I thought it would be interesting to observe the linearity of stats going from level 35 -> 60. I’m only collecting data every 5 levels, so 35, 40, 45, 50, 55 and 60.
So far, I’ve managed to collect Data for levels 35, 40 and 45. Here are the results and some are rather… Interesting to say the least.
Here’s an update to the list! Now included Ninja and Summoner!. I’ll get to work with Machinist and maybe Black Mage in the near future.
Dragoon:
WD: 10.625
STR: 1.000
DET: 0.139
CRT: 0.162
SS: 0.104
Monk:
WD: 10.714
STR: 1.000
DET: 0.139
CRT: 0.167
SS: 0.116
All Formulas are up in the “Damage Mechanics and Formulas” page. If you click on the 3 lines on the top right corner, it’ll open up a menu which you can use to navigate my posts. Anyway, here’s the current and most updated damage formulas:
Critical Hit Chance
((CRT-354)/(858*5))+0.05
Critical Hit Damage
((CRT-354)/(858*5))+1.45
Skillspeed on GCD
2.50245-((SS-354)*0.0003776)
http://www.bluegartr.com/threads/125988-Critical-Hit-Rate-3.0-Bardshrekt
Critical Hit Chance
I saw this amazing open source google.doc on the blue garter forums which had A LOT of Critical Hit Data, so I gathered up my own Data-Points, then asked Viridiana, who I know has been gathering hundreds of Data Points, to add his to this google.doc… 354-760 Critical Hit Rating almost. Fantastic.
Continue reading “3.0 Critical Hit Chance & Damage Formulae”
https://www.reddit.com/r/ffxiv/comments/3ayt9l/spellskill_speed_is_no_longer_worth_0000952s/
Mikal Mirkas has done an absolutely beautiful analysis on Skillspeed and came to the conclusion that there’s a 26-27 flip-flop each new “Skillspeed Threshold”, averaging out the Skillspeed shift to 26.5.
However, the first few points of skillspeed always seem to scale Skillspeed differently. As a precaution, I’m going to use the Japanese base GCD of 2.50256, giving us a new Skillspeed formula of:
2.50256-(0.01*(Skillspeed-354)/26.5)
Next, we know that Skillspeed Affects DoT damage. I don’t have many Data-Points for this unfortunately, but they’re roughly accurate and are tied to my Damage Calculation Formulae.
It’s pretty tight, but it looks like the Skillspeed Mod is approximately:
(1+(skillspeed-354)*0.000138)
If you’ve read my previous post about the Damage Formulae, this Skillspeed Mod will fit under the buffs section when calculating damage. Alternatively, you can multiply it by the DoT potency. It’s the same effect.
Okay, here it goes. SquareEnix looks like they may of changed more than just the stats themselves. It appears there’s no longer a base damage value which my old model used to plus on top of the Strength modifier, after it’s been multiplied by 0.075.
This change to damage changes a lot of things. For example, the Strength Coefficient is now 0.011, and the Determination Coefficient is 0.0001368, which makes Determination scale almost 4.5 times less than 2.5X for Weapon Skills.
So, here’s the Updated Models, which, are a WIP:
Weapon Skills:
((Potency*0.01005)*(WD*0.0400573+1)*(STR*0.1104171)*(DET*0.0001368+1)*BUFFS)-1
Auto-Attacks:
((WD*0.0319937+1)*(STR*0.1387779)*(*DET*0.0001513+1)*BUFFS)-1
Skillspeed Mod for DoTs:
(1+(skillspeed-354)*0.0001285) – Note, this fits into the Buffs section.
GCD Calculation:
I read through a few sources about GCD on Bluegatr, Reddit and the JP one and now for GCDs I’m using:
2.50256-(0.01*(SS-354)/26.5)
The reason for 2.50256 is because the first 6 values of Skillspeed are valued differently. The Japanese use the 2.50256 number to by-pass the modifier for the first 6 values.
Of course, my Damage Models only work for the Dragoon and are still a WIP, so a few numbers will change.
Also, I spoke to Clorifex (Garlandtools guys) and he mentioned that the game uses whole numbers in every formula he’s encountered so far. Oddly:
1 / 0.0400573 = 24.96426889
1 / 0.1104181 = 9.056486629
1 / 0.0001368 = 7308.289367
There’s a good probability that the game actually Divides WD/STR by whole numbers, rather than multiplying by long decimals.
So, when I rounded these values, I got this:
((POTENCY*0.01005)*(WD/25+1)*(STR/9.05)*(DET/7308+1)*buffs)-1
This only gives me a 0.2 variation error over my model which multiplies values by decimals.
Total cumulative variation error is 3.8871, which is good, considering I still haven’t a clue how the Game rounds up numbers.
4 days until Heavensward hype!
It’s been a while since my last post. There’s been a few numerical changes to my Dragoon model, but nothing dramatic. My Model is more accurate now, so that’s something I’m happy about, but that’s not what I’m here to discuss.
A while back if you may remember, I discussed Job Coefficients, or at least, a possibility of Job Coefficients.
Well, I’ve spent a very bloody long time recording data from a Dragoon, Ninja, Paladin and a Warrior and plotted them all onto a graph to see the linear regression of DMG vs STR, given 51 WD and 224 DET.
There’s 5 plotted lines, 4 of which are my parsed data points of the 4 jobs. “Dragoon Expected” is the linear regression of my Weapon Skill model for the Dragoon and as you can see, it’s very consistent with my parsed Dragoon data.
This graph is further proof that each job scales Strength differently. But, what’s even more alerting to be is the differences in the “+1X.XXX” part of my linear slope. This is the base damage given 0 Strength, or the combination of Weapon Damage, Determination, and our base damage value. For the Dragoon, this value is 4.49ish.
The Japanese blogs however, with their AA testing, state that Determination scales similarly between all jobs. If this is correct, then there has to be a vast difference in the base damage of each job and/or Weapon Damage.
Next update, hopefully before Early Access, will be a similar Graph at 0 WD. This should provide me with base damage values and hopefully, we can see exactly what’s going on with our jobs.
Valky had a Theory back in Beta Phase 3. His assumption, if he was correct, was that the only stat that scales differently between Auto-Attacks and Ability Attacks was Determination. Our original formula which was made by EasyModeX, as well as the Japanese Formula, were both constructed with this in mind. But, this rule does not fit my data samples what so ever.
Let’s look at the solver I’m using, and yes EMX, this is your original solver, albeit modified slightly.
If you managed to make it through my previous Determination Update post, you would of seen that I posted an updated Ability Damage Formula for the Dragoon.
((WD*0.0402502+0.9209679)*(AP*0.0447422+2.6089895)*(DET*0.0011331+1.8536433))
I wanted to streamline it a tad bit more. I was messing around with a few more data sets, and I realised that that the +X parts of Weapon Damage and Determination were fairly static. They were always accurate to the thousandth decimal place. I also noticed how close these numbers actually were to 1… And I thought to myself… “are they just multipliers and nothing more?”
Oddly, my assumption was accurate.
((WD*0.0437052+1)*(AP*0.0763653+4.4581293)*(DET*0.0006116+1))
Or, a rounded version of:
((WD*0.044+1)*(AP*0.076+4.46)*(DET*0.00061+1))
This formula above streamlines all of the excessive numbers I used and also provides me with a damage variation exactly similar to my previous formula.
I still need to gather low AP/High DET data samples and run it by this formula, but it’s looking like this is the expression I’ll be sticking with.
Now that I’ve established the final Expression will be (WD*A+1)*(STR*B+C)*(DET*D+1), the next thing I’ll be doing is continuing my work on Job Coefficients. As we all know, each job affects damage differently and what I’d like to do next is find out whether jobs affect the value of stats individually, or if it’s just another external coefficient.
It’s been a while since I did my last update. I thought I’d share some of my work thus far. I’ve spent the last couple of weeks gathering additional armour pieces so I can eventually have a massive spread of data for Determination information. I’m also working on obtaining more Bard, Ninja and Warrior FCOB armour pieces, so I can assess Job Coefficients more carefully and whether Jobs add an additional Coefficient to your Main Stat, Weapon Damage, or both (as a multiplier at the end).
Anyway, here’s what I’ve got so far.
Dervy does Theorycrafting 