Item Search
     
BG-Wiki Search
+ Reply to Thread
Page 5 of 15 FirstFirst ... 3 4 5 6 7 ... LastLast
Results 81 to 100 of 294

Thread: Weapon Skills     submit to reddit submit to twitter

  1. #81
    Custom Title
    Join Date
    Nov 2008
    Posts
    1,066
    BG Level
    6
    FFXI Server
    Diabolos

    <3 Nameless. That came up elsewhere, though Entropy was the subject if you think you could give it a similar test, and turns out to be bunk for Resolution at least. I think in the Entropy case people just forget they're still WSing in Twilight body and Bale legs, which alone are the whole of a level 75's accuracy gear for meriting.

  2. #82
    E. Body
    Join Date
    Dec 2008
    Posts
    2,408
    BG Level
    7
    FFXI Server
    Quetzalcoatl

    What would be a good place/method to test the potency of the -acc down effect on Externator?

    It definitely seems to be exceptionally strong eyeballing it, as in dynamis using my tping gear I'll go from "going to die quickly if I don't cure myself" to "you even aiming brah?" after using externator. To the point I literally have gone back to level 5 dancer style, in that Drain samba III + tping gear + Saber Dance keeps me almost permenantly at full health.

    Just eyeballing, it looks like I'm going from 50%ish eva all the way to capped after popping Externator. But that's eyeballing and I have 0 idea as to what the values really are.

    So, what do I go get punched by and what would be the best way to parse the eva values before/after externator to get an accurate reading on the potency? I can post the numbers here and let you guys go to town on mathing out what the exact value of the weaponskill is, because I'm super curious.

    I have 99 bard and 99 dancer, with a level 85 white mage mule I can use to heal myself. The mule doesn't have access to Hero's areas yet though

  3. #83
    Ridill
    Join Date
    Apr 2011
    Posts
    23,736
    BG Level
    10
    FFXI Server
    Bahamut

    Easiest way would be to get a limule to transfer it to you. No parsing required. That being said beleive someone already did that and came up with 20

  4. #84
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    Yeah, it was -20 Acc on the test server.
    https://www.bg-wiki.com/bg/Exenterator

  5. #85
    Custom Title
    Join Date
    Nov 2008
    Posts
    1,066
    BG Level
    6
    FFXI Server
    Diabolos

    Oh oh oh! Side question!

    Quote Originally Posted by Nameless View Post
    normal hit accuracy was 55% throughout the test, no gear change during ws.

    [...] all subsequent hit accuracy were 62% (using total subsequent hits/(43x4)).
    I was gonna ask if Belt/Gorget was giving you that 7% accuracy, but then I remembered you weren't swapping gear at all.

    Can you do this again with Belt/Gorget by chance to give a decent idea of their accuracy bonus? Would be something very nice to pin down since they're practically mandatory for these shared-fTP WS.

  6. #86
    Old Merits
    Join Date
    Aug 2007
    Posts
    1,088
    BG Level
    6
    FFXI Server
    Ragnarok

    No, didn't use gorget/belt. I think that extra accuracy was just due to the small sample cuz at one point it fall to 57% during the test.

    And I will do that test soon, just not sure how accurate it will be cuz the sample will be small again. It takes a long time to do cuz I am missing half my swings so it takes forever to get 100% tp xD.

    PS: aren't they 10 accuracy tho?

  7. #87
    Ridill
    Join Date
    Apr 2011
    Posts
    23,736
    BG Level
    10
    FFXI Server
    Bahamut

    You could cheat. Since you already have a large sample size with a given target with a certain acc you could melee in more acc then ws in the same amount of acc you were before with the exception of the gorget/belt of course. Just have to make sure your acc is exactly the same besides that. Including things like more dex from different subjobs if you are uncapped skill skillups since then ja/traits that increase it food etc

  8. #88
    Hydra
    Join Date
    Feb 2009
    Posts
    123
    BG Level
    3
    FFXI Server
    Bahamut

    Any suggestions for testing an attack bonus? I feel like Ruinator may have a decent attack bonus. Just having a hard time getting good results to tell me about how much or if there is an attack bonus on Ruinator.

  9. #89
    Chram
    Join Date
    Sep 2007
    Posts
    2,526
    BG Level
    7
    FFXI Server
    Fenrir

    Find a target where you can reliably get an exact cRatio. EG: only target EM mobs, and figure out exactly how much attack is needed to drop them to Low Defense, and weaponskill in that.

    Determine roughly how much damage a single hit of the weaponskill will do, and drop the mobs' HP to the point where you can kill them with just the single hit. This is easier to do if you have a partner, so you can reach a certain TP amount and then wait for the partner to finish dropping the mob's HP. Change your target to a different mob so that you don't accidentally melee it to death, and weaponskill your low HP mob.

    Collect 30-50 weaponskill samples for each TP level you want to test. The more the better. Provide all sample results for analysis.

    The lower the damage on the weapon you use, the easier it will be to find the edges of the pDif spread (up to a point), which makes it easier to identify any attack boost. 1/5 merits and a d10-d20 weapon would probably be easiest (unless that makes it too difficult to kill the mob), but if you've already got 5/5 just use the lowest weapon damage and WSC you can manage.

  10. #90
    Old Merits
    Join Date
    Aug 2007
    Posts
    1,088
    BG Level
    6
    FFXI Server
    Ragnarok

    Just giving you a head start, JP tested it to be above 35%, some seem to think it is 37.5% ratio bonus.

  11. #91
    Hydra
    Join Date
    Feb 2009
    Posts
    123
    BG Level
    3
    FFXI Server
    Bahamut

    I can verify that Ruinator has an attack bonus of some sort.

    With recent test has 64 base damage weapon, +173 str, nothing to increase ws damage, or damage in general. the mobs I was after have somewhere around 508 def, I had 719 attack.

    so had around 1.4 attack/def ratio.

    given the min pdif of 1.035, and a min damage of 77 we get 74 total base damage for weapon + fstr. Also the max of 140 dmg puts base dmg at 74.

    now we have 173 str, 74 base, you get 174*.85+74 = 221.

    The max pdif with ~1.4 cRatio is ~1.91. Then 221*1.91 should put the max at 422 damage on a single hit, and 221*1.035 = 228 min damage on a single hit.

    I hit a max single hit for 490, and second highest single hit I saw was 459. Weakest being 275. Now I didn't do enough tests to get the full spread only did 9 single hit's, I expect the top to be higher and low to be lower. 490/221 = 2.217. Which would put cratio at atleast 1.7. which would put my attack at 867 or a little more than +20%.

    The average counting all hits for where I hit for up to 4 hits. I ended with an average of about 360 dmg per hit on ruinator. which would put average at ~1.63 pdif, which is well above the expected average.

    These numbers aren't exact, but this does show that there is an attack bonus on Ruinator, and that bonus is at minimum around 20%.

  12. #92
    Nidhogg
    Join Date
    Jul 2008
    Posts
    3,746
    BG Level
    7
    FFXIV Character
    Seraphus Highwynn
    FFXIV Server
    Gilgamesh
    FFXI Server
    Diabolos

    Side question, any way to test if DRG somehow got Accuracy Bonus IV 96-99? I heard you can remove skills by changing to jobs that have no polearm skill like RDM/SCH on the test server then warping to Rabao and changing jobs at the nomad moogle to DRG and you'll have 0 polearm skill, then maybe I can test acc on low level mobs (I should only have DEX/2 and Acc Bonus III- +35, hypothetically)?

  13. #93
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    Yeah, I already told you that I tested it and they don't get it.

  14. #94
    Nidhogg
    Join Date
    Jul 2008
    Posts
    3,746
    BG Level
    7
    FFXIV Character
    Seraphus Highwynn
    FFXIV Server
    Gilgamesh
    FFXI Server
    Diabolos

    That was 90-95 I thought

    Any way we can test Critical Defense Bonus providing a "Critical Damage-x% Taken" (Negative version of Critical Damage+x%) using someone sub THF in ballista using Sneak attack to see if damage is reduced?

  15. #95
    Chram
    Join Date
    Sep 2007
    Posts
    2,526
    BG Level
    7
    FFXI Server
    Fenrir

    @Darkhorror: The numbers you are providing do not quite match what you claim they do, if my understanding of pDif is correct.

    Mob defense: how was this determined? You say 'around' 508 defense. What attack did you have in order to cross the neutral/low defense boundary?

    You state you have "+173 str". What is this in addition to, since you use the '+' symbol? If that's your total str, it should be written differently (eg: base+add=total).

    Ratio: "around" 1.4 should be more precisely defined. Assuming the defense is correct, with 719 attack it would be 1.415, or 1449/1024.

    "given the min pdif of 1.035" - You state this as given, but don't show evidence of how you determined it. Note that since your cRatio was less than 1.5, your min pDif will be exactly 1.0.

    "Also the max of 140 dmg puts base dmg at 74." - Again, how did you determine this?

    A max observed damage of 140 implies a max pDif damage of 134 (after factoring out the 1.05 spread). Given my notes (in the other pDif thread) about calculating max pDif using the 236/1024 multiplier, that implies a base damage of 77, not 74. Since that also matches your min damage value (which should have been at 1.0 pDif anyway), we can conclude that your base weapon+fStr damage is in fact 77.

    WSC with 173 str and 5/5 merits is 147. Total weaponskill base damage is 147 + 77 = 224

    "The max pdif with ~1.4 cRatio is ~1.91" - I'm not sure how you're reaching that value at all. Max pDif for the 1.415 ratio would be 1782/1024, or 1.740.

    Max expected damage without an attack bonus would thus be 224*1782/1024 = 389 * 1.05 = 408. Minimum damage will remain 224 unless you increase attack by more than 43 points, or 6%.


    "Now I didn't do enough tests to get the full spread only did 9 single hit's" - As I mentioned before, the higher your base damage, the more difficult it is to be sure that your sample values cover the full range with a given number of attempts. However we can at least see if any particular threshholds were crossed.


    "I hit a max single hit for 490" - This, however, gives us something to work with. Extracting the 1.05 multiplier from 490 means pDif damage had to be at least 467. 467 damage on 224 base means a pDif multiplier of at least 2.085 (2135/1024).

    If we consider that this attack multiplier is applied in the same phase as other such multipliers (ie: before cRatio is calculated, similar to other weaponskills), we need to back-calculate the cRatio that could support that max pDif. The max pDif addition to base cRatio is capped at 384/1024 (probably; not completely pinned down) for cRatios higher than 1666/1024. 1666+384 = 2050, lower than our required max. Therefore we can consider the max pDif to be using the capped version, and the actual cRatio to be 2135 - 384 = 1751/1024.

    Going from 1449/1024 to 1751/1024 is an increase of 302/1024, and would require a multiplier of at least +20.84%.


    "Weakest being 275." - Would seem impossible to draw any conclusions, but can make some guesses based on my older formulas.

    275 damage on 224 base would be ~1258/1024 pDif. Min pDif is calculated as an offset from a (factor*cRatio) value. For one-handed weapons, estimates were factor=1.13, offset = -711/1024.

    Working backwards, 1258 + 711 = 1969. 1969 / 1.13 = 1743. 1743 is remarkably close to max pDif value of 1751, though since it's lower it's suspicious.

    The max pDif that could result in 275 damage is 1261. Using 132/1024 as the multiplier on 1751 (which we know is a min value for the cRatio) we get 1751+225 = 1976. 1976 - 1261 = 715, which is a possible offset for the min pDif value (min pDif formulas were not completely nailed down, and I haven't restarted the testing yet).


    20.84% is 213.4/1024. Rounding up to the nearest interval of 4 gives us 216/1024.

    719 att plus a 216/1024 attack boost gives us +151 att, for a total of 870 att.

    Aside: I should also add a query as to how you reached 719 attack. That seems higher than reasonably achievable without food or buffs like Berserk. If any secondary buffs were in play, that can affect the final reasult, since percentage attack buffs stack additively (eg: 400 base attack with two 25% attack buffs would give you 600 total attack; if you didn't know about one of the attack buffs, it would look like it increased from 500 to 600 attack, or just a 20% boost). If that's the case, all of the above calculations are thrown out the window and we have to start over again.


    In any case, the above calculations for min and max don't quite sync up, though they're quite close. It's possibly a 21% attack boost, or possibly a 25% attack boost applied to an undetermined base attack value.

  16. #96
    Hydra
    Join Date
    Feb 2009
    Posts
    123
    BG Level
    3
    FFXI Server
    Bahamut

    Quote Originally Posted by Motenten View Post
    @Darkhorror: The numbers you are providing do not quite match what you claim they do, if my understanding of pDif is correct.
    https://www.bg-wiki.com/bg/PDIF

    those are what should be used. the ones on other site are incorrect.

    Mob defense: how was this determined? You say 'around' 508 defense. What attack did you have in order to cross the neutral/low defense boundary?
    around 508 because the low/regular def boundary changes depending on which mob I fight. Some are slightly stronger others slightly lower. But the variation is just about 2 or 3 levels, nothing worth while.


    Ratio: "around" 1.4 should be more precisely defined. Assuming the defense is correct, with 719 attack it would be 1.415, or 1449/1024.
    the exact value doesn't matter and isn't known considering the change in levels of the mob. You are talking a 1% difference here will make very little difference in the outcome unless you are looking for exact values.

    "given the min pdif of 1.035" - You state this as given, but don't show evidence of how you determined it. Note that since your cRatio was less than 1.5, your min pDif will be exactly 1.0.

    "Also the max of 140 dmg puts base dmg at 74." - Again, how did you determine this?

    A max observed damage of 140 implies a max pDif damage of 134 (after factoring out the 1.05 spread). Given my notes (in the other pDif thread) about calculating max pDif using the 236/1024 multiplier, that implies a base damage of 77, not 74. Since that also matches your min damage value (which should have been at 1.0 pDif anyway), we can conclude that your base weapon+fStr damage is in fact 77.

    WSC with 173 str and 5/5 merits is 147. Total weaponskill base damage is 147 + 77 = 224

    "The max pdif with ~1.4 cRatio is ~1.91" - I'm not sure how you're reaching that value at all. Max pDif for the 1.415 ratio would be 1782/1024, or 1.740.

    Max expected damage without an attack bonus would thus be 224*1782/1024 = 389 * 1.05 = 408. Minimum damage will remain 224 unless you increase attack by more than 43 points, or 6%.
    as I said before look at the link, the numbers you are using for pdif are incorrect.


    "Now I didn't do enough tests to get the full spread only did 9 single hit's" - As I mentioned before, the higher your base damage, the more difficult it is to be sure that your sample values cover the full range with a given number of attempts. However we can at least see if any particular threshholds were crossed.


    "I hit a max single hit for 490" - This, however, gives us something to work with. Extracting the 1.05 multiplier from 490 means pDif damage had to be at least 467. 467 damage on 224 base means a pDif multiplier of at least 2.085 (2135/1024).

    If we consider that this attack multiplier is applied in the same phase as other such multipliers (ie: before cRatio is calculated, similar to other weaponskills), we need to back-calculate the cRatio that could support that max pDif. The max pDif addition to base cRatio is capped at 384/1024 (probably; not completely pinned down) for cRatios higher than 1666/1024. 1666+384 = 2050, lower than our required max. Therefore we can consider the max pDif to be using the capped version, and the actual cRatio to be 2135 - 384 = 1751/1024.

    Going from 1449/1024 to 1751/1024 is an increase of 302/1024, and would require a multiplier of at least +20.84%.


    "Weakest being 275." - Would seem impossible to draw any conclusions, but can make some guesses based on my older formulas.

    275 damage on 224 base would be ~1258/1024 pDif. Min pDif is calculated as an offset from a (factor*cRatio) value. For one-handed weapons, estimates were factor=1.13, offset = -711/1024.

    Working backwards, 1258 + 711 = 1969. 1969 / 1.13 = 1743. 1743 is remarkably close to max pDif value of 1751, though since it's lower it's suspicious.

    The max pDif that could result in 275 damage is 1261. Using 132/1024 as the multiplier on 1751 (which we know is a min value for the cRatio) we get 1751+225 = 1976. 1976 - 1261 = 715, which is a possible offset for the min pDif value (min pDif formulas were not completely nailed down, and I haven't restarted the testing yet).


    20.84% is 213.4/1024. Rounding up to the nearest interval of 4 gives us 216/1024.

    719 att plus a 216/1024 attack boost gives us +151 att, for a total of 870 att.

    Aside: I should also add a query as to how you reached 719 attack. That seems higher than reasonably achievable without food or buffs like Berserk. If any secondary buffs were in play, that can affect the final reasult, since percentage attack buffs stack additively (eg: 400 base attack with two 25% attack buffs would give you 600 total attack; if you didn't know about one of the attack buffs, it would look like it increased from 500 to 600 attack, or just a 20% boost). If that's the case, all of the above calculations are thrown out the window and we have to start over again.


    In any case, the above calculations for min and max don't quite sync up, though they're quite close. It's possibly a 21% attack boost, or possibly a 25% attack boost applied to an undetermined base attack value.
    If you want more precise numbers you need to do another test vs something where you know the def of every mob you fight. Plus you need to do far more single hit kills. This was a test to make sure there is an attack bonus, give a general min and max attack bonus. Which is also why using /1024 isn't going to be much help because the margin of error is large enough that just using strait numbers is much easier and won't change results.

  17. #97
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    Motenten has his own pDIF model that is probably as/more accurate than the regressions I posted.

  18. #98
    xXNyteFyreXx420Sharingan
    Join Date
    May 2009
    Posts
    3,710
    BG Level
    7
    FFXI Server
    Fenrir

    Quote Originally Posted by darkhorror View Post
    those are what should be used. the ones on other site are incorrect.
    Those equations aren't fully accurate either, as others have noted.

  19. #99
    Hydra
    Join Date
    Feb 2009
    Posts
    123
    BG Level
    3
    FFXI Server
    Bahamut

    None of it really matters anyway, they are all very close the base numbers aren't exact enough for it to matter. We will have a margin of error but it won't be more than a couple percentage points. What is shown is that we have a decent sized attack bonus, what looks like is somewhere around 20% +/- a percentage or two, or a strait attack bonus of that much. Really the focus on that testing was to test to see if there is an attack bonus, it's clear there is, and get atleast an idea of how much that bonus is. Was no where even close looking for an exact amount, but what we do have is a good value for an approximation of what it may be. The changes in numbers that are being argued are most likely with in the cratio changes based on the mobs def anyway so changing the numbers a percent one way or another doesn't change the outcome.

  20. #100
    Chram
    Join Date
    Sep 2007
    Posts
    2,526
    BG Level
    7
    FFXI Server
    Fenrir

    Quote Originally Posted by darkhorror View Post
    https://www.bg-wiki.com/bg/PDIF

    those are what should be used. the ones on other site are incorrect.

    ...

    as I said before look at the link, the numbers you are using for pdif are incorrect.
    Err. No. Both sites have significant issues with the equations. I don't use either.


    Quote Originally Posted by darkhorror View Post
    around 508 because the low/regular def boundary changes depending on which mob I fight. Some are slightly stronger others slightly lower. But the variation is just about 2 or 3 levels, nothing worth while.
    Since you'd asked for advice on how to do the testing, I'd assumed that you had in fact followed the advice given:

    Quote Originally Posted by Motenten View Post
    Find a target where you can reliably get an exact cRatio. EG: only target EM mobs, and figure out exactly how much attack is needed to drop them to Low Defense, and weaponskill in that.
    As such, yes, the best you can say about your damage values is that you got an approximation of a result. 20% is close enough for a rough guess, and just to show that the attack bonus does in fact exist, and is fine if that's all you were going for. However all of the exact analysis is useless because the mob levels varied.


    Quote Originally Posted by darkhorror View Post
    the exact value doesn't matter and isn't known considering the change in levels of the mob. You are talking a 1% difference here will make very little difference in the outcome unless you are looking for exact values.
    Err.. well, yes, I was in fact looking for exact values.

+ Reply to Thread
Page 5 of 15 FirstFirst ... 3 4 5 6 7 ... LastLast