Black Mage Sanctuary - A Guide to DPS

[HaroldSaxon]

Source is FFXIV, 3.2x
Data points are various, with low/mid/high single variable

Your definition of stat weights appears leave spell speed variable by person, fight, party members, and probably other things that don't immediately come to mind. The numbers I put were the marginal dps increase to single spell damage for 1 of each of the stats weighted against the marginal dps increase to single spell damage for 1 INT.

To repeat what you quoted, then reworded to make it sound like I ignored it:

I never reworded anything of what I quoted from you. And of course spellspeed is variable by party members - just like Fey Glow used to alter the weight of spellspeed back in 2.x.

I'd still like to see a source of your datapoints. Do you have a spreadsheet with your raw data and damage formula calculations?

Stat weights are not accurate if you ignore rotation. They have always been variable dependent on fight and rotation, and usually things that are "meta" like Foe's (Bard) and Fey Glow (from 2.x) are assumed. Now, with MCH's place in the meta, foe's shouldn't be assumed and weights should be separate to be accurate. But it always assumes optimal rotation and optimal play.

Your "weights" are far too simplistic and honestly until they are backed up with datapoints and a data formula, they are as valid as the weights I linked to earlier - because none have corroborating evidence.

[StouterTaru]

Your "weights" are far too simplistic and honestly until they are backed up with datapoints and a data formula, they are as valid as the weights I linked to earlier - because none have corroborating evidence.

It is that simple for WD, INT, Det, and Crit. Changing any of them does not have any affect on your rotation, and your rotation doesn't change their weights in any way.

I could post a clutter of numbers; my data points aren't in any wonderfully organized and/or labeled manner. It was mostly min/max non-crit with fixed Det values and variable INT, then again reversed to find the trendlines. Crit hit was just min/max crit hit damage vs non-crit damage, and assumption that +1% crit damage correlated to +1% crit hit chance. Min/0.95 was compared to max/1.05 to gauge precision, and determine the base value.

[HaroldSaxon]

It is that simple for WD, INT, Det, and Crit. Changing any of them does not have any affect on your rotation, and your rotation doesn't change their weights in any way.

I could post a clutter of numbers; my data points aren't in any wonderfully organized and/or labeled manner. It was mostly min/max non-crit with fixed Det values and variable INT, then again reversed to find the trendlines. Crit hit was just min/max crit hit damage vs non-crit damage, and assumption that +1% crit damage correlated to +1% crit hit chance. Min/0.95 was compared to max/1.05 to gauge precision, and determine the base value.

Your rotation does change based on getting more spellspeed.
Furthermore, any increase in one stat increases the value of another. Have a higher crit chance? Spellspeed is valued better because more attacks = more crits. Have more attack power (int)? Then all of your secondaries get better.

Every single stat is linked, and it isn't anywhere near as simple as what you are saying.

If your data points aren't organised or labeled and you refuse to post them, then your stat weights have absolutely no meaning if you can't prove it.

It can be done that way, but the tradition that's been built up over the past couple years is that stat weights are relative to one set only (BiS).

Those are stat values, not stat weights. I agree that stat values for BiS is more accurate, and comparing values for gearsets is more accurate. On top of that, there isn't any concrete accuracy amounts out there (just approximations) as far as I know, so you can't have stat weights out when you don't know the BiS.

[Rongway]

Spellspeed is valued better because more attacks = more crits.

More attacks = more attacks.
To say "more attacks = more crits" is to say that x% of (y + 1) is greater than x% of y.
Crit doesn't care what your spell speed is.

[HaroldSaxon]

More attacks = more attacks.
To say "more attacks = more crits" is to say that x% of (y + 1) is greater than x% of y.
Crit doesn't care what your spell speed is.

https://www.youtube.com/watch?v=A4IdTQtICdo

I suggest you give this a watch. A higher amount of crit will give increased value to other stats, including spellspeed.

[MiniTyra]

https://www.youtube.com/watch?v=A4IdTQtICdo

I suggest you give this a watch. A higher amount of crit will give increased value to other stats, including spellspeed.

Your rotation does change based on getting more spellspeed.
Furthermore, any increase in one stat increases the value of another. Have a higher crit chance? Spellspeed is valued better because more attacks = more crits. Have more attack power (int)? Then all of your secondaries get better.

Every single stat is linked, and it isn't anywhere near as simple as what you are saying.

Those are stat values, not stat weights. I agree that stat values for BiS is more accurate, and comparing values for gearsets is more accurate. On top of that, there isn't any concrete accuracy amounts out there (just approximations) as far as I know, so you can't have stat weights out when you don't know the BiS.

Wat. About that.



1) The only interactions you see of other stats affecting everything else are:

AP/MP: Affects all secondaries
WD/MD: Insane amounts (I'm talking +200 MD/WD) affect the value of SS on Jobs that use Autos (So Physical Jobs)

And that's it. Every other stat is affected by itself.

2) That video was done using EMX's old Damage Formula where DET was an additive function of STR and WD, which Sunny Hirose found to be incorrect, Puros SS formula which was half broken and Valkys Crit Formula where 2 years later, again, Sunny Hirose found to be completely off target.

I'd take everything in that Video with a Grain of Salt.

3) Where are you getting More Attacks = More Crits? Wat.

More crits in the same time frame? Maybe. But whether you do 30 GCDs in a minute, or 60 GCDs in a minute, if you have a 25% Crit Chance, 25% of your total skills will be Crits and no matter what, 25% of your damage will always be multiplied by the same... multiplier.

4) You can get stat-weights without the need of accuracy requirements, as everything is proportionate to 1 point of AP/MP. You can't find BiS gear without stat-weights. In fact, stat-weights are never 100% accurate. The only accurate values you can get are Stat Equivalence values, which is what the JPs do.

You're confusing a lot of things.

[Waliel]

3) Where are you getting More Attacks = More Crits? Wat.

More crits in the same time frame? Maybe. But whether you do 30 GCDs in a minute, or 60 GCDs in a minute, if you have a 25% Crit Chance, 25% of your total skills will be Crits and no matter what, 25% of your damage will always be multiplied by the same... multiplier.

Well, it's technically true, the best kind of true.

[SunnyHirose]

Those are stat values, not stat weights. I agree that stat values for BiS is more accurate, and comparing values for gearsets is more accurate. On top of that, there isn't any concrete accuracy amounts out there (just approximations) as far as I know, so you can't have stat weights out when you don't know the BiS.

Use what nomenclature you must, but I am unaware of anyone drawing this distinction before with that exact terminology in this game's theorycrafting; cf. Puro's rant, T0rin explains weights. I don't recall anyone giving out any widely accepted weights using an artificial baseline since EMX. People have, however, done so when true BiS wasn't clear, and even with inaccurate models and it didn't hurt much (e.g. skill speed was known to be pretty bad on physical DPS, then we re-checked it, turned out it was even worse).