Item Search
     
BG-Wiki Search
+ Reply to Thread
Page 2 of 15 FirstFirst 1 2 3 4 12 ... LastLast
Results 21 to 40 of 294

Thread: Weapon Skills     submit to reddit submit to twitter

  1. #21
    Masamune
    Guest

    Weird that ppl missed that thread to post testings for the new WSs.

    I'll try to update from test server thread back here...

    EDIT: nvm looks like Byrth already made those links on each WS page @ BGwiki.

    So... is this thread any useful ? delete ?

    EDIT2: nvm there are my own WS tests results here :s

  2. #22
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    I've made a halfassed attempt to nail down Evisceration's 300TP crit value last night/this morning using the Abyssea/100% Crit rate testing method, but I think there is a better way to test it after I'm close. I'm thinking of something along these lines:
    * My Critical hits do way more (~4x as much?) damage as my normal hits with Sanguine Scythe, Hero, and Razed Ruins.
    * All I'm interested in is whether I non-crit at all
    * fTP of all hits is known.
    * If I go a job without Conserve TP, I know exactly how many hits my WS had.
    And then one more optional part:
    * The more hits (crits) we have, the more likely we are to see an average crit pDIF near the average crit pDIF for my Ratio.

    So the questions are:
    1) How many hits can I allow on WS with 1 Crit ~= 4x Normal damage before I'm no longer completely confident that I can tell the difference between a crit and a non-crit?
    - For instance, if my single-hit Crit WSs do between 550 and 700 damage, my two-hit Crit WSs would do between 1100 and 1400 damage. If I suddenly get an 800 damage WS with a 10 TP return, I can be entirely sure that one of them was a non-crit.
    2) If we're willing to accept the final assumption, can we tell the difference between an 8-crit and 7-crit/1-normal WS with a tolerable degree of confidence? Is it worth the trade-off for the speed increase compared to 1?

  3. #23
    Masamune
    Guest

    Honestly, i prefer the hard way by far, ie trying to test faster trading off even a little confidence can only lead to errors, and they stack...

    hard way = test only 1 WShit, that's why can't do that in abyssea since can't oneshot any mob (and variable levels don't help either...).

    WS Crit rate testing is the kind of shit that should be done in group, with several testers, like we did for crit%=f(dDEX), each testers adding his numbers to the pool here, pretty much like they did for Evisc testing at Allak.

  4. #24
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    Quote Originally Posted by Masamune View Post
    Honestly, i prefer the hard way by far, ie trying to test faster trading off even a little confidence can only lead to errors, and they stack...

    hard way = test only 1 WShit, that's why can't do that in abyssea since can't oneshot any mob (and variable levels don't help either...).

    WS Crit rate testing is the kind of shit that should be done in group, with several testers, like we did for crit%=f(dDEX), each testers adding his numbers to the pool here, pretty much like they did for Evisc testing at Allak.
    Well, I just melee the monster down to 1~10% and WS it when I want to only test one hit in Abyssea. It's more of a pain in some ways and it's hard for me to hit the 200TP point easily, but if I can melee it down to 12~15% and WS it for twice the N then I'm even better off. If I can WS it from 100% with an 8-hit Evisceration and collect data 8 times as fast . . . baller! I don't think I can, though.

  5. #25
    Masamune
    Guest

    You thinking too much imho :D

    Just keep it simple and reliable on rabbits Ronfaure, but ask for help, you can't test all critbased WSs at 100/200§300TP alone.

  6. #26
    Masamune
    Guest

    Quote Originally Posted by Byrthnoth
    I don't know why I did this on the test server instead of real server, but I cranked out some 300TP Ukko's tonight. I got a little over 70 in 2 hours (and change) and didn't get a single non-crit... with . . . . . . . capped dDEX/merits and no additional crit hit rate. 1308-1750 damage (variety of monsters/Ratios).

    So yeah, looks like Ukko's probably gets a pretty substantial crit rate bonus at 300 TP?

    I'm interpreting my results to mean it's at least 70%. Keep in mind that the other Ukko's Crit testing I know about was at 100TP and showed that the crit rate boost was greater than 30%. I confirmed that it's less than 40% at 100TP tonight, but not a lot less (took me a while to get a non-crit). I'd be unsurprised to find the progression is something like 35% to 70% or 75%. If the curve is really that steep, a 25TP Moonshade would add ~5% Crit rate. I guess it depends what the 200TP crit rate is though.... maybe for another night.
    i found this in random fact thread, but i'm wondering if you deduced base crit% you had vs. the mob from all those crit% in your post ? ie if you parse "less than 40% @ 100TP" while having a base natural crit rate of 15% vs that mob -> Ukko crit% @ 100TP = 40-15 = "less than 25%".

    Also, since it has been shown crit dmg bonus is no longer capped @ +50%, i'm not sure anymore if there is still a dDEX "cap" (crit%max=50%) ?

  7. #27
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    The 100% Crit rate method can never absolutely confirm a lower bound for a crit rate. It can only confirm the upper bound. So when I say the bonus is less than 40%, it means I had 60% from other sources and saw a non-crit. If you look at the tests that have been done on Ukko's Fury (all linked or on the talk page), there's my non-crit showing it's less than a 40% boost and a ton of crits by another guy showing that it's probably more than 30%. There's also a ton of crits by me showing the 300TP boost is likely 80% or higher.

    dDEX definitely does still cap at 15% (+base+merits+gear+atma) like it always did, unless they changed it last patch or something. All of my Abyssea parses (from back when I really did Abyssea) came out exactly where I would have expected for a 15% Crit rate from dDEX.

    It's faster to absolutely determine critical hit rate (with 95% confidence) using the 100% crit method. Here is a link that explains it again.

  8. #28
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    Studio Gobli appears to disagree with quite a few of our Mythic WS mods, even including easy shit like Magic WSs. The thread where we tested the Mythic WS properties is huge and full of people that didn't know how to test WSs which means there's a lot of shit to sift through and you can't always trust the results.

    I'm 100% sure at least one of their WS mods/fTPs is wrong (Pyrrhic Kleos), but other mods of theirs match up with things I've observed better than ours. For instance, a while ago I wasted a bit of time trying to find a hidden effect on Twilight Helm because Primal Rend was getting extra damage with it compared to 10 INT/CHR. If Primal Rend is also a DEX mod, it would make sense. It would also make me pretty incredibly lucky for not switching in/out DEX gear when doing my killer tests. The magic WSs, at least, should take about 30 seconds in Abyssea to test and verify.

  9. #29
    First invited, last in the zone.
    Join Date
    Sep 2008
    Posts
    1,449
    BG Level
    6
    FFXI Server
    Lakshmi

    Blaaah, Drakesbane does have an attack penalty. I don't have enough data to nail down what it is, but the ~19% from gobli is plausible.

    SE, Y U HATE DRG?!

    We can sub war for berserk and we don't have a native attack bonus JA and you cripple all our multi-hit WS with attack penalties >_<

  10. #30
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    >:/

    So Garland of Bliss and Primal Rend are the two WSs with very obvious differences between Studio Gobli and the NA testing, and Gobli has a weird dSTAT for almost every Mythic WS. I don't know how/whether they cap, either. I don't remember if I ever bothered to unlock Garland of Bliss, but I can certainly brew-confirm Primal Rend tomorrow.

    Also, Gobli can't decide if Mandalic Stab is +2/3 or +3/4 Attack (they have two pages that each say each thing). I think there was some NA testing on it, but I have lost it. Does anyone know where it went?

    Also also, they have True Strike getting a 2x Attack Multiplier, which explains why it always did so much better than people expected. tmyk

    Other things:
    * Infernal Scythe's Attack Down effect is probably fairly potent (judging by Tachi: Ageha), but we don't know what it is.
    * I have filed away that fINT = 0 for Herculean Slash (and Infernal Scythe?). Where did I get that? Is it true?

  11. #31
    An exploitable mess of a card game
    Join Date
    Sep 2008
    Posts
    13,258
    BG Level
    9
    FFXIV Character
    Gouka Mekkyaku
    FFXIV Server
    Gilgamesh
    FFXI Server
    Diabolos

    Where is the 19% listed on Gobli?

  12. #32
    First invited, last in the zone.
    Join Date
    Sep 2008
    Posts
    1,449
    BG Level
    6
    FFXI Server
    Lakshmi

    Quote Originally Posted by Yugl View Post
    Where is the 19% listed on Gobli?
    http://www32.atwiki.jp/studiogobli/pages/93.html

  13. #33
    Masamune
    Guest

    Quote Originally Posted by Byrthnoth View Post
    Studio Gobli appears to disagree with quite a few of our Mythic WS mods, even including easy shit like Magic WSs. The thread where we tested the Mythic WS properties is huge and full of people that didn't know how to test WSs which means there's a lot of shit to sift through and you can't always trust the results.

    I'm 100% sure at least one of their WS mods/fTPs is wrong (Pyrrhic Kleos), but other mods of theirs match up with things I've observed better than ours. For instance, a while ago I wasted a bit of time trying to find a hidden effect on Twilight Helm because Primal Rend was getting extra damage with it compared to 10 INT/CHR. If Primal Rend is also a DEX mod, it would make sense. It would also make me pretty incredibly lucky for not switching in/out DEX gear when doing my killer tests. The magic WSs, at least, should take about 30 seconds in Abyssea to test and verify.
    don't even need abyssea: test on Lesser colibris, i got their stats already just check all the Magic WSs i already tested and results posted in this thread.
    And yes i noticed a lot of differences between Gobli and us... and also questionable methods from testers here :s (but hey gobli is no better where we can't even see how they tested lol...)

    BTW, i have a serious doubt on all those empyrean WSs tests about their critrate: the aftermath can screw results.
    Are testers patient enough to wait for aftermath to wear off (or desequip/reequip weapon) before readying another WS ?
    Here is an example just to clarify:
    - i know normal/crit ranges are 100-200 300-600.
    - i do my 1st WS: 150dmg , ok this a normal dmg.
    - i do my 2nd WS (aftermath still up since im impatient): 350dmg, er... is this a crit dmg, or a normal 125dmg aftermath'd ?
    - /em blames BG
    ... so yea wouldnot surprises me much if actual empyrean critrate bonus are like 2/3 what has been currently found.

    Is there a command to set XXX TP on Test Server ?

  14. #34
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    Quote Originally Posted by Masamune View Post
    don't even need abyssea: test on Lesser colibris, i got their stats already just check all the Magic WSs i already tested and results posted in this thread.
    Well, I brew-tested it. I confirmed I was stat capped (I think) by killing a DC and EP Buzzfly with the same damage Primal Rend. Neither set of equations works, but the JP equation is slightly more viable.

  15. #35
    Masamune
    Guest

    Quote Originally Posted by Byrthnoth View Post
    * I have filed away that fINT = 0 for Herculean Slash (and Infernal Scythe?). Where did I get that? Is it true?
    In this very same thread lol, check main page and follow the white rabbit... ooops Infernal Scythe link :D

    I'll make some tests on Primal Rend, but on my dmg calc (WS tab) i have:
    CHR30%, fTP100=3.75+1/16, fTP200=4+1/16, fTP300=4.5+1/16.
    Sadly, i dunno where i got those numbers...

    EDIT: Infernal Scyther 15% attack down, again i dunno where i got this number :s

  16. #36
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

    I link the source of those numbers from the bgwiki page for it, and I found your testing on fINT=0 for those WSs and linked them after posting that. The JP formulas totally work if you assume dSTAT caps at 434. I further confirmed them on Wild Rabbits. They're right.

  17. #37
    Masamune
    Guest

    Quote Originally Posted by Byrthnoth View Post
    ... dSTAT caps at 434. I further confirmed them on Wild Rabbits.
    what do you mean by "dSTAT"? i don't understand how you can get a 430+ INT difference outside abyssea o.O ?
    Can you make a post with your testing numbers confirming JP numbers please ?

  18. #38
    BG Content
    Join Date
    Jul 2007
    Posts
    21,105
    BG Level
    10
    FFXI Server
    Lakshmi
    Blog Entries
    1

  19. #39
    xXNyteFyreXx420Sharingan
    Join Date
    May 2009
    Posts
    3,710
    BG Level
    7
    FFXI Server
    Fenrir

    Quote Originally Posted by Masamune View Post
    BTW, i have a serious doubt on all those empyrean WSs tests about their critrate: the aftermath can screw results.
    Either I'm horribly misunderstanding your post or you're the impression that Empyrean aftermath procs on WS. As far as I know it doesn't, I've certainly never seen a damage spike that couldn't be explained through other mechanics.

    I have a little bit of time, so I'll go look into Infernal Scythe's attack down value real quick.

  20. #40
    Masamune
    Guest

    Quote Originally Posted by Nightfyre View Post
    Either I'm horribly misunderstanding your post or you're the impression that Empyrean aftermath procs on WS. As far as I know it doesn't, I've certainly never seen a damage spike that couldn't be explained through other mechanics.
    Nope you understood correctly, and what "could" be the proofs are quite rare but happens: when you have an average of 2k-4k spiking at very best 6k and suddenly out of nowhere yu hit a 9k ??!! A monk friend also mentioned those very high spikes on his VSmites, but rare.
    As far i paid attention, the very few WTFspikes i got on my Ukkos were always with aftermath up.