Random facts about abilities. Self-explanatory.
Random facts about abilities. Self-explanatory.
Ramuh's Favor gives Critical hit rate +21% with 512 Summoning Magic Skill (the last Favor tier reachable) and Caller's Horn +2. Used the 100% Critical hit rate method with RR/GH/4 merits/Iga Ningi +2 and Haste/double March to speed up attacks.
The Double Attack rate for Ifrit's Favor should be at least the same or higher (I had parsed a 22-24% Double Attack rate in the past using a Ram Staff on a Zvahl Fortalice)
Inferno Howl (Ifrit's Enfire buff) caps at 45 points of damage with 500 Summoning Magic Skill.
Manawell doesn't seem to activate on conserve MP procs...
I noticed this while trying to proc Goetia +2 set bonus on Ul'hpemde mobs in Sea.
I have a screenshot but I'm not good at uploading images.
Here's the chat log though:
> Manawell JA used.
> MP = 193
> Blizzard IV cast. (164 MP)
> MP = 50 (MP lost = 143 -- 2/16 MP conserved, no Goetia+2 set proc.)
> Convert JA used.
> MP = 918
> Blizzard IV cast. (164 MP)
> MP = 806 (MP lost = 112 -- 5/16 MP conserved, no Goetia+2 set proc.)
> Manawell effect wears off.
Thanks to Ophannus and Yugl for helping out with the data collection.
Total attempts: 210
Angons consumed: 191 -> 90.9524%
Standard Error: 0.019795 -> ~91% ± ~2%
Expected true value: 90%.
Did 60 angons this morning and saved 5. It's probably a 10% proc on Dragoon's Earring it seems.
Even if you add that to the data, it's still about 9%.
With Ophannus's new data it becomes 91.(1)% probability of being consumed, ± 1.73%.
Total used 246, total attempts 270.
Used about 1000 Bounty Shots between last night and now using Sylvan Glovelettes +2. Got stuck at TH11 for the last 600~ or so, although it took 250~ between TH9 and TH10, so it might not be 'maxed out'. Going to try Feint on the test server I suppose.
Oh yeah, should be worth mentioning that I had WHM subjob, not THF subjob.
I believe SE posted on the official forums about how they were adjusting Tier 2 merit JAs that only had reduced recasts. Either I missed something or they changed their minds about some of them, because now I'm seeing:
* Warrior's Charge, "Increase chance of triple attack by 5%"
* Assassin's Charge, "Increase chance of fourfold attack by 5%"
* Feint, "Increase chance of Treasure Hunter level up by 25%"
* Martyr, "Increase amount of HP healed by 5%"
* Devotion, "Increase amount of MP granted by 5%"
* Dark Seal, "Shorten cast time of all dark magic spells by 10%"
* Diabolic Eye, "Increase physical attack accuracy by 5%" (10 Acc per merit level)
* Killer Instinct, "Increase duration of effect by 10 seconds"
* Feral Howl, "Increase accuracy by 5%"
* Troubadour, "Increase chance of unerring song accuracy by 25%"
* Nightingale, "Increase chance of spectacular "Fast Cast" effect by 25%"
** Not actually Fast Cast (Reraise unaffected) but it causes you to instant-cast songs (recast still only -50%)
* Flashy Shot, "Increases Ranged Attack by 5%"
* Shikikoyo, Still not adjusted
* Blade Bash, Still not adjusted
* Sange, Still not adjusted
* Deep Breathing, Still not adjusted
* Diffusion, Still not adjusted
* Snake Eye, Still not adjusted
* Fold, Still not adjusted
* Role Reversal, Still not adjusted
* Ventriloquy, Still not adjusted
* Saber Dance, Still not adjusted
* Fan Dance, Still not adjusted
* Enlightenment, Still not adjusted
It looks like they are adjusting things from top to bottom in the merit menu, and all of these changes are based purely on the descriptions.
Chi Blast Damage
I tested Taurine Cesti's interaction with boost and chi awhile back and came a simple realization very quickly. The known chi formula, wherever it came from, is the worst thing I have ever seen.
Chi Blast formula from wiki:
There are a few things with this that just make no sense, like the placement of the random number and such. If we plug 0 boosts into this, we would come out with DMG=MND with no fluctuation in damage at all, so the random should be multiplied by MND before # of boosts ever enters the equation. Logic would dictate that chi in it's simplest form would be:The generally accepted damage formula for Chi Blast is as follows:
MND * (#Boost * Rand + 1) = DMG
MND is the player's MND stat
#Boost is the number of times the player has boosted
Rand is a game determined number between 0.25 to 0.60 (without Temple Gloves) or 0.75 to 1.10 (with Temple Gloves).
MND * Rand= DMG
Chi damage when boost is factored in would then be:
MND * Rand(x+1) where x equals boosts interaction with chi. Now it's time to figure out why and how boost increase chi damage.
It is known that boost increases a player's attack by 12.5% and that boosting with Temple Gloves(enhance boost) attack is increased by 18.75%. Taurine Cesti increase boost by an additional 6.25% as well meaning with all 3 attack is boosted by 25%. The increase itself shows that either on their own give a 1.5 multiplier to boost increase itself and 2.0 multiplier when both are equipped. This is what explained so much of Chi's Formula to me.
What has been common knowledge for years is that a 12 boost chi while wearing temple gloves give a cap of mnd x 10, on top of which with 12 boosts damage will ALWAYS be divisible by 10. What may not be common knowledge is 12 boosts with no equipment will always be divisible by 7, additionally damage with gloves and cesti with 12 boosts I found to be always divisible by 13. While the numbers 7,10, and 13 seem to have no exact correlation with each other other than each equip adds 3, you need to factor out 1 from each to remove what the damage would have been without boosting. After doing so you are left with 6,9, and 12 being the boost interaction. They have the same 1,1.5,2 ratio as boost's 12.5%,18.75%,25%.
Nothing equipped 12 boosts gives a 6x damage multiplier or #boosts * 50%.
With temple Gloves or Taurine Cesti gives a 9x multiplier or #boosts * 75%.
With gloves and cesti gives a 12x multiplier or #boosts * 100%.
50%=4 * 12.5%, 75%= 4 * 18.75%, 100%= 4 * 25%.
Boost's Interaction with chi appears to be MND * Rand(# of boosts * 4 * Boost Increase Percentage)
Chi then equals MND*Rand(#boost*4*BoostIncrease+1)
The formula gets floored after MND*Rand or final damage wouldn't always be divisible by the same number and is floored again at the end since you can't have decimals in damage.
Chi Damage = floor(floor(MND*Rand)*(#boost*4*BoostIncrease+1))
Finding the Rand:
I ended up wasting much more time than was needed with this part, I seem to do that often with tests.
The Random number is always between .5 and 1. I have never seen a chi do more than 10*MND nor have I ever see one do less than .5MND*10.
I did all tests with 0 boosts.
Test 1 134 MND over 200 boosts gave a Dmg spread of 67-134,hitting all values in between, never higher, never lower. Odds are slim that a chi outside that range is possible. If random is out of x/128 I would have not been able to hit some of those damages.
Test 2 I realized that was a huge waste of time and decided to next use a brew to get 999MND. I lucked out and got 976 damage on first chi.
976=999*1001/1024, you can not get that value when using x/256 or x/512, only if random number is out of 1024.
Final Formula
Chi Damage = floor(floor(MND*Rand)*(#boost*4*BoostIncrease+1))
MND= Player's MND Stat
Rand= random number between 512/1024 and 1024/1024
#boost= number of boosts used
BoostIncrease= 12.5%, Temple Gloves and Taurine Cesti each raise this an additional 6.25%.
Why do you separate BoostIncrease and 4 instead of combining them?
Then: BoostIncrease= 50%, Temple Gloves and Taurine Cesti each raise this an additional 25%.
Regardless, thanks! That was one of the formulas that seemed omni-present but I was uncomfortable with.
I only separate them to show where the 50% etc. stem from. Although I technically explain all that in my post, if someone reads just the formula as boost*boostincrease where boostincrease =50% without reading my reasoning then they wouldn't have that understanding of boostincrease= 4 times expected value of what applies to melee attack increase.
In the long run I suppose that doesn't matter at all as either someone would read my testing and understand, or not care at all. Both result in the same fact knowing the formula wouldnt change the fact that if actually using chi blast you would stack the hell out of mind, and boost while in gloves and cesti. If you would like to simplify it and just call boostincrease=50% that is fine with me though, I am not exactly the best at formula notation anyway. As an example at first I was going to use BM(Boost Multiplier) then I thought "hmm, maybe Bowel Movement abbreviation would be a bad idea".
I like the boost increase separation just because 'Boost Number' and 'Boost Multiplier' are universal terms. That provides a universal term for the effect. Rather than writing two different effects on the boost/Taurus/Future Boost enhancement pages, you can just write "Increases boost #" or "Increases boost multiplier" and these terms will have the same meaning across formulas.
Counterstance
After reading Motenten's data on counter rate with melee gaiters+1 and +2, used Mandies as the enemy for all tests I decided to do some own data gathering of my own.
I have always been a firm believer that monk type enemies, while great for speeding up hits taken are a bad choice for counter testing. I have had the suspicion for years that a counter could be countered or guarded itself, basically making it appear that the initial counter never occurred. Martel's Post about the mechanics of shield block and counter furthered to almost cement my suspicions.
All tests were done on Abyssea LaTheine frogs. These neither Guard, Parry, Counter, nor Shield Block, making them a good enemy to use. Tests ranged from between 4 frogs to 8 frogs used. I would only start parsing when I was fully sure that all frogs were in front of me and counterable, so no stray damage was taken.
Test 1: 99RDM/MNK counterstance rate
Counters: 1228Code:Melee Damage Taken Player Melee Dmg Melee % Hit/Miss M.Low/Hi M.Avg #Crit C.Low/Hi C.Avg Crit% Diasetsu 221862 100.00 % 2032/3925 0/198 108.40 19 156/246 191.79 0.94 % Defenses Passive Defenses Player Evasion Evasion % Parry Parry % Counter Counter % Intimidate Intimidate % Diasetsu 1697 26.44 % 462 9.78 % 2228 52.30 % 0 0.00 %
Times Hit: 2032
Counter Rate: 52.3% , 55.05% before accuracy check
Confidence Interval: (1.5%) 50.8%-53.8%
CI factoring out 95% accuracy: 53.47%-56.63%
RDM/MNK gets 55% counter under counterstance. This is also what Montenten concluded for MNK main. Does counterstance just floor rate to 55% and not include counter merits, or does it add 45% to counter trait 1s base 10%.
Test 2: 99MNK/DNC Counter Rate with 5/5 Counter Merits
Counters: 1078Code:Melee Damage Taken Player Melee Dmg Melee % Hit/Miss M.Low/Hi M.Avg #Crit C.Low/Hi C.Avg Crit% Diasetsu 307155 100.00 % 5587/6561 11/81 54.45 54 54/138 109.15 0.97 % Defenses Passive Defenses Player Evasion Evasion % Parry Parry % Counter Counter % Intimidate Intimidate % Diasetsu 5483 45.14 % 0 0.00 % 1078 16.17 % 0 0.00 %
Times Hit: 5587
Counter Rate: 16.17%(17.02%)
CI: (.88%) 15.29%-17.05%
CI/95%: 16.09%-17.95%
99MNK Counter appears to be 17% with 5 Merits, making Counter II 12%
Test 3: 99MNK/DNC Counterstance (No Gaiters)
Counters: 3256Code:Melee Damage Taken Player Melee Dmg Melee % Hit/Miss M.Low/Hi M.Avg #Crit C.Low/Hi C.Avg Crit% Diasetsu 219101 100.00 % 2238/8296 11/134 97.52 19 81/166 142.47 0.85 % Defenses Passive Defenses Player Evasion Evasion % Parry Parry % Counter Counter % Intimidate Intimidate % Diasetsu 5040 47.85 % 0 0.00 % 3256 59.26 % 0 0.00 %
Times Hit: 5494
Counter Rate: 59.26%(62.38%)
CI: (1.3%) 57.96%-60.56%
CI/95%: 61.01%-63.75%
Counterstance does not floor rate to 55%, it does seem to increase rate by 45%. This shows that 99MNK with merits most likely has a 62% counter rate under base Counterstance.
Test 4: 99MNK/DNC Counterstance (Gaiters/Gaiters+1)
I only had Melee Gaiters equipped during activation, all data in parses is with no footwear otherwise.
Counters: 1858Code:Melee Damage Taken Player Melee Dmg Melee % Hit/Miss M.Low/Hi M.Avg #Crit C.Low/Hi C.Avg Crit% Diasetsu 93681 100.00 % 1075/4002 38/126 86.71 9 100/164 138.56 0.84 % Defenses Passive Defenses Player Evasion Evasion % Parry Parry % Counter Counter % Intimidate Intimidate % Diasetsu 2144 42.23 % 0 0.00 % 1858 63.35 % 0 0.00 %
Times Hit: 1075
Counter Rate: 63.35%(66.68%)
CI: (1.74%) 61.61%-65.09%
CI/95%: 64.85%-68.51%
Melee Gaiters appear to add 5% to counter under Counterstance and only need to be equipped at time of JA activation.
Counter Rate is 67% or +50 counter over gear, merits and trait.
Test 5: 99MNK/DNC Counterstance (Gaiters +2)
Counters: 2283Code:Melee Damage Taken Player Melee Dmg Melee % Hit/Miss M.Low/Hi M.Avg #Crit C.Low/Hi C.Avg Crit% Diasetsu 110405 100.00 % 1082/6215 40/135 101.65 9 82/192 148.22 0.83 % Defenses Passive Defenses Player Evasion Evasion % Parry Parry % Counter Counter % Intimidate Intimidate % Diasetsu 3932 53.89 % 0 0.00 % 2283 67.85 % 0 0.00 %
Times Hit: 1082
Counter Rate: 67.85% (71.42%)
CI: (1.58%) 66.25%-69.43%
CI/95%: 69.76%-73.08%
Melee Gaiters+2 add an additional 5% counter over Melee Gaiters, or 10% counter over using anything else for Counterstance.
99MNK with 5 merits has a 72% counter rate prior to accuracy check. 99 Spharai puts you 5% over the counter cap of 80%.
In case people were wondering, Banish and Tomahawk seem to share the same "Special Defense Down" term. I used a Tomahawk against a Corse monday night and got a "no effect" message.
Has anyone gotten around to testing the level 87 Buster JA's? (Dragon Breaker/Arcane Crest/Sepulcher/Hamanoha) at least to quantify the Inhibit TP Gain effect and -evasion if possible? Speculating that the Buster abilities stack with other forms of stat down since it says "Dragon Breaker effects wears off" not the individual -eva, -acc etc. So Inhibit TP from Penance or the -evasion from Feint could stack with the buster JAs.
Also interestingly the help text in the game says nothing about lowering the target's defense, only Accuracy,Evasion,Magic Accuracy, Magic Evasion and TP gain, but Julian's .DAT mine'd files shows "Lowers Defense" under Dragon Breaker for some reason, dunno if its a mistranslation or not.
Angon does not overwrite and is overwritten by Bilgestorm. Got a "No Effect" message when used ontop of Bilge and Bilge makes Angon's Effect wear off as soon as it's cast.
Spirits Within against Cerberus:
Seems that when using this WS against Cerberus, there is an inconsistency.Code:HP 1620 1720 1845 1955 1970 1970 1970 100 TP 200 214 213 247 262 ? ? 100 TP(Theoretical) 202 215 230 244 246 246 246 300 TP 770 803 824 953 984 839 815 300 TP(Theoretical) 759 806 864 916 923 923 923
SAM/RDM for each WS.
Removed breath gear. Tested against NQ Mobs as a control (Worked as BG-Wiki lists). Sekkanoki-ed the 100% TP WSs.
Though not listed, the order of testing is
1955/1970 HP
1620-1845HP
The far right two 1970 HP tests
Based on this, I'm guessing the "bonus" is variable to moonphase. As a note, decreasing phase is at some point up until Waxing Gibbous and increasing in percentage.