Connecting the DoTs: A Summoner Guide

[Sleigh]

No one is getting enhanced current modes. No Astral Heat IV, no GL4, no Aetherflow 4 stacks, etc.

BLM is getting a new stance, similar to how MNK is getting Chakra and SMN is getting Bahamut stacks.

[Dervy]

Yup. 3,0 is the official "YOU GET A STANCE, YOU GET A STANCE, EVERYONE GETS A STANCE" patch. We'll most probably see Rank 4 variants in 4.0.. Heuheu.

[pandabearcat]

stuff

You say "very likely" etc, I say "find the average case and remove randomness from your model".

Essentially the prorate on Bio II means, there will ALWAYS be one tick, and a SLIGHT chance of 2 ticks.

Similarly, for miasma, there is a decent chance of 1 tick, but not guaranteed, which you already agree with.

When taking the average case, we can assign expected values to the damage, and the expected value is not "very likely" it is "average case", which means out of 100 bosses you'll maybe get 90% with 2 ticks and 10% with 1 tick, but the model already takes that into consideration.

I'm not sure what the deal is here.

Lets give another example, a simple one, with just 1 dot.

I cast my dot, then wait 4.5 seconds.

So basically you have

[0-2.5] - no dot
[2.5 - 7] - dot

Now how much damage did that dot do?

It has a chance of two ticks (if first tick is between 2.5 and 4 sec, or to be more precise if the first tick is between -0.5 and 1 sec, so that the second tick with dots is between 2.5 and 4 sec), and chance of 1 tick (if first tick is between 4 sec and 5.5 sec, or, to be more precise, if the first tick is between 1 and 2.5 second, so that the second tick with the dots comes up between 4 and 5.5 sec).

So instead of going, at 7 seconds, it has a 50% chance of 1 tick and 50% chance of 2, you say, okay, the time the dot is active is 7-2.5 = 4.5 seconds.

This is equivalent to an average case of 4.5/3 = 1.5 ticks, and we multiply damage. Will you ever get 1.5 ticks? NO. But you will get 1 tick 50% of the time and 2 ticks 50% of the time.

This is an appropriate method of finding expected value of dots when their application times do not sync up...otherwise you use terms like "very likely", and you have fuzzy math.

EDIT: I also want to clarify why I chose 6 seconds as the "stopping point". At 6 seconds you are guaranteed EXACTLY 2 ticks, no more, no less. So, we can guarantee 2 ticks for the tridisaster model. By also prorating to 6 seconds, we guarantee that our model of "ticks every 3 seconds" now syncs up, regardless of how we applied our dots, so we no longer have to prorate anything. We assume at that point onwards all dots are ticking for both models.

[TraeSnow]

No one is getting enhanced current modes. No Astral Heat IV, no GL4, no Aetherflow 4 stacks, etc.

BLM is getting a new stance, similar to how MNK is getting Chakra and SMN is getting Bahamut stacks.

Yea, my bad I read it wrong. xD
Though still...am I the only one that feels like we are being limited more with getting more skills that require a stack from Aetherflow but still being stuck at 3 max?

[Havenchild]

So basically you have

[0-2.5] - no dot
[2.5 - 7] - dot

Isn't this assuming 0 precasting? So literally, The tank pulls, THEN I decide to use Bio II?

When in actuality what I do, is let's say at 4 seconds before the mob is pulled, Im already popping RS + POT + Casting Bio II. So I mean unless by some terrible stroke of luck, I just missed the DoT tick by micro seconds, I would get a first tick in that 0 - 2.5 window a Bio II dot tick and then by roughly 4 - 5 seconds I have all 3 up.

I guess the difference between what I'm saying and what you're saying, largely seems when do you actually count 0 seconds.
When the encounter starts? or when the first DoT goes up.

I use the first DoT (which falls in that 0 - 2.5s window) and work with potency per sec while if I'm not mistaken, you are using when the encounter start time and looking at it from a potency per GCD.

[pandabearcat]

I am not really talking about precasting here (if you're really into that, why not precast sflare > swiftcast > tri-disaster rotation?). I assume you know your high level play and am not going to make judgements on that XD Merely that your motivations are incorrect and your math is not solid.

I'm simply talking about the matter of prorating dots and why it makes sense mathematically.

Sure we can define the fight start at some given point, and shorten the dots there. These are two separate arguments though, you're talking about chances of catching a tick and I'm talking about how you can remove all randomness. You questioned my math and my precasting - I'm saying the former is rock solid, the latter we can certainly look at it your way and see if it is correct.

I don't necessarily disagree with your conclusion, only your motivations - thinks like unlucky and very likely imo have no place in models.

Anyway if we assume precasting and lets just say you get off Bio II at 0.5 seconds after pull, then it looks like this:

[0.5] - Bio II
[3.0] - Miasma
[3.5] - Bio
[5.5-6] - Swiftcast Sflare

We'll prorate everything to 6.5 sec cause thats how long it takes for the second ruin to pop out. Yes I know sflare has 0 chance of applying, so we will take that out of consideration for now.

Total damage is (6/3*35)[Bio II] + (3.5/3*35+20)[Miasma] + (3/3*40)[Bio] = 170.83

So if we assume precasting for tri-disaster it will look something like this. Bear in mind I am guessing at delays here, but I think they are reasonable.

[0.5] - Sflare
[1.5] - Swiftcast + Tri disaster
[4] - Ruin II
[6.5] - Ruin II

So for this we have (5/3*110)[DoTs] + (2*80)[Ruin II] + (2*25)[Sflare] = 393.3 + Tri-disaster upfront damage

So the difference is only 222.5 potency.

Tri-disaster would have to hit for 80+ potency to make it worthwhile. Unsure also if it will apply miasma direct damage, but it makes sense for it to not - because it will have its own potency, and applying auras is not contingent on direct damage.

It is interesting to note that this only results in slightly higher damage than simply precasting tri-disaster (or swiftcasting it). But it -does- result in higher damage, so its worthwhile if you know tank positioning. If you also know the positioning AND its not on the boss's current position, you could precast sflare for both cases and its pretty good.

Also to answer Judge's question about why not only prorate the first tick...well yes, that is essentially exactly what its doing.

Any tick that is equivalent to floor(duration/3) will basically mean a guaranteed tick. For example, prorating Bio II to 5 seconds results in 1 + 2/3 ticks...this is equivalent to only prorating the fractional tick and adding in the guaranteed one.

I think it will make sense once I get my simulator working, and we can see the differences in openers.

[Judge_Xero]

So apparently it might be a miss translation that 3D costs Aether. Which I hope to be true since I can't picture any situation where it would be good to use Aether like that other than
3D -> Bane

[pandabearcat]

Yea I just read that too, would be great to have a cooldown (maybe 60 sec) instead of aether cost.

Also maybe we can get aetherdam 4 =DDDDDD

[Sunako]

I think tanks are gonna love the idea if you open with tri disaster and fester without quelling once they have one shield lob into it XD.

[Hakmatic]

[1.5] - Swiftcast + Tri disaster

What I dont understand is why you keep saying swift+tri. I mean everyone knows that any AF move is OGCD. Pretty sure its safe to say its going be a OGCD since it uses AF.