Now, I plan on doing some of this testing on Drg tonight so can vary str a bit more. Basically naked except weapon, and all gear except +ws damage(will find something to put in place of those peices from what I have on hand). If the damage increase on weapon skills is close to 1/3 of my Str again, would that be enough to indicate that the minimum increase is close to that for you? As I have seen that as 50sam/23mnk and 23mnk/blu so far when done weaponskill testing. If it shows up on Drg, that would be up to 4 different str values that it has happened with. Will again be doing it on tunnel worms in south gusta, and be using Leg Sweep, TP affects Stun chance, fTP of 1.0, and while it has a STR mod of 100% itself I will establish a baseline Leg Sweep value for both naked and with gear before using boost, like I have with the others. I don't care to find out what the exact % of Str it is adding, but just about everyone else was saying that boost wasn't affecting WS's at all prior to the linked dev posts spurring new testing, and my testing was showing it was. I also have given enough info here so that others could at least repeat the process to verify or disprove if they wanted, and yet to see someone else disprove or verify. I just haven't put in a lot of time to get enough numbers to rule out variation enough to pinpoint things so one could make a formula.Quote:
Originally Posted by Urmom
And the 1/3rd str is kinda important, as look at the patch notes Nyarlko posted. With it not being noticed prior, I was guessing that it might be a % based stat mod like Weaponskills normally have, and 33% would be low enough to be missed.
From what I have seen in practice(yet to test the Boost->wait->WS pattern to verify myself) Odds are best method to use boost is waiting as close to its expiration as possible before using a WS or JA, or let it run out for an enhanced auto-attack.Quote:
Originally Posted by Morovich
Problem is weighing out if the combination of wait time(and no tp feed to monster), increased TP gain on the hit, and increased damage is worth it. If it multiplies into weapon skills like auto attacks, high base fTP(such as Howling fist and its 8.0), or ones that duplicate fTP for all hits(like Stardiver, not sure if any H2H ones do) may be the best ones to use it with. The mathy stuff is to try and get to the bottom of how to best use it, and with what.
-
08-11-2017, 09:23 AMHawklaser
-
08-11-2017, 09:56 AMNyarlkoEngage > Boost > wait til 1sec left on Boost duration > WSQuote:
Originally Posted by Morovich
Does not work w/ magic WS, best to be fully haste/delay capped before popping Boost to minimize the standing around time. -
08-11-2017, 11:48 AMHawklaser
Well here is what I found testing on Drg/23mnk tonight. Again, only used tunnel worms in south gusta for test targets, and as little wait as possible from using boost to ws.
Ungeared with 103 str Leg sweep did 1619-1677 for its damage range. Based on earlier testing I had done, I expected Boosted range to go up by 34,for an expected range of 1653-1711. I saw a range of 1677-1718.
With gear for a Str of 291, Leg sweep did 2384-2480. Expected damage to go up 97, for a range of 2481-2577. I saw a range of 2481-2587.
Now while I don't have enough numbers to be significant enough to pin things down precisely. I'd say being able to predict in advance the boosted ws range by simply adding 1/3 of my Str to normal Leg Sweeps, and be that close is a pretty clear indication the base +str modifier for boost to weapon skills is somewhere in the range of 30-40% of ones Str.
After this, I did a couple of tests with gear on and varying amount of boost time passed. I tried to break it up into quarters. So waiting for 0-25% of Boosts delay, I was seeing same numbers as did with no wait. If waited until between 25-50% passed, I saw the added damage about double. Waited for 50-75% it about tripled. From 75-100% it was about quadruple. This was with very little testing, just trying to see where breakpoints might be, and how much they may vary. I don't plan to do much more testing on this, but I hope it helps those interested in figuring things out more. -
08-12-2017, 12:33 AMUrmomYou'd need to account for all variables in the dmg calculation. Doing hitting low level stuff like that more or less accounts for pdif and fstr but would need decently varying weapon base dmg too. But even then it's not 1/3 str or whatever number getting added it'd be more like 1/12 getting added to base dmg being multiplied by pdif. Also given that and roundings that are done and the variability of pdif to begin with larger differences are much more conclusive. Well there is also the small possibility the boost counts as an "extra hit" so might need something you are one shottingQuote:
Originally Posted by Hawklaser
Again no base stat directly adds to dmg it adds to base dmg which gets multiplied by things like pdif so there is multiple stages of rounding and such. 1/400 would be even harder to notice too. Which is the number gotten for regular hits. I know what notes say also know this is supposed to give a boost independent on str which testing has shown to be much larger than the str bonus but the dev note Nyarlko posted suggests is wait time dependent. So whatever boost to dmg you get isn't going to be all from strength to begin with and apparently it's going to be variable if you don't wait -
08-12-2017, 03:27 AMHawklaserLets see what have accounted for.Quote:
Originally Posted by Urmom
Low level mob, and same for all tests rules out pdif and fSTR.
Tachi: Hobaku, Shoulder Tackle, and Leg Sweep all have a fTP of 1.0, Tp to increased status chance, and established baseline damages prior to boost to account for native stat mods on the weaponskills. That takes care of normal ws mods, fTP, and variation from TP.
Weapons used were Mumetio (lv 1 G.katana), unarmed 23 mnk(would have to log in for skill so could give so could figure it out) and Gae derg+1 (119 lance with 299 damage). I'd say that covers weapon variation quite well.
Varied Str was 30, 58, 103, and 291. 103 and 291 were with same weapon. So I'd say that covers Str variation.
What else besides large testing sample is needed?
-Edit-Needed to mention that establishing a baseline damage for the WS prior to using boost should also eliminate any variation besides the normal randomization inherent in the damage system. Meaning the only changes in damage should be from what boost was providing and that normal variation which would get ruled out quickly with a large sample size.
So how about this, verify or disprove my testing with some of your own. Steps would be simple, record your STR when have no +WS damage gear on, go to south gustaberg so can do testing on the same tunnel worms, establish a baseline damage range with a one-hit fTP 1.0 weaponskill that has TP -> status before using boost, estimate the expected damage range by adding 1/3 of your STR to the baseline range,then engage a worm in a manner so can boost-> ws as fast as possible without accidentally having an auto-attack kill the worm before boost to check the prediction. Note:Thanks to knowing a multiplier is tied to the damage, ensure your boost duration is at least high enough that 25% of it will not pass before a ws is used, I had over 10s durations for all of my testing, Gae Derg was over 20s.
So until someone proves otherwise with a repeatable method, I am fairly sure this is how boost is affecting a 1-hit, 1.0 fTP weapon skills with tp only affecting status like stun. Normal Weapon skill damage + (1/3 STR * Time multiplier). Time multiplier I am guessing is around 1.0, 2.0, 3.0, 4.0 for Mnk subjob broken up in 1/4 chunks of boosts elapsed duration. IE 1.0 for less than 25% duration elapsed, 4.0 for 75%+ elapsed. Now if the 1/3 STR * Time multipler is before or after fTP or is duplicated to all hits of a weapon skill, I don't know. My original intention was to prove that Boost was affecting WS's or verify if it indeed was broken, which I saw Boost affecting WS's before the waiting came into play. Everything after that, was more to help others who enjoy really digging into the math stuff pin down how its working more accurately.