Differences in JP and NA/EU Damage Formulas

Unfortunately, just like society on Earth, Theorycrafting In the land of the West (I’m including you aussies here!) and the Japanese are vastly different. Our core understanding of the games mechanics are very similar – We both recognise that Critical Hit Rating has Marginal Diminishing Returns and that Skillspeed is rather Exponential. We also both recognise that Weapon Damage is the highest contributing factor for damage and potency is essentially a percentage of our base damage… Also that the Global 3s timer for all DoTs are very problematic.

Of course, this is mainly due to how basic these rules are. Everyone who plays the game will understand these mechanics someway or another… Except for the actual formulas to calculate Critcal Hit Rating, Skillspeed, and especially Damage. The damage formula is the largest difference and their understanding of Determination is completely different, including another big shocker that I suspect 99% of you would never of guessed!

Majority of the Japanese Theorycrafting work can be found on this Gentleman’s blog, who goes by the alias “neetsha”. (I think Neetsha is a dude at least…)

http://ku-so.hatenadiary.jp/

You may also find him on Twitter via this link:

https://www.twitter.com/arisue_neetsha

There’s also an included damage calculator that he coded in Javascript, which is where I robbed some of his formula’s from, hehe ty m8!

http://www.eonet.ne.jp/~versatile/ku-so/ff14dps.html

 Before we go any further, I want people to understand that this document will become outdated by the 3.0 Heavensward Expansion and maybe even before then, as I’ll be going over the damage formula over the next few coming days/weeks/months/expansion packs (lol). When that time comparison, I’ll do another comparison post… But for now, enjoy!

We’ll start off with the Damage Formula as this is the juiciest portion of this entire topic. This is also the hardest to explain… and the longest to prove… and the hardest to understand as I have to use Googles shitty translator to help me read Japanese Blogs…


The Damage Formula

The first thing we should do is break down the Damage Formula. This isn’t supposed to be 100% accurate either, but rather, the formula gave an accurate damage measurement when compared against the collected data points.  If you’ve done any reading of this games mechanics, you may of encountered this formula many times:

(WD * 0.2714745 +  AP * 1.1006032 + (DTR-202) * 0.0241327 + WD * AP * 0.0036167 + WD * (DTR-202)*.0010800 – 1) * (Potency/100)

(AP, or Attack Power, = STR/DEX/INT)

Of course, you have to take into consideration order of operations (BODMAS) when using this formula to calculate expected damage.

 Let’s take it apart. We have a few sections which are all independent of each other:

  • WD*.2714745 – First, we’re seeing the impact of Weapon Damage on Damage on its own
  • AP*1.1006032 – Then add the value Attack Power has on its own
  • (DRT-202)*.0241327 – We do the same for Determination. We do DTR-202 to see the impact of Determination from our base-value at level 50. I’ll go more in depth why this is later on.
  • WD*AP*.0036167 – Next, we’ll add on the impact Weapon Damage has on Attack Power
  • WD*(DTR-202)*.0010800 – Plus the impact Weapon Damage has on Determination
  • -1 – God knows why the solver which included the original data-points added this in.
  • (Potency/100) – And as Potency is a percentage of damage, we divide it by 2, and multiply it by the overall base damage.

 Time for some explanations.

A long time ago, if you’ve been playing since the Beta Phases and/or launch, you may remember a person going by the name “Valky”, who had a blog we call B.L.I.T.Z.B.A.L.L. We owe much work to this gentleman has he set the core foundations of everything we know about this game – this is also the same on the JP side.

He theorised these core rules:

  1. A change in Weapon damage alone will affect our Damage
  2. A change in Attack Power alone will affect our Damage
  3. A change in Determination alone will affect our Damage
  4. The both the value of Attack Power and Determination are independent of each other and do not affect each other.
  5. Weapon Damage affects the value of Strength and Determination.

http://valk.dancing-mad.com/methodology/

His original stuff, whilst good, was inaccurate when accounting for high levels of Determination. Then a guy we all know as EasyModeX, with the help from people such as T0rin, Kenji1134 and Ein to gather data points, updated the formula to what we currently have.

 As a result, this formula has given us even more interesting rules about how our stat’s work, but I’ll go over that later as well, so we can do a nice comparison with Japanese understanding of stats.

 The core understanding of Weapon Damage, Attack Power and Determination are vastly different than the Japanese, which also means our Damage formula’s are waaaay different.

 The most important thing about our Formula is that we value all of our stats to be rather independent – But what I feel is the most important thing is our calculation of Determination. We assume that Determination a value which provides us Linear Gains, however, the value or the weighting does not follow typical Linear gain rules, such as Critical Hit Rating. Linear Gains usually result in Marginal Diminishing Returns, but that simply isn’t the case with our formula, as observed in these two images.

 

 



The Japanese Formula

((WD * 0.26 + AP * (WD * 0.00366 + 0.0745) + 4.95) * ((DTR – 202) * 0.0005 + 1.00) * job * (Potency/100)))

 Above is the Japanese damage formula.

 Now, if I’ve been reading those poorly translated blog posts correctly, the Japanese assume that the value of Determination is derived from the impact your Attack Power. Because it’s a measurement of the impact of Attack Power, rather than the value of Attack Power itself, Weapon Damage is also a contributing factor in this.

 Not only this, their formula assumes that Determination follows your typical Linear gains with marginal diminishing returns, similar to that of Critical Hit Rating.

 http://puu.sh/gHWK8/febe8b31e8.png

 http://puu.sh/gHWKZ/896ab2b7f9.png

 As you can see, with an increase of 300+ Determination, the value of Determination decreased.

 Because of this relationship between Determination and Attack Power, their formula is broken down into two sections.

 Section 1:

  • WD * 0.26 – Impact of Weapon Damage
  • +
  • AP * (WD * 0.00366 + 0.0745) – Impact of Weapon Damage on  Attack Power

 Section 2:

  • (DTR – 202) * 0.0005 – Correction Formula of the value Determination has
  • + 1.00 – Because why not?

One thing I’ll point out now, is that the formula uses something we call a “correction formula”. Essentially, we’re guessing the impact of a variable from it’s maximum base level – in this case, we’re the guestimating impact determination has on damage from it’s base value at level 50, 202. This is why we do DRT-202. This also applies to both NA/EU and Japanese Formula and as a result, as our values of Determination increases, we’ll slowly have issues with both formulas and will need to be reworked, again.

 Both of these sections are then multiplied together and… Voila. Damage. But, that isn’t the entire formula… what you might of noticed is the ( * job) coefficient in the formula.

 The Japanese have a theory that every job has a different coefficient that scales up the damage slightly. In other words, if every job has equal Weapon Damage, Attack Power and Determination, each job will deal different values of damage. This originates to something Valky noticed a few eons ago.

 Back in Phase 3 when Valky was creating the Damage formula, he noticed that different jobs were dealing different damage values above his own calculations. He formulated everything on his monk and the damage formula worked perfectly! However, the damage values for the Dragoon and Bard always ended up slightly greater than his formula.

 This was eventually refuted by a lot of us on the NA/EU side, but the Japanese investigated this even further.

Job Modifiers
DRG/BRD 1.046875
MNK/NIN 1.03125
WAR/LNC/ARC 1.015625
MRD/ROG 1
PLD 0.98437
SMN/SCH? 0.96875
WHM 0.859375
BLM 0.828125

Above is a list of all of the Jobs and their damage modifiers that are used in the Japanese Formula. All you do is add the corresponding job coefficient into the formula and… It’s done!

Reading this kind of surprised me. I originally thought that each job did equal amounts myself, and that Valks Data was just due to rounding errors on his behalf.

One benefit of this that I can see, is that it makes job balancing exponentially easier. Instead of having to worry about job abilities constantly and their rotation performance, they can just alter a simple job coefficient… and job done!

I’ve also done some investigating about this myself and I’m still currently gathering a lot of data. I’ve made a write up in my blog about this, which you can find at this link:

https://dervyxiv.wordpress.com/2015/03/18/jobclasses-their-impact-on-in-game-damage/

So, we can derive these rules from the Japanese damage formula:

  1. Weapon Damage is the highest impact of Damage
  2. Attack Power is directly impacted by Weapon Damage
  3. The damage formula has been simplified so it doesn’t have two different variables of Attack Power impacting damage,  (AP) + (WD*AP)
  4. Determination doesn’t have a value of it’s own, and is a resulting factor of the impact of damage Attack Power has
  5. Determination is 100% linear with marginal diminishing returns.

 


Stat-weight Comparisons

If you want to learn more about stat-weights, I suggest you go and read up on this post:

https://dervyxiv.wordpress.com/stat-weights/

Now as I’ve already explained, there are differences in the calculations of stats on the NA/EU and JP side. As a consequence, there will be different stat-weightings for each job.

Let’s go back to our original understanding of values on the NA/EU side.

  • We assume that Weapon Damage affects the value of Attack Power.
  • We assume that Determination and Attack are independent of each other.
  • We assume that Determination, whilst having linear gains, it’s a static value only affected by Weapon Damage
  • Due to this static value of Determination, an increase or decrease in Determination will never change value. It’s only affected by Weapon Damage
  • Therefore, we can make the statement that Determination works similarly to our Attack Power.
  • A large WD:AP ratio will result in a lower Weapon Damage weighting. This is why Black Mages have 6.XX weighting for Weapon Damage.
  • Both Weapon Damage and AP/STR directly affect Critical Hit Rating and Skillspeed. If there’s an increase/decrease in Weapon Damage and AP, Critical Hit Rating and Skillspeed will follow suit.

So pretty much, once we’ve calculated our stat-weights for a certain ilvl range, they’re fairly accurate and don’t budge, as long as you’re calculating weights within the baseline.

So as a comparison, here’s some Dragoon stat-weight calculations using the NA/EU and JP Damage Formula.

NA/EU Stat-Weightings

JP Stat-Weightings

As you can see, the Japanese Value Determination much more, as it’s directly affected by the STR*WD value. The higher the STR*WD value, the greater DET will be… And inversely, the lower your Strength, the lower your Determination value will be:

The change in Weapon Damage, Skillspeed and Critical Hit Rating are expected and follows through with what we already know about stats, similar to that of NA/EU.

But what happens to Determination? I’ve already explained how it works in NA/EU land… So, let’s take a look! I’ll create the most extreme scenario and increase Determination by a whooping 300+

NA/EU +300 DET

JP Changes + 300 DET

Japanese version of Determination, as I stated previously, follows typical linear gains. Meaning, it has marginal diminishing returns when stacked up. Yes, the damage value is exactly the same per point of determination gained as it’s linear, but the percentage of damage decreases.

Okay, so what about if we increase Strength by 300?

NA/EU +300 STR

JP +300 STR

Okay. So as I explained earlier, when the WD:AP ratio is large, the weighting of Weapon Damage decreases. But as we’ve gained a lot of AP, this ratio is much lower. Therefore, the weighting of Weapon Damage increases. Similarly, as Attack Power has increases, the values of Skillspeed and Critical Hit Rating also increased. Determination on the other hand?

Well, there’s a difference. Where as NA/EUs Determination is static/constant, the Japanese Determination will always fluctuate depending on the value of Attack Power. As we have a much higher Attack Power (in extreme measures), Determinations value has exponentially increased.

The same can also be said if we look at an increase in Weapon Damage:

NA/EU +300 Weapon Damage

JP +300 Weapon Damage

And of course, with an increase in Weapon Damage, Determination for both NA/EU and JP increases.

I can seriously go on forever about this, but it’ll be too long. So I’ll end it here and continue later on about Japanese formula’s in another blog post.

If you’re interested in seeing all of the Japanese Formula’s, you can take a look at this spreadsheet I made:

https://docs.google.com/spreadsheets/d/1jShiXKLE5ro93Lqfj25PrLCJPr7-1Z9j9ow7dO-Syvc/

That document should have every single Japanese Formula known and all resources are posted in the “resources” tab.

Differences in JP and NA/EU Damage Formulas

17 thoughts on “Differences in JP and NA/EU Damage Formulas

  1. Kaesebrezen says:

    I’m posting this here, because i lack the classes to test things out, but at least on my scholar i can reduce damage numbers and healing numbers to – because it simple works:

    Weapon Damage * (1.0 + Potency/Coefficient/100) * (1.0 + Det/56/100).* Coefficient.

    (http://www.reddit.com/r/ffxiv/comments/2zpk11/comparison_between_naeu_and_jp_damage_formula/cplzg2m
    http://www.reddit.com/r/ffxiv/comments/2zpk11/comparison_between_naeu_and_jp_damage_formula/cpm02ss)

    My general idea is that:
    – Crit and DET are balanced against each other, i.e 28 CRT = 1% Crit chance = 0.5% damage increase -> 28 DET == 0.5% damage multiplier
    – Potency behaves similar to WoW-Mastery, i.e. it might feature a spec-specific (or at least role-specific) conversion rate.
    – Formula: Weapon Damage * (1.0 + Potency/Coefficient/100) * (1.0 + Det/56/100).* Coefficient.

    This has several major benefits for DEVELOPERS:
    – Assuring that CRIT and DET are in a reasonable competition and balanced.
    – Balancing specs against each other can be done by changing spec conversion of their main-potency, without affecting DET/CRIT balance.
    – It’s easy to understand, easy for the server to calculate.
    – Overall system is VERY similar to WoW (this would make sense)

    Like

    1. The main question is this, how is Determination affected by Weapon Damage and Attack Power (Or Potency)? This relationship between Determination, Weapon Damage and Strength is what really decides the structure of the Damage Formula.

      If determination is truly a modifier, or a stat that is affected by the values of both Weapon Damage and Attack Power, then I can see the formula resembling what you’ve posted, or something like:
      – Weapon Damage*A * (Attack Power*B) * (DET*C) * Job Coefficient * (Skill Potency/100)

      B and C values could either be a single Coefficient, or something like *X+Y, or as you put it, /X/Y

      But, if Determination is affected by Weapon Damage and not Strength, then it’ll have to be sectioned into two parts.
      – (Weapon Damage*A * (Attack Power*B)) + (WD * (DET*C)) * Job Coefficient * (Skill Potency/100)

      At least, that’s my impressions of it.

      Like

      1. Kaesebrezen says:

        I honestly think DET is a simple multiplier – Potency on the other hand is the odd stat.

        Just looking at the base stat table. DET and CRIT are equal across all classes – Potency is the one stat that is not. If DET and CRIT are equal across all classes, you’d expect that it affects all classes equally. The best way to achieve that is to have it be a simple multiplier.

        Potency is the odd sheep. I’m taking healing spells as they have one very big advantage here: Cleric Stance. You can alter your potency without changing any other stats.

        77 WD – 377 DET – 602/223 Potency.
        Physik: ~1420
        PhysikCleric: ~480

        While Cleric stance applies a 20% healing potency penalty, this can only be applied at two places (theoretically there are more, wait for later in this post): A final 0.8 multiplier, or a 0.8 multiplier bringing the 223 Potency down to 178.4

        Taking the Japanese Value for DET 377 = 8.75% and a 1 time det multiplier.

        Supposed Case 1: Cleric stance is a final 0.8 multiplier.
        Potency increased by 170%, healing increased by 200%.

        Supposed Case 2: Potency is multiplied 0.8 first.
        Potency increased by 238%, healing increased by 200%.

        And looking at that – neither of these two cases is good. Both are equally bad. One means Potency affects potency, the other is a massive diminishing return on potency.

        The only way i can even get to 480 healing is by applying the 0.8 three times (using my formula):
        1) Multiply weapon damage//spell coefficient
        2) Multiply your total potency
        3) Reduce healing potency conversion to 80%.

        I’ve another idea on how potency might work, but that’s pretty warfetched (but it would actually fit with cleric stance – and a possible behaviour of crit, det and base stats):

        There is a global “base” value for every stat in %. You will always have 12% crit, always have 3% det, and always have (for example) 400% ” healing potency”. Base Crit and Det are equal across all classes, so their conversion is equal for all classes. Class potencies are different across all classes, while the global value is equal (look WoW Mastery). Now you take your global 400% value and divide it across your base potency. Then you get your class-specific conversion rate.

        For Scholar this would be at 205 Base Healing Potency a conversion of 1.95, while Whitemage is looking at 1.632.

        Using my formula with a 1.632 Potency conversion – i actually get the correct value on my level 32 Whitemage.

        Like

      2. Do you have any parsed logs of you casting Physick on your 77WD Scholar with and without Clerics? I’d like to take a look at your parsed numbers. That’s actually rather interesting. I can’t do much as I don’t have a Scholar (or any healer for that matter) at 50, but your findings may be very, very helpful.

        Like

  2. Kaesebrezen says:

    I sadly kind of new to FFXIV, been playing 1.5 months now, and one month at ARR release (but not theorycrafting – pretty experienced with that on WoW).

    I just got ACT and am trying to figuring out how to expert the data from there, but meanwhile some better average numbers for all 3 heal spells taken from ACT

    Physik: 1409
    Adlo: 1060
    Succor: 537

    Now reverse engineering that with my stats and accounting for Maim and Mend at 30%:

    1409 == 4 * 1.3 * 77 * 1.0875 * Potency -> Potency = 3.23
    1060 == 3 * 1.3 * 77 * 1.0875 * Potency -> Potency = 3.24
    536 == 1.5 * 1.3 * 77 * 1.0875 * Potency -> Potency = 3.28 (Succor numbers take a lot longer to generate due to MP – so this is largely the avg is on much less casts)

    Healing Potency of 602 would then equal an increase of 224% – or 2.7 potency for 1%.

    If i apply that potency value to Cleric Stance, it’s impossible for me to get the logged numbers, regardless of how i place it (487/370/181 – but all of them share a identical Potency), that’s why
    I’m currently switching on how Cleric Stance actually works:

    You have a Healing Multiplier determined by Healing Potency (== Mind) and a Magic Damage Multiplier determined by Magic Potency (== Int). All Cleric stance does is to swap these values – and then adds anothe 1.1 multiplier to damage done, and a 0.8 multiplier to healing done.

    Very Important Note/Observation:

    Cleric Stance tooltip specifically says it “swaps INT and MIND — RATINGS–“. Ratings in WoW terminology have a unique conversion factor associated to each rating. If Cleric Stance does in fact swap ratings, then it’s pretty clear why the following actually works:

    Energy Drain with a Potency of 150 deals the same damage out of cleric stance (with 223 Magic Potency) as Succor with a Potency of 150 heals in Cleric Stance multiplied by 0.8 (now 223 healing potency – the previous 223 magic potency with it’s associated conversion factor).

    (YoshiP is looking on many other MMOs and how they do things – i personally wouldn’t be surprised if they are pretty close to WoW in calculations. WoW over time switched from additive back in Classic/BC to pretty much everything is multiplicative.)

    I’d say Cleric Stance is actually quite a helpful mechanic in this discussion – as it might tell us something that’s not so clear otherwise.

    Like

    1. You pretty much described how Clerics Stance works – Which we all ready know. The “Ratings” are the numbers derived from your Attributes, Strength, Dexterity, Intelligence and Mind and they literally have a 1:1 affect on your Magic/Healing Potency, as well as Attack Power. Right now I have a Strength Value of 653, with an Attack Power of 653. It’s just a much more simple way of managing all of our Stats. This is also precisely the reason why no one uses the terms “Healing Potency” or “Magic Attack Potency”, because they mean exactly the same thing and act exactly the same.

      All Clerics Stance does is swap Healing and Magic Attack Potency, then adds a 1.1/0.8 modifier on top. It’s very similar to how abilities like Defiance works on a Warrior, Shield Oath for Paladins, or Venoms for a Ninja.

      I might need to actually level up a Healer now to do some investigating, because you’ve brought up some very important points, especially about Determination, which I need to consider.

      One thing, when you say potency, make sure you’re stating whether it’s Healing/Magic Potency, or Skill Potency. You use the term “Potency” to talk about all 3, and it kind of makes it harder for my brain to process, especially this early in the morning for me haha.

      As for recording logs in ACT. ACT allows you to break down Skill/Ability or Healing Ability output as well. All you do is right-click on the log which displays all of your healing/damage values for that ability, click “Copy as Plain Formatted Text”. Then drop it into pastebin or something.

      Like

  3. Kaesebrezen says:

    I decided to get a step back today and fully understand the base mechanics today.

    Critical Hit Rate:

    Base Critical Hit Rate is 5%.
    For a level 1 char you have that at 56 Critical Hit Rate.
    For a level 50 char you have that at 341 Critical Hit Rate.

    Now i looked at http://www.bluegartr.com/threads/120327-Critical-Rate-Testing-%28And-other-stuff-%29 and the general consensus is that CRIT increases linear.

    Attack/Magic/Healing Potency Increases:
    Keep WD and DET constant – alternate Primary Stat – You’ll see a linear increase. (PrimaryStat multiplier somewhere)

    Now what i’m lacking is ( i don’t have weather daystar and scylla gear to swap large amounts of DET and constant MIND) data on Constant WD – Constant Primary Stat – Alternating DET.

    I tried doing that with buff food – but that’s incredible small jumps of 4,8,11 up to 17 det. While with these values i also get a linear increase, it’s just an extremely small DET range. (I used DET-only buff food because that’s also allows naked testing for DET)

    Luckily as a Scholar Constant DET – Constant Mind – Alternate WD is extremely simple due to <30 weapons. (This is actually the only of the tree setup that returns R^2=1 on a linear trendline for me)

    Class-Multiplier: Why would i need them?
    Weapons already feature different damage for several classes. All classes have a different base value in their Primary Stat, so this can already be taken as some kind of class-specific modification (compare your and the japanese ranking to a ranking created by sorting primary stats of all classes – there are similar trends – how similar are they?)

    Base Value of Certain Stats:
    – Why is Crit at 341? Why is Det at 202? Why at level 1 do i have Crit/Det = 56/20 = 2.8 and at 50 i have 341/202 = 1.688. Are these value random or do they actually carry a deeper meaning (i.e. they are used in some calculations, even be it as simple as stat conversion)

    Especially the last two points are somthing i really want answers to – I want to know if there is a reason for 341 CRT, 202 DET as a base value.

    Like

    1. If you’re trying to understand some of the base mechanics, I’d also go ahead and read http://valk.dancing-mad.com/methodology/. We don’t use his Damage Formula as it’s inconsistent with Determination, but we use his Critical Hit Formula still to this day. If I remember correctly, Critical Hit starts off at a base 5.33%, or something like that. Each point of Critical Hit Rating you gain increases your PPS (Potency Per Second) by 0.05318 Raw PPS.

      As for Class/Job Multipliers, I can see why SE would do this: Balancing. If every Job never had the multiplier, (or a multiplier of 1 for arguments sake), the Dragoon would never be able to compete with the Monk or Ninja. As an easier way to balance out jobs, they just need to change a single multiplier and that’s… that. Easy. As for my Ranking System, I’m still sorting it out, but there’s a lot of rounding errors I’ve encountered so far. I need a lot more high-potency data to really split the jobs/classes apart. For reference, the Japanese ranked the jobs using very, very low potency data.

      Like

  4. Kaesebrezen says:

    New day, new thoughts:

    Valky has a decent amount of talk about “What if WD = 0” talk – which is a combination about your weapon breaking and you still dealing damage, and then concludes that there is some +STR term in the formula – but has it ever defintivly proven that there does not exists a base physical or magic damage value defined by level?

    Taking the na/eu formula and rewrite it to the following:
    (AP*0.0036167+(DET-202)*0.00108 + 0.2714745)*WD + AP*1.1006 + (DET-202)*0.024 -1

    Why can’t it be something like
    (AP*rate + Det*rate) ( WD + Base Damage Value) == AP*WD + Det*WD + number*AP + number*DET

    So has it been proven that there is not a base magic or physical damage value defined by level?
    Wouldn’t breaking a weapon take 30 deaths (is there a faster way to do that?) i would test it myself (get naked, broken weapon – then all ability damage done has to be (AP*1.1006-1)*coefficient)

    I rewrote the reorganized the formula a little bit:
    (AP*0.0036167+(DET-202)*0.00108 + 0.2714745)*WD + AP*1.1006 + (DET-202)*0.024 -1

    And then wondered:
    What if it is ((AP-BaseAP)*Conversion + (Det-BaseDet) + BaseValue of AP + Base Value of Det) * (WD + Base Level damage)

    So the “0.27*WD” term that is described as “impact of weapon damage” would be replaced by “BaseAPValue + BaseDet” value. Then you add the value of Det and AP to the conversion.

    Lastly – are there bigger active sites/forums where FFXIV theorycrafter gather? I’d really like to get a lot more information on pretty much everything. (i have that BLM google docs damage sheet i’m lookng through right now to get a better understanding for some classes)

    Like

    1. Anonymous says:

      Unfortunately, no. Majority of the hardcore theorycraters left the game. At launch, there was a group who was going to release a website of such, but they also quit. Such a bummer.

      Also. I’d take that Black Mage spreadsheet with a grain of salt. We all had arguments with that fellow for weeks and weeks and never managed to get around to him why a lot of his stuff is simply inaccurate.

      Like

    2. Kaesebrezen says:

      Onto reordering the formula’s again: This time for the japanese one

      You can do these two things
      1) WD(0.26 + 0.00366AP) + 0.0745AP + 4.95
      2) AP(0.075 + WD*0.00366) + 4.95

      Looking at formula 1) – Here you can do a vastly different interpretation:

      You have WD multiplied by some AP term. The AP Term could be interpreted at (BaseAPValue(ClassBaseAP) + (AP-ClassBase)*conversion) – whereas the job multiplier now could be direct result of classes having different base AP values – and you’re overshooting on AP (or going too low)

      Looking at formula 2):
      There isn’t much room for interpretation here. It’s just bringing me to the next point (and links to my previous post)

      Both reorders have a remaining 4.95 – there isn’t much to say aside of that the japanese assume the existence of a raw base damage value (which das not exist in the NA/EU formula – unless you count the -1 as a remnant of the actual value that kind of disappeared along the way)

      Like

  5. Kaesebrezen says:

    Looking at the Stat Weights – the sole reason why japanese ranks DET so high is because their DET conversion is so stupidly high.

    13.5 CRT ~= 1% Crit ~= 0.5% DPS increase ->
    1 CRT ~= 0.037% DPS increase
    1 DET ~= 0.05% DPS increase in the japanese formula.

    That’s effectivly 33% stronger than CRT is. I really want to know how they ested DET, especially if they assume it multiplies, since…

    Testing DET with low WD – > bad.
    Testing DET with low Primary Stat -> bad.
    Testing DET with low Ability Potency -> bad.

    Low number * even lower multiplier + variance/rounding and you can’t really see much.

    The only class/spell that seems to be remotely useable to test DET is WHM + Cure II with it’s lovely 650 Potency.

    And even then – to actually calculate the real conversion from that in case it’s a multiplier – you need the exact base damage. (If base det = 1.0 then that’s easy, if it’s not, then it’s getting a lot harder)

    Like

    1. http://forum.square-enix.com/ffxiv/threads/79042-Dragoons-A-Rotation-Reborn/page555

      Here’s something you might like from myself and Sunny Hirose. Maybe you can work some of your magic here?

      2 things we’ve worked out over the past week:

      1) Base Weapon Damage IS 0, therefore there is an existing STR*WD + STR

      2) Determination IS affected by both WD and AP

      3) Affect of Determination scales very strongly with Strength

      Now, the main issue really is being able to stack enough determination whilst keeping a constant Attack Power/Healing & Magic Potency to actually have a noticeable difference in damage if I’m honest… Sunny Hirose was only able to gather a few Data Points as it’s so hard =/

      Like

      1. Kaesebrezen says:

        I did something similar with these numbers as i did when testing with constant det (altering WD/Potency), and it behaved exactly like them.

        My general opinion is now: WD * DET * Potency + X *WD + Y * Potency + Z * DET

        Just the biggest gripe i have is that X*WD + Y*Potency + Z*DET could always be a result of some final added base damage value that exists because of WD=0 being possible

        You always can solve (positive) Number = x*a + y*b + z*c – so whatever +WD + POT + DET term there is the the formula could be a direct result of some existing base damage.

        The hardest – yet probably most important test – is to go fully naked, break your weapon, reset your bonus attributes, race change to the neutral race for your chosen primary stat (i’m unsure wether racial is always a benefit, i.e. +0 – +6 or we’re actually looking at -3 to +3 – but Au’Ra might help here) – i.e Wildwood Elezen for Heal, Dunesfolk Lalafell for STR damage, Hellsguard Roegadynn for Dex-Damage, Seawolf Roegadynn for Int-Damage, and finally unequip your job stone.

        And at that point you alternate potency and DET (det altering works okish with no-primary attribute right side gear – you have a pretty big pool to choose from)

        Like

  6. With 0 WD, 0 STR & 202 Det, base damage appears to be 4.75ish, on a Dragoon at least.

    0.0846*STR + 4.7483 is the model I currently have for calculating damage with 0 WD, 202 DET. I don’t think race would really impact anything to be honest with you. It’s just +6 strength, exactly the same as equiping a right side piece.

    Would it make sense to you, if we were to get a very accurate formula going to calculate the damage with alternating STR/WD (with 202 DET), then afterwards reverse engineer DET into the equation and observe at which points DET is influencing the damage formula?

    That way, our STR/WD damage model could serve as the control to calculate the scaling of DET. It could seem like a lot of effort, but I guess this will be a definite way to how the hell DET actually works.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s