1-proc rounds (26.1%) -- estimated 28.0% 167/425 2-proc rounds (39.3%) -- estimated 41.4% 117/425 3-proc rounds (27.5%) -- estimated 25.2% 30/425 4-proc rounds (7.1%) -- estimated 5.4% Charged, unaimed -- exactly 2 or 4 procs from 1 or 2 rounds Has a 100% chance for one impact proc from the first round, and a 100% chance for an additional proc, with a 60% chance of an additional round which also has a 100% chance for one impact proc and a 100% chance for an additional proc I think this works out to be a 40% chance of 2 procs, and a 60% chance of 4 procs I counted 49/85 4-proc rounds (57.6%) Charged, aimed -- 2-3 procs from 1 round, and 4-6 procs from 2 rounds Has a 100% chance for one impact proc from the first round, and a 130% chance for an additional proc, with a 60% chance of an additional round which also has a 100% chance for one impact proc and a 130% chance for an additional proc Has a theoretical bonus 130% chance for each round to proc (in addition to the forced Impact) which means a 70% chance for two procs, and a 30% chance for three procs, for each round 100% chance for one impact proc on 1st round, a 100% chance for an additional proc, and a 30% chance for a 3rd proc on 1st round 60% chance for a 2nd round 100% chance for one impact proc on 2nd round, a 100% for an additional proc on 2nd round, and a 30% chance for a 3rd proc on 2nd round I think this works out to be a 5.4% chance for 6 procs I counted 0/85 0-proc rounds (0.0%) 0/85 1-proc rounds (0.0%) 34/85 2-proc rounds (40.0%) 0/85 3-proc rounds (0.0%) /85 4-proc rounds (60.0%) 0/85 5-proc rounds (0.0%) 0/85 6-proc rounds (0.0%) Oddly enough, I got the exact calculated values for only the procs from the base 100% status chance on top of the forced proc, with 1 or 2 rounds. 40% chance for 1 round (with 1 forced, and 1 additional proc) 60% chance for 2 rounds (with 1 forced, and 1 additional proc) As best I can tell from my tests, Marked Target is applying most of the time, but not all the time. Specifically, it doesn't seem to be properly applied to Charged Shots that are Aimed. I'm not sure why, but both tests for Charged & Aimed shots got exactly what I'd expect from having 100% status chance, rather than the boosted 130% they should have if Marked Target is being properly applied. Link to comment Share on other sites More sharing options...

IamLoco Posted October 2, 2020 Share Posted October 2, 2020 This thread needs a big PUSH. I googled for "archguns impact proc" and it led me here. I was testing the FLUCTUS myself today, as I had built it specifically for the high Slash damage and was going to install it on my necramech while necros´ desecrate is up, to farm toroids in OV. What I discovered was very similar. Out of 10 status procs, 8-9 were impact and 1-2 were slash, although it SHOULD be the other way round. This test was done on purpose without any elemental mods and a status chance as high as I could go without 60/60 mods. I came to the conclusion that archguns might have a forced impact proc due to their nature of being big, heavy guns - but then this is completely misleading and not documented. To me it looks more like a faulty damage calculation than an intended behaviour. Link to comment Share on other sites More sharing options...

## Recommended Posts

## Archived

This topic is now archived and is closed to further replies.