Early Vs Delay Openers

[Kabooa]

You should open a discord named the Bole.

You laugh, but it was the only card that kept its 20% base!

[Satarn]

But I'm just a simple MNK using simple math to try and get people to cater to buffing my preferred, sub-optimal TK playstyle.

Fixed that for ya.

[Oxdarock]

Hello again, everyone! I've just recently updated the Early vs Delayed Openers Doc I started up a month and a half ago.

I'm actually pretty glad I did this. Not only did it allow me to find mistakes and make adjustments, but what was once a theory that relied on faith has now become conviction backed by math. There definitely might be some timing differences when put into actual practice with some GCDs and oGCDs missing buff windows by the slimmest of margins, but hopefully this is enough to convince people of my claim about the potential of delayed openers. Regardless, I believe it whole heartedly, but people can play the game how they want.

Anyway, shout out to Izsha here on the forums for their advice (and persuasive word choice) on providing evidence to back up my claims (which I really should have done at the start, lol) and to everyone who provided feedback through the several threads regarding these openers. Until SE decides to make a drastic change to a class (hint, hint, fix Anatman), I think I'll take a break on the theorycrafting for a while. Hope this helps everybody!

[Yandere-chan]

Looking at the document, I take it that the goal is to get a net gain by readjusting every class's opener so every possible buff lines up together with every class's strongest bursts? I'm just curious how this would work if not all the buffs are present for the particular party or how it affects damage down the line since these adjustments would affect everyone's rotations post the opener. Would they realign again naturally later in the fight? Will it still be a net gain with the adjustments needed between then? Would this slowly lose damage the longer the fight goes on? How short would a fight need to be for this to be a net gain if that's the case?

Obviously, this would require a lot of coordination too since everyone would actually lose damage if all the buffs didn't line up as planned, yeah? So in pugs, it's safe to assume that the more generally optimal early opener is better either way? Just curious of these things since they do of course matter. I think it's an interesting concept either way, adjusting a team's openings to better line up with buffs; I'm just curious how it all loops back to retain this synchronization of buffs and bursts and how it pans out depending on the team makeup.

[Oxdarock]

Looking at the document, I take it that the goal is to get a net gain by readjusting every class's opener so every possible buff lines up together with every class's strongest bursts? I'm just curious how this would work if not all the buffs are present for the particular party or how it affects damage down the line since these adjustments would affect everyone's rotations post the opener. Would they realign again naturally later in the fight? Will it still be a net gain with the adjustments needed between then? Would this slowly lose damage the longer the fight goes on? How short would a fight need to be for this to be a net gain if that's the case?

Obviously, this would require a lot of coordination too since everyone would actually lose damage if all the buffs didn't line up as planned, yeah? So in pugs, it's safe to assume that the more generally optimal early opener is better either way? Just curious of these things since they do of course matter. I think it's an interesting concept either way, adjusting a team's openings to better line up with buffs; I'm just curious how it all loops back to retain this synchronization of buffs and bursts and how it pans out depending on the team makeup.

Yes, that’s exactly the goal of this document, to get a net gain by lining buffs with every classes' strongest burst. It uses an extreme example of having the most percentage based raid buffs in the party, but if buffs are missing, then naturally the gain is less, though this is true for early openers as well. The most noticeable change to the rotation with delayed openers is that certain oGCDs come off cooldown 1-3 GCDs later than their earlier counterparts. Other than that, there’s no particularly drastic change in the rotation, which means everything should keep this alignment so long as nothing needs to be held, with everything lining up again at the 6 minute mark.

As for net gain in relation to fight duration, it should continue to increase so long as nothing is held and the rotation is continued like normal. However, it becomes a loss if that 1-3 GCD delay causes one to miss a portion of their burst window, or a particularly hard hitting GCD/oGCD that they’d normally be able to fit in using an earlier opener mid-fight due to mechanics, phase shifting, and sometimes add phases or dps checks (not the end of the fight since the boss dying sooner is typically a good thing). No amount of raid buffs can make up for an extra cast. The most well-known example of this is the GNB opener for E3S, in which they go into their burst window even sooner than in the early version, all so that they don’t miss a portion of a burst later on due to Maelstrom.

As for what openers to use in a pug scenario, I’d agree that using the early openers would be best in terms of keeping everything optimal with no coordination. The early openers have been around long enough to be considered standard, so there’s a good chance that other people will be using them. The only openers where the version you use is mostly irrelevant are PLD, BLM, and SAM, all of whom don’t particularly care about team comp and do their own thing. I will say that while the delayed SAM opener does more damage due to being ahead two GCDs, I’m not 100% sure if it’s an overall gain due to things like filler combos and/or ad hoc style play. However, the BLM delayed opener is a gain in most scenarios as the combination of F1 and F3 does more damage than a sixth F4 while still fitting in the same time frame, with Swiftcast timing being the main difference. MNK’s also in a similar situation since Anatman shenanigans make that opener inconsistent while the TK opener can still fit pretty well into most other classes’ rotations, even with the delayed BH timing.

[OneTrueMiqote]

However, the BLM delayed opener is a gain in most scenarios as the combination of F1 and F3 does more damage than a sixth F4 while still fitting in the same time frame, with Swiftcast timing being the main difference. MNK’s also in a similar situation since Anatman shenanigans make that opener inconsistent while the TK opener can still fit pretty well into most other classes’ rotations, even with the delayed BH timing.

If you had said something like "here is an idea, it may be true, it might not be true, but feel free to explore it!", it would have been fine. But this is exactly why people accuse you of spreading misinformation. Not because they are against your "new" ideas, but because you give out information that are based on your FLAWED logic and calculations and state them to be TRUE. I have seen you admit individual mistakes and correct them, but you have been unwilling to admit that you do not have sufficient understanding of this game's jobs nor the mathematical skills required (yes, your spreadsheet with elementary math is not enough) to analyze the complicity of openers. Being dismissive of this criticism will not help you improve at what you set out to do and will only cause more confusion to uninformed players, which is why your posts here have been bashed to oblivion.

[PyurBlue]

the BLM delayed opener is a gain in most scenarios as the combination of F1 and F3 does more damage than a sixth F4 while still fitting in the same time frame

F1 + F3 proc is less DPS than F4. The casts happen back to back, but the former still costs you 2 GCD's of time.

180+240/5 = 84 PPS
300/2.8 = 107 PPS

[xxPheNoMeNa]

I'm not gonna discuss every job upon reading this, although I can tell the information is not correct. But strictly referring to the SMN portions, your logic is extremely flawed and misleading.

You even make a generic explanation that says "As you can see, the delayed opener pulls ahead of the early opener by 444.5 potency, mainly due to the extra Ifrit attack, the extra R4 and both Akh Morns within Devotion, and the extra DoT ticks of the first Tri. And like the BRD opener, this gap will remain unless the boss becomes untargetable with 12 seconds left on both DoTs in the delayed rotation, although it’ll still fall short due to the way Devotion is set up for the remainder of the fight (i.e. allowing the use of two Akh Morns/Revelations and an extra R4 in every Devotion window)."

In reality, such a potency gain doesn't exist, and I'll explain why:

1. You don't "gain" an extra ifrit auto attack just by delaying the Bahamut. You could argue that you'll "gain" a very small amount of potency by pushing an additional auto attack into raid buffs, but you don't gain an auto attack outright. If Bahamut is used earlier, that auto attack will just occur after Bahamut finishes.

2. I doubt you can even get both akh morns on Devotion using your proposed opener, considering you're listing the 2nd Akh Morn 7 GCDs after the Devotion weave (and assuming the pet AI isn't botched, Devotion would cast right around the time you're using it). 7 GCDs @ 2.5s = 17.5s of time, not including the additional time that would be added on due to GCD delays with hardcast GCDs (what we refer to as the "Caster tax"), or the time required to actually weave oGCDs. Devotion is only a 15 second duration.

3. You aren't "gaining" DoT ticks either. What you are doing is simply avoiding clipping your Tri-Disaster into opener raid buffs and instead electing to hold it (though, you don't list a 2nd Tri-D in your opener despite it being 16 GCDs long, so I don't know where you're proposing to use your 2nd Tri-Disaster). But either way, the choice of clipping your Tri-Disaster early vs. not does not magically gain you DoT ticks, as DoTs will always be up on your target to gain server ticks regardless.

4. You probably didn't even realize that this would be an issue, but the proposed Bahamut rotation you're advocating would even cost you a Wyrmwave cast. If you're weaving your 2nd Akh Morn in the 6th GCD of your bahamut rotation, your final 2 GCDs MUST be instant casts to ensure those final Wyrmwaves don't ghost (by ghost, you'll see the animation but 0 damage actually registers, and you can confirm this using your in game battle log). Yet you're advocating to use a hardcast Ruin 3 following the 2nd Akh Morn weave, meaning your final 2 GCDs would be R3 > R4. This means your final wyrmwave will ghost pretty much every time you try to do this.

All of this leads me to believe that you didn't even bother to play test your ideas, let alone do proper research on how to implement them

And based on our own rotation sim sheet, your opener actually performs worse than our actual openers by roughly 20,000 damage overall, and this is just personal damage comparisons without other party buffs factored into it (with ideal rotation timings, your opener would be even worse off).

[Oxdarock]

F1 + F3 proc is less DPS than F4. The casts happen back to back, but the former still costs you 2 GCD's of time.

180+240/5 = 84 PPS
300/2.8 = 107 PPS

The reason why I say that the F1 and F3 combo happens in the same timeframe as F4 in the opener isn't because of the CD of the Fire spells, but rather the timing of Xeno, which is the same in both openers.

[Oxdarock]

snip

Thank you for pointing out the mistake with the second Tri cast. Had a feeling that I’d make a typo of some such somewhere. What’s worse is that I put an R3 where the second Tri should have been after the first Enkindle Baha. ^^;

Admittedly, I wasn’t 100% sure on the extra Ifrit Burning Strike due to the AI interaction with Devotion. Upon testing it once more, I’ve found that you do not gain an extra Ifrit attack, so that was a mistake on my part. I’ve fixed the portion that states this in the doc and as for the tables, because of this variable in Ifrit timings, I’d have to remove Burning Strike entirely, which would bring them closer by 80 potency, give or take.

That said, everything I said about Baha and Akh Morn is true. After fixing the delayed opener, putting the second Tri after the first Enkindle Baha (where it should have gone to begin with), the last two casts of the Baha phase are R4, and a swift R3. I’ve actually tested this opener at with a 2.43sec GCD and was able to fit them just fine. With even an even slower GCD, you’d probably just have to switch the order of the second Enkindle Baha and Fester in order for the second Akh Morn to fit within Devotion (though if you have rock bottom sps, whether or not you’d still fit it in is something even I’m not 100% sure of). I should also mention that, if you hold Aetherpact until the middle of the GCD cooldown, it won't be applied until at least 1.5secs later, right before Deathflare. You're only really able to get 6 casts off in Devotion regardless of which opener you use.

You’re correct in that you don’t “magically” gain DoT ticks in the opener and I actually had a difficult time figuring out how to word this. What I came to realize is that, in SMN’s case, the difference in DoT damage depends on how many DoTs tick via a Bio and Miasma refresh before a boss jumps, which is a very specific situation (which I’ve also added into the doc just now). So, assuming this isn’t the case, if you remove Tri damage from both tables (and the extra Burn in the delayed opener), you then get a difference of 47.9 potency, again, in favor of the delayed opener.