The Monk Temple

[pandabearcat]

Is the dps loss you stated - 1.4% - regarding only situations where you're not putting up your dots?

If you put up dots the dps loss is even lower, you will never lose Twin for a coeurl move (unless you had to disconnect, in which case you should twin for that), so the only attack you will usually be in danger of losing twin for is the 2nd twin (3rd Raptor attack).

A loss of ~14 pp is worth it for the TP efficiency gain in my opinion, but it will depend on the fight and your comp of course.

[Allyrion]

If you put up dots the dps loss is even lower, you will never lose Twin for a coeurl move (unless you had to disconnect, in which case you should twin for that), so the only attack you will usually be in danger of losing twin for is the 2nd twin (3rd Raptor attack).

A loss of ~14 pp is worth it for the TP efficiency gain in my opinion, but it will depend on the fight and your comp of course.

What I meant is that fracture/ToD should make you lose twin on the next dragon kick. I guess you wouldn't be using fracture though when you're going for TP efficiency.
Is ToD not enough to lose the buff on the next dragon? I'd believe that, just wondering.

Whenever I've tried to include True Strike in the rotation, it was always the most intuitive to couple it with Bootshine usage.

[pandabearcat]

Oh, I see what you're getting at!

Yes, if you have fillered in a DoT then don't use True.

Using True with Bootshine is actually a really good idea, it gets you into your rotation easier and its easier to remember.

I would not use Fracture unless during BfB or IR, as it is still inefficient compared to your combo.

Unless you don't have TP problems.

[Allyrion]

Oh, I see what you're getting at!

Yes, if you have fillered in a DoT then don't use True.

Using True with Bootshine is actually a really good idea, it gets you into your rotation easier and its easier to remember.

I would not use Fracture unless during BfB or IR, as it is still inefficient compared to your combo.

Unless you don't have TP problems.

I only have TP problems on Turn 4 when we don't have two BLMs or Turn 5 (definitely don't use it there).
Turn 1 - my tp gets low using fracture but it dies before I get to a point where I have to stop using my abilities.

[NoctisUmbra]

Since all the monks I know complain about TP usage, it is a fairly significant (you gain almost a minute more of full boss uptime) TP efficiency increase to always swap out True/Twin.

Unless you know a disconnect is coming and you need Twin for a Demo/Snap coming up, I would say always alternate, even if it means the damage falling off. You lose roughly 1.4% dps (less than that actually, but i'm lazy) and gain quite a bit of efficiency.

You gain 3% absolute efficiency, but keep in mind TP drain is coming off of margins, so it is closer to something like 20-30% slower TP drain. It makes a huge difference, esp with the amount of relative SS monks have with GL3.

I don't have TP issues in any instance I can recall that isn't resolved by cutting ID and Fracture from my rotation. So yea, I don't see myself risking the loss of a +10% damage buff, even briefly, for further Tp efficiency. If I ever come across a situation where even after dropping ID, Fracture and ToD I still have TP issues, then I will work True Strike into my rotation.

My overall point is, a lot of MNKs underperform when it is in fact quite simple to put out solid DPS on MNK by just sticking to the flank bread and butter combo. So many MNKs take theorycrafted advice, try to pull off a rather complex rotation sequence that isn't necessarily set in stone, since a lot of it has to be played by ear during content, and end up losing GL, Twin Snakes, Dragon Kick, etc and resulting in worse DPS than if they had just used the 3 main skills back to back.

People need to take it one step at a time and realize that theorycrafting is a means to obtain better performance in content, not a universal recipe.

[EasymodeX]

I don't have TP issues in any instance I can recall that isn't resolved by cutting ID and Fracture from my rotation. So yea, I don't see myself risking the loss of a +10% damage buff, even briefly, for further Tp efficiency.

Humm, that isn't really the full picture. Keep in mind that the "further TP efficiency" allows you to inject additional Fractures.

That aside, I just ran the exact option in my model and the damage result was actually the same even if the TW buff drops for the following TW (when you swap in TR).

Simply put, dropping the TW buff off of TW is a loss of 14p on the attack. Replacing the middle TW with TR is a gain of 10p. The remaining 4p is gained if you get the +5% crit (successfeal rear position). It actually averages out pretty well. Net DPS of 0 with a gain in TP efficiency.

Edit: More detail:

(1) Assume a base 20% chance to crit (13-ish + average IR)
(2) TW-buffed TW is 140 * (1 + .5*20%) * 1.1 = 169.4
(3) TW-buffed TR is 150 * (1 + .5*25%) * 1.1 = 185.6
(4) Sad TW is 140 * (1 + 1.1) = 154
(5) Consistent TW-only is 169.4 * 2 = 338.8
(6) TR + sad TW = 185.6+154 = 339.6

GG.

Beyond that, you save what, 10 TP every 12s?

A normal MNK rotation (with TR) costs 56 P/GCD. FR costs 80P/GCD. That's a delta of 24. So every 12s*24/10 = 29 seconds you can use an extra Fracture using the TP you saved.

FR is a potency gain of 355/291-1 = 22% for a 1 GCD every 29 seconds. 29s is 13.9 GCDs, so 22%/13.9 = 1.6% rotational damage gained by switching to True Strike -- not directly from TR, but by saving more TP to be used on FR, even when you cause a drop of the TW buff on the next TW.

Sidenote: 355 and 291 are potency adjusted for crit, blunt resist debuff, FoF, GL3, etc.


TLDR: Using True Strike is DPS neutral and a TP gain even if you lose the Twin buff on the following Twin, assuming you get the successful rear positional. Even if you don't get the rear positional, it is a DPS gain if you use the extra TP to provide more Fractures.

[Allyrion]

Not to switch topics too much, but I've been playing with logrep.
Both logrep and ffxivapp still give me a dps loss with dots. Anyone else experiencing this? I thought the dot simulation was more accurate now.
Note: the 'loss' is not nearly as big as it used to be.

[EasymodeX]

How are you observing the "loss" with DOTs? Are you just comparing an X-minute parse without using DOTs to one while using DOTs?

Due to RNG with crits in general, that's not an effective way of assessing how well the parser is picking up DOTs, unless the difference you're observing is major.

I would suggest taking both detailed parses with %crit for each ability, de-factoring the crit to isolate their base damages, and re-applying a standard crit% rate to re-create the parse equally between both. Compare those to observe what your results are between "DoTs" and "no DoTs".

Or are you eyeballing the DOT tick for individual DOTs and checking what the parser is recording them as?

That result should be pretty accurate for regular damage (but prone to RNG crits).

Edit:

Also, try ACT!

[Allyrion]

How are you observing the "loss" with DOTs? Are you just comparing an X-minute parse without using DOTs to one while using DOTs?

Yep that's all I did. Just very basic testing, which is why I was wondering if any of you knew that it was accurate regardless.
Does it properly scale the dots with temporary buffs? I can always coincide them with either IR or BfB and sometimes both.
I'll actually try testing it today with the temp buffs.

One of the eyebrow raisers was getting an 8% crit on a parse (though that never happens in a real fight), while all the abilities were closer to 20% except the dots. That happened in ffxivapp, which made me try out logrep.
Also the overall dot damage just seemed so low overall. It was higher in logrep though.

I'll try what you said and maybe ACT.

[EasymodeX]

Does it properly scale the dots with temporary buffs? I can always coincide them with either IR or BfB and sometimes both.

The specifics of DOT modeling is parser-specific. Each parser is going to end up doing it in different ways. All I know is that the ACT plugin author uses the parse data for other attacks to model the DOT damage, and uses a crit% based on the other attacks. Finally, he tracks the data with and without buffs like BFB/Raging Strikes to model the DOT damage more accurately (e.g. apply RS to the DOT). There was a thread on Reddit where he was responding to questions and providing more detail. I don't know how IR was handled. In general it should at least be reflected in the estimated crit% applied to the DOT model based on other attacks. The result should be fairly close to accurate.

The only fundamental errors IIRC with ACT's DOT modeling were: (1) it uses the debuff timer for DOT ticks, not server timing (small difference), and (2) no Contagion (for SMN), although he was working on it.

Edit:

One of the eyebrow raisers was getting an 8% crit on a parse (though that never happens in a real fight)

There was a troll with the DRG threads who was posting his ultimate rotation backed by a parse where he RNG'd a 40-100% crit rate on all the major attacks (44% on autoattacks!!). Lols.