[Balance] 5.4 Changes and Thoughts

[Cyous]

At this point, I think we could greatly benefit from extended DoT duration. 18-20 sec MF/SF could in theory help us offset the nerf to proc chance by giving us more bang for each GCD we spend dotting. They just buffed almost every other caster DoT out there, it would just go with the flow.

Not the most exciting of solutions, but unless math proves me wrong, I believe it would be a happy middle -regardless of a much needed buff to Hurricane.

At 18sec (16 ticks), we would expect 2.4 procs per DOT.
With 2 Moonfires, we would expect 1.7 procs per DOT (3.4 procs overall). (0.38 pps)
With 3 Moonfires, we would expect 1.4 procs per DOT (4.2 procs overall). (0.46 pps)

[halfawake]

From what I've seen on the ptr both lunar shower and hurricane should extend natures grace. With less starsurge price on multi dot we may run intonthe same problem that our ale gives us. Then again I'm cynical anything will happen. Just generally very displeased with the nerf.

[Slapsgiving]

At 18sec (16 ticks), we would expect 2.4 procs per DOT.
With 2 Moonfires, we would expect 1.7 procs per DOT (3.4 procs overall). (0.38 pps)
With 3 Moonfires, we would expect 1.4 procs per DOT (4.2 procs overall). (0.46 pps)

Thanks for the math, Cyous, but does this mean there is no gain whatsoever from it? Shouldn't longer DoT duration at least free up a couple of GCDs for other stuff and help cycle faster?

[Cyous]

Thanks for the math, Cyous, but does this mean there is no gain whatsoever from it? Shouldn't longer DoT duration at least free up a couple of GCDs for other stuff and help cycle faster?

I think you're missing a key detail.
Applying DOTs also consume extra GCDs. To keep it simple, we need to adjust the "total duration" for the varying number of instances. (basically, add a GCD per DOT.)
TotalDuration = (NumMoonfire + Duration) - 1
(The " -1" is there to offset the application of the first DOT coupled with the duration (our Duration is 14 sec, not 15 [1 dot active]). It costs 1 GCD, yes, but technically the first DOT starts at 0.00sec, not 1.00).

I honestly have no idea how to math this next part...it's just speculation on how it might be done, but I would only assume that if we don't get at least 1.5 SS procs per DOT cast, it's not worth it.

Crit: 50%
Ticks: 12
TickDur: 14sec

At 4 instances of Moonfire... 1.2 procs per DOT (4.8 procs overall). (0.533 pps)

All that I can say is that the effective proc rate is simply leveling out, but you're not getting much in return. At best you're making up for GCDs spent on DOTing at 4 targets (not gaining anything at 5 targets). The idea of SS-weaving was to gain GCDs over hardcasting nukes.

With 4 DOTs... on live...we're expecting 7.20 procs. We only spend 4 GCDs DOT'ing. We're gaining 3 GCDs back over 17sec. Since we spend 4 GCDs DOT'ing, our NG should only have 12sec remaining (the 0.00 concept, so our first DOT comes at 15.00sec, and the second DOT would be at 14.00sec). Considering we use NG as our baseline, we should have 3sec of non-NG uptime with 4 DOTs rolling, and without refreshes.

(([NG_Dur] - [GCDs spent]) / [TotalDuration] - [ (Number of Expected SS procs) - (GCDs spent)] = NG uptime))
(([15.00sec] - [4.00]) / [14+4-1=17] - [7.2-4 = 3.2] = 11.00/17 - 3.2 = 11/13.8 = 80%))
80% uptime of 15sec = 3sec downtime (18sec total)

^ this is done using 5.3 values -- not the SS changes for 5.4
v this is done using 5.4 values -- yes, this is going to suck...

(([NG_Dur] - [GCDs spent]) / [TotalDuration] - [ (Number of Expected SS procs) - (GCDs spent)] = NG uptime))
(([15.00sec] - [4.00]) / [14+4-1=17] - [3.6 - 4 = 0.4] = 11.00/17 - 0.4 = 11/16.6 = 66%))
66% uptime of 15sec = 5sec of downtime. (20sec total)

Remember, this assumes no DOT refreshing. Factoring that in...well just add GCDs for every refresh (1sec added to downtime per refresh).

Again, I really don't know. We need to simulate this kind of thing, 1 instance of math won't be enough. There is...well..nothing to compare it to. I cannot test it either due to bugs as well.

[Slapsgiving]

I think you're missing a key detail.
Applying DOTs also consume extra GCDs. To keep it simple, we need to adjust the "total duration" for the varying number of instances. (basically, add a GCD per DOT.)
TotalDuration = (NumMoonfire + Duration) - 1
(The " -1" is there to offset the application of the first DOT coupled with the duration (our Duration is 14 sec, not 15 [1 dot active]). It costs 1 GCD, yes, but technically the first DOT starts at 0.00sec, not 1.00).

I honestly have no idea how to math this next part...it's just speculation on how it might be done, but I would only assume that if we don't get at least 1.5 SS procs per DOT cast, it's not worth it.

Crit: 50%
Ticks: 12
TickDur: 14sec

At 4 instances of Moonfire... 1.2 procs per DOT (4.8 procs overall). (0.533 pps)

All that I can say is that the effective proc rate is simply leveling out, but you're not getting much in return. At best you're making up for GCDs spent on DOTing at 4 targets (not gaining anything at 5 targets). The idea of SS-weaving was to gain GCDs over hardcasting nukes.

With 4 DOTs... on live...we're expecting 7.20 procs. We only spend 4 GCDs DOT'ing. We're gaining 3 GCDs back over 17sec. Since we spend 4 GCDs DOT'ing, our NG should only have 12sec remaining (the 0.00 concept, so our first DOT comes at 15.00sec, and the second DOT would be at 14.00sec). Considering we use NG as our baseline, we should have 3sec of non-NG uptime with 4 DOTs rolling, and without refreshes.

(([NG_Dur] - [GCDs spent]) / [TotalDuration] - [ (Number of Expected SS procs) - (GCDs spent)] = NG uptime))
(([15.00sec] - [4.00]) / [14+4-1=17] - [7.2-4 = 3.2] = 11.00/17 - 3.2 = 11/13.8 = 80%))
80% uptime of 15sec = 3sec downtime

^ this is done using 5.3 values -- not the SS changes for 5.4
v this is done using 5.4 values -- yes, this is going to suck...

(([NG_Dur] - [GCDs spent]) / [TotalDuration] - [ (Number of Expected SS procs) - (GCDs spent)] = NG uptime))
(([15.00sec] - [4.00]) / [14+4-1=17] - [3.6 - 4 = 0.4] = 11.00/17 - 0.4 = 11/16.6 = 66%))
66% uptime of 15sec = 5sec of downtime.

Remember, this assumes no DOT refreshing. Factoring that in...well just add GCDs for every refresh (1sec added to downtime per refresh).

Again, I really don't know. We need to simulate this kind of thing, 1 instance of math won't be enough.

Thanks again for the further explanations, that was quite a... erm... mathful? (I can follow, don't feel like it is totally wasted).

If you consider demo locks, for instance, they work pretty much with a very strong dot with a long duration and a shorter one that can be extended. They also must be refreshed every so often with the right procs (it is the UVLS spec), but overall I feel they are less cumbersome to manage and more efficiently kept up.

Oh well, one week to the patch so hoping for a bandaid of some sort before it launches. Thanks again for all the modelling and ideas, I truly hope some of it falls on not completely deaf ears.

[Cyous]

I'm looking forward to reviewing some of the logs (I won't be raiding progression unfortunately.) Dropping a bomb to our only niche was a mean thing to do...but the reasons are understandable. With that said, 1 week is still enough time to get an AOE buff (QoL buff, not a numbers buff), though I don't think it will happen at this rate.

I did a quick test on live to compare my numbers to the math posted above. 2-3sec downtime. So it looks correct (I removed my helm and thus the meta gem, as it skews the total ticks by second...thus increasing the frequency of procs).

I guess it's safe to say that the math is a decent indicator of what to expect.

- - - Updated - - -

Possible culprit for the low SS proc rates we discussed a couple pages ago...SS is nerfed by 50% (* 0.5).

When I set the number of DOTs to 2, the "effective Crit" to 40%, and the frequency of DOTs to 12 ticks over 14sec...

... 382.5 ticks / 35.23 procs / 60sec = 0.181 procs per second.
... Using the same numbers as above...but using 15% proc rate... we have 0.182 procs per second.

Could we get a test without the 4pc? And another with it? Trying to isolate the issue here...

[Fleuria]

Hopefully if it is bugged they quickly pick it up and fix it otherwise if it sneaks through it might be crippling for progression term but if its been nerfed by 50% does it also affect the single target?

--Update--

Although just played around on ptr with dotting 2 targets and doing a standard rotation, seemed to be fixed or well I had way to many procs, rarely even had to hard cast in general, woulda logged but then i realised that i hadn't set it to ptr combat log zz.

[Grubjuice]

I would love LOVE LOVE if Dream of Cenarius caused Wrath to refresh Harmony (for Resto) and caused Healing Touch to generate Lunar/Solar Energy (for Balance)

i might actually take DoC if that were true.

[Glurp]

I would love LOVE LOVE if Dream of Cenarius caused Wrath to refresh Harmony (for Resto) and caused Healing Touch to generate Lunar/Solar Energy (for Balance)

i might actually take DoC if that were true.

I'm not sure why you'd want that. The current version of DoC is useless for Balance but it's still better than your suggestion.

[Deleted]

and caused Healing Touch to generate Lunar/Solar Energy (for Balance)

i might actually take DoC if that were true.

Even if it did that it would still be terrible. Balance doesnt want to throw out heals into what is already a clunky rotation at times. DoC needs a complete redesign for pve atleast.

[Deleted]

What do you guys think about the idea that astral comunion should be able to be cast while moving making it easyer to be in eclipse when you have some downtime while moving.
Baseline instead of a pvp set bonus.

Thinking about fights like lei shen heroic or iron juggernaut

[Deleted]

As for AoE fixing, I'm thinking something simple.

Hurricane: "Creates a violent storm in the target area causing 339 [+ 31% of Spell power] Nature damage to enemies every 1 sec and increasing the duration of Nature's Grace by 1 second. Each time Hurricane ticks while not in an eclipse state, it generates 10 Lunar or Solar energy, whichever is more beneficial to you.. Lasts 10 sec. Druid must channel to maintain the spell.

This would allow us to keep NG up while AoEing, without hurting us when we transistion back into single target, and would not make Hurricane completely crap outside of eclipse. Of course the NG duration bonus is reduced by haste, like the ticks of the spell.

It's still an extremely weak spell because we have to be stationary while channeling, and the damage is pretty bad, but at least it's something.

Another cool, minor thing could be allowing us to pop mushrooms while channeling Hurricane.

[Grubjuice]

Another cool, minor thing could be allowing us to pop mushrooms while channeling Hurricane.

WM: Detonate is off GCD so you can pop Mushrooms a the beginning or end of any channel and not triger a GCD, so that's not the big issue per se,

rather it would be nice if we could place AND pop mushrooms while channeling hurricane (or if mushrooms spouted up automatically under the field of our hurricane's AoE 1 every 3 seconds over 9 seconds)

[Meejum]

As for AoE fixing, I'm thinking something simple.

Hurricane: "Creates a violent storm in the target area causing 339 [+ 31% of Spell power] Nature damage to enemies every 1 sec and increasing the duration of Nature's Grace by 1 second. Each time Hurricane ticks while not in an eclipse state, it generates 10 Lunar or Solar energy, whichever is more beneficial to you.. Lasts 10 sec. Druid must channel to maintain the spell.

This would allow us to keep NG up while AoEing, without hurting us when we transistion back into single target, and would not make Hurricane completely crap outside of eclipse. Of course the NG duration bonus is reduced by haste, like the ticks of the spell.

It's still an extremely weak spell because we have to be stationary while channeling, and the damage is pretty bad, but at least it's something.

Another cool, minor thing could be allowing us to pop mushrooms while channeling Hurricane.

Paraphrasing another's idea I have seen about this, which is the most elegant solution I have seen:

They could fix it by allowing SS procs to be able to be cast during the hurricane channel. They have the technology it seems...MW monks can instant cast some heals when channeling soothing mist without breaking the channel. Not a programmer, so not sure if offensive vs defensive casts would be problematic.

[earthwormjim]

As for AoE fixing, I'm thinking something simple.

Hurricane: "Creates a violent storm in the target area causing 339 [+ 31% of Spell power] Nature damage to enemies every 1 sec and increasing the duration of Nature's Grace by 1 second. Each time Hurricane ticks while not in an eclipse state, it generates 10 Lunar or Solar energy, whichever is more beneficial to you.. Lasts 10 sec. Druid must channel to maintain the spell.

This would allow us to keep NG up while AoEing, without hurting us when we transistion back into single target, and would not make Hurricane completely crap outside of eclipse. Of course the NG duration bonus is reduced by haste, like the ticks of the spell.

It's still an extremely weak spell because we have to be stationary while channeling, and the damage is pretty bad, but at least it's something.

Another cool, minor thing could be allowing us to pop mushrooms while channeling Hurricane.

It doesn't need to grant eclipse energy, you never want to be casting it outside of solar, let alone outside of any eclipse state.

Its damage is actually not terribly far off other classes, provided you have NG and are in Solar.

Whenever I pad on heroic Lei-shen, I'm pretty middle of the pack as far as AOE damage on balls goes.

[Deleted]

It doesn't need to grant eclipse energy, you never want to be casting it outside of solar, let alone outside of any eclipse state.

Its damage is actually not terribly far off other classes, provided you have NG and are in Solar.

Whenever I pad on heroic Lei-shen, I'm pretty middle of the pack as far as AOE damage on balls goes.

The problem is that we have to line it up with solar and only for duration of NG to even be middle of the pack. Its just too complicated and dependent on too many things to be a good aoe.

[Cyous]

Reviewing logs from earlier today (Garrosh testing)...

With 6 active DOTs (CA = 12), and 55% Crit (effectively), the 4pc bonus proc rate of these DOTs is 15.5%.

Using a very small window, I had 145 DOT ticks and 12 SS procs. (0-17sec window)

Astonishingly enough, that is an expected amount of SS procs...I'm really, REALLY shocked at this... SS proc'ing 12 times with an expected 12.3 procs. (My SS proc rate was exactly 12.3 as well). I honestly expected to have...well...more procs with CA active. It just seems...so...weak. Its niche is dead. It's just raw damage output....

When I shift the logs to 24sec ...(0-24sec window)
17.8 expected SS procs (15.5% * 55% * 208)
18 SS procs in logs.

Seriously? That's just so disappointing. :C

[Mirri]

Reviewing logs from earlier today (Garrosh testing)...

With 6 active DOTs (CA = 12), and 55% Crit (effectively), the 4pc bonus proc rate of these DOTs is 15.5%.

Using a very small window, I had 145 DOT ticks and 12 SS procs. (0-17sec window)

Astonishingly enough, that is an expected amount of SS procs...I'm really, REALLY shocked at this... SS proc'ing 12 times with an expected 12.3 procs. (My SS proc rate was exactly 12.3 as well). I honestly expected to have...well...more procs with CA active. It just seems...so...weak. Its niche is dead. It's just raw damage output....

When I shift the logs to 24sec ...(0-24sec window)
17.8 expected SS procs (15.5% * 55% * 208)
18 SS procs in logs.

Seriously? That's just so disappointing. :C

What are you grumbling about ? 12 SS procs in 17 seconds and 18 procs in 24 seconds. Isn't that like chain casting SS procs pretty much accountin for gcd and server lag. What am I missing that is making you feel so disappointed ?

[Cyous]

What are you grumbling about ? 12 SS procs in 17 seconds and 18 procs in 24 seconds. Isn't that like chain casting SS procs pretty much accountin for gcd and server lag. What am I missing that is making you feel so disappointed ?

That's with 8.67 ticks every 1sec (effectively). It's like having 12 DOTs rolling (6 GCDs -- due to CA) for the price of 6 DOTs (3 GCDs -- due to CA.) That's what's making me grumble.

[Mirri]

That's with 8.67 ticks every 1sec (effectively). It's like having 12 DOTs rolling (6 GCDs -- due to CA) for the price of 6 DOTs (3 GCDs -- due to CA.) That's what's making me grumble.

Isn't that the intended effect of the changes anyway? To prevent us from ONLY dotting and SS on multi target fights or at least reduce the effectiveness. At least we still have CA to pop during multi target encounters to quickly double dot and chain cast SS (as proven by your Maths above!).