Monk 3.0 CHAKRA UNLEASHED

Quote:

---

Originally Posted by Umbrax

In a scale from 1 to 100, how accurate do you think your weights are?

---

Exact same set, different sim:

WD 14.789
STR 1.000
CRT 0.292
DET 0.205
SS 0.171

2 significant figures is "good enough" for weighting purposes in this game, and if you compare the numbers, you'll see his Critical, Determination, and Weapon Damage are the same within that tolerance level.

Quote:

---

Originally Posted by SunnyHirose

Exact same set, different sim:

WD 14.789
STR 1.000
CRT 0.292
DET 0.205
SS 0.171

2 significant figures is "good enough" for weighting purposes in this game, and if you compare the numbers, you'll see his Critical, Determination, and Weapon Damage are the same within that tolerance level.

---

Just a question, did you include materia and food in these calculations? If so it explains the difference between our two stat weights a bit. When I calculate mine I only use the base set and then calculate the difference at +100 stats so as to find out the most efficient way to meld over multiple points. This might explain the slightly higher WD and Det and also explain why my crit is slightly lower.

I used stats from the bottom of the screen, so yes. Normally I would add party buff, but I figured that's not the same comparison.

I only do +1 per stat because if you'll recall our exchange on reddit (a) I consider stat weights to be most useful as an abstraction, so they must be considered continuous, even though by now we reckon some secondaries really aren't and SS definitely isn't, and (b) considering I'm getting this from a formulative calculation and not a simulative one, it'd be bizarre not to approximate dy/dx with ∆y/∆x for very small x when the model actually allows very small x.

I'd think scaling should make +100 on crit/ss have higher weights than +1 regardless, though... whereas WD/AP/DET lean far more on damage formulas than potency estimates.

In That Other MMO where there are procs on everything, they rely heavily on simulators, and in that case 0.5% margin of error (which is basically 2 significant figures...ish) is a default "good enough" setting for most people's purposes, so I wouldn't read too much into such small differences.

Quote:

---

Originally Posted by Krindor

Preliminary MNK stat weights, will probably be updated in the future:
STR: 1
WD: 14.8312045494837
Crit: 0.291785864914651
DET: 0.209996953500461
SKS: tbh, all I can say is that it's close to equal to det but because of tiering it's hard to properly get a stat weight.

As for BiS from calculations it's this: http://ffxiv.ariyala.com/VSWY. To be noted is that both the Alex and Script belt are equally as good and can therefore be changed.

Those interested in why I'm saying that SS is close to Det can look at this image: Blue = Crit, Yellow = Det, Red = SS. To be noted as this was done with a sim some changes causes unwanted effects that can either increase or decrease the value of SS.
https://i.gyazo.com/6d4d0aae4659da6d...c441cb6e6f.png

---

http://ffxiv.ariyala.com/W3BM

This set is better than the one you posted; using your stat weights

Quote:

---

Originally Posted by SunnyHirose

I used stats from the bottom of the screen, so yes. Normally I would add party buff, but I figured that's not the same comparison.

I'd think scaling should make +100 on crit/ss have higher weights than +1 regardless, though... whereas WD/AP/DET lean far more on damage formulas than potency estimates.

---

When I run my sim I simulate the "BiS" unmelded pieces at a certain acc cap(They might change after melding), meaning that my stats end up a bit lower than what the final melded BiS does, roughly 210 total stats. After this there's 2 reasons I use +100 instead of +1, one is to see how much value each meld is as well as seeing the effect of changing pieces. The second is to get a somewhat reasonable weight on SS as each tier is on average every 31 points (Making it that only 7 points are worth less with +100) and my sim takes into account of 10ms tiering, though the result is still a mess for SS.

Quote:

---

Originally Posted by Ryythe

http://ffxiv.ariyala.com/W3BM

This set is better than the one you posted; using your stat weights

---

This indeed better than the one I posted, when simulated as well.

Quote:

---

Originally Posted by Krindor

When I run my sim I simulate the "BiS" unmelded pieces at a certain acc cap(They might change after melding), meaning that my stats end up a bit lower than what the final melded BiS does, roughly 210 total stats. After this there's 2 reasons I use +100 instead of +1, one is to see how much value each meld is as well as seeing the effect of changing pieces. The second is to get a somewhat reasonable weight on SS as each tier is on average every 31 points (Making it that only 7 points are worth less with +100) and my sim takes into account of 10ms tiering, though the result is still a mess for SS.

---

I want to reiterate that this isn't the right way to look at weighting. To wit:

- Once you meld, you are no longer using the same set. Yeah, it's not different by much, but the weights are altered.
- For weights to have any theoretical basis, you have to ignore tiering. This is a hard and fast requirement. Measuring a point farther away doesn't really accomplish this any better than a near one if you can't negate the discontinuity.
- Weights are little more than a heuristic for finding BiS, and a flawed one at that. Our reliance on them stems back to the BCoB days when all secondaries were so small and so close to linear that whatever you did with them didn't matter much. In ARR, they usually did just find BiS for you (as I understand it, EMX even based weights on averaged-out stats to tune it to item level instead of to BiS). But we haven't been working with this situation since Gordias. If you have a sim that can spit out that much data, just find the BiS and don't bother with weights.

Quote:

---

Originally Posted by natyusha

Pre Pull Form Shift into Coeurl [Shoulder Tackle] > Demo [PB] > Snap > Snap [B4B] > DK [IR] > TwS [Pot] > ToD [Elixir Field] > Boot [Steel Peak] > True [Howling Fist] > Demo [Forbidden Chakra] > Fracture > DK > TwS > Snap

everything after using a potion up to using fracture should be within the trick attack

---

What skill speed have you tried this on? I've done it several times now and I don't even come close to fitting it all. The second demo just barely makes it before trick attack falls so I'm roughly 2 seconds behind. My gcd is currently at 2.35 sec.

Quote:

---

Originally Posted by Claymore022

What skill speed have you tried this on? I've done it several times now and I don't even come close to fitting it all. The second demo just barely makes it before trick attack falls so I'm roughly 2 seconds behind. My gcd is currently at 2.35 sec.

---

Trick attack should be coming out after a ninja's 4th global which is a little over 8 seconds into a fight. Add the 10 second trick duration + animations and you basically have a 19 second window to get everything up to the fracture off. Cases where the ninja hits the boss before you do and you just can't get the fracture in are quite common. Fracture will hit at about 21 seconds with low skillspeed (570~) [2.42+2.3+2.17+(7*2.05)] with skill speed that low it comes down to making sure that you get your globals rolling faster than they do which shoulder tackle allows you to do (on any fight where the ninja has to run in). If you have higher skillspeed you should be able to fit it in on any fight with a run up.

sophia ex out

https://youtu.be/4acOtQjmIvQ

Quote:

---

Originally Posted by natyusha

Pre Pull Form Shift into Coeurl [Shoulder Tackle] > Demo [PB] > Snap > Snap [B4B] > DK [IR] > TwS [Pot] > ToD [Elixir Field] > Boot [Steel Peak] > True [Howling Fist] > Demo [Forbidden Chakra] > Fracture > DK > TwS > Snap

everything after using a potion up to using fracture should be within the trick attack

---

Why ToD before PB falls off, why not throw in another Snap, or even a True if you have the SS to make it back to GL? Mine is usually Snap in place of ToD, then after PB falls off I use ToD...