DoT changes and future Spell Rotations?

[Deleted]

But Haunt and Shadow Embrace are the only things keeping us balanced in PVP, so I'm not sure how you mean.

Which bit? The corruption style refreshes, or the other part?

All i meant was that it looks as though they may intend us to dot up a target at the beginning of an encounter, and more or less forget about the dots and debuffs, refreshing them all with a rotation of nukes/drains rather than directly recasting. Haunt and SE would still be there - in fact, they'd be easier to maintain.

I know they talked about fel fire being intended for demo or destro, but if UA has no initial damage (and the shard changes indicate it hasn't) which do you choose: recast UA or refresh it with an instant cast direct damage spell.

[gherkin]

I must have missed the Felfire Bolt refresh mechanic. I have some thinking to do.

[vkei]

they just have to make sure letting DoTs run off will not taxing by the new system

like said u have 6.8 ticks corr for 18 sec , a 0.8 tick will comes when corr ends with no refresh ,so that u don't need to clip the last half tick for the last 0.8 to actually goes on , cliping the last half tick still give u the most up time and less GCD usage ,but having to clip the last half tick is just the same as don't clip DoTs in an other way ,even harder ......

the other thing is CoD ,yes it can crit....but......we will nerver having the haste to make it ticks 2 times , and can't refresh it

[gherkin]

A .8 tick is even more confusing to you average user than dot clipping. They will never implement that, and even if they did it's still a damage done loss if not a dps loss.

I should animate the three solutions somehow...

[sycoman2]

Theres some pretty weird theories flying around about how Dot refreshing will work in cata...

When reapplying a DoT, you can no longer "clip" the final tick. Instead, this will just add duration to the spell, similar to how Everlasting Affliction currently works.

Right now if you refresh a dot before it expires you reset its tick timer. The corruption refreshes from everlasting affliction do not do this however, the duration is refreshed but the tick timer is not.

What i got from the above quote is that all dots will work this way by default.

Maybe im missing something, but i think we can all agree that that was very poorly worded and we can not make any solid conclusions until we get more details.

[vkei]

clipping DOTs at cata will work the same as how SB/haunt/DL/DS refresh corr but it always refresh to 18sec

when we look at UA , u want to clip it after the last tick and b4 the duration ends if they are not giving any damage done unless your haste give a full tick when duration ends without refresh

if u have 2sec between the last tick and duration end and not able to clip at this 2sec , u start your UA again that the 1st tick comes 2.x sec after u reapply your UA , u get totally (2+2.x+time between your old UA fade and new UA apply) sec gap between the tick , this kind of DOT refresh will be even harded than not clip the last tick , u will always fail to refresh at the best time unless u stop casting and wait

all i can think of is giving those last half tick doing thier damage when it ends , so if u can always refresh them at the best time u get 100% up time and doing the best damage , if u clip it too early u waste more casting time , if u let is fades and reapply , u lose up time , but in all case u will not feel like your haste is waste or u actually fucked up by the haste u have

also they have to work on CoA and CoD , u refresh CoA to full 24sec and lose the hardest tick? your CoD with 60 sec and tick at 50 sec when u can never clip it with the 60sec CD?

oh , don't forget DL/DS are also DOTs that u have no way to clip it , u lose damage if u let it run its full duration seems a stupid idea to me

[Goatonlb]

I say we get away from the convention of saying "clip" if it's going away, as we still seem to be grasping at straws that are no longer there. If the new system is going to work like Everlasting Affliction, then there's no such thing as "clip", just "refresh". Put that word out of your mind, and a lot of this gets simpler in design.

The problem as I see it, is that it's desireable to refresh DoTs at the last possible second (instead of the soonest possible second after they expire) to minimize the number of GCD's you have to spend casting them. In turn, this may simply make certain values of haste more desireable to have instead of a nice linear scaling. If every 100 haste, lets say, gave you an additional tick, then anything between would be valued a lot less for DoT classes.

However, if ALL our DoTs get spells to refresh them without having to use a GCD to do it, then none of this discussion about 1/2, 1/3, 1/4 ticks matters. Fel Flame refreshing UA will mean we cast that as part of our rotation to keep it going. The DPET won't be high enough to cast often though, since it's supposed to be used for mobility, we'll likely use it for that and JUST to refresh UA. Corruption will still likely be refreshed via most of the spells it already is, so Sub 25% BoA, UA, Corruption will all be refreshed naturally. Above 25% though, this still leaves the problem of WHEN to refresh.

Some valid points have been brought up though - such as Agony: Would it be worth to let fall off as we do now to insure the hardest hitting tick lands, or are they going to change the spell to have normalized damage throughout the duration instead? Sub 25% will this spell be worse since it'll be constantly refreshed via Pandemic?

While they mention all DoTs getting the benefit of haste, I'd just as soon throw CoD (BoD now) off that list, I see no way for haste to be helpful to the spell unless it also reduces the CD of the cast, as mentioned by vkei.

So we need to find out about Agony and how it'll interact with refreshing, and we need to find out how UA will work with various amounts of haste. I just don't feel we have enough information at this point to tell exactly - I would wager that they won't make us worry about partial ticks though, so we'll see what they implement to cover that.

[cronix]

These are in fact the same thing. The reason they are worded differently is to emphasise the main change, which is that when haste increases the tick rate of dots in cata, they will not lose any of their maximum duration.

This actually means that scaling will favour dot heavy classes, since haste will increase the output of our dots at the same rate it does for dd spells, however it will not increase the rate at which we have to expend gcds to do so.

I am actually a little skeptical about these changes being as beneficial as they portray. On fights like lich king for example during phase 1. I throw up corruption and ua on the other mobs while my eradication and 4piece are down and this helps be keep about an 80% uptime for this buff (it refreshes over the top of itself often), i do similar things on tovoron and anub ect.. My point is that with the dots no longer ticking as fast, it means that the chance of the procs like eradication, nightfall and 4 piece t10 type buffs will be less frequent and thus will either zero out the dps difference with having to refresh dots less or may even outweigh it all together, it all depends on the types of buffs and design they set in place. All im saying is it could go either way. What do you guys think?

[Deleted]

UA will be ticking faster than it is now, corruption ticks will be at the same rate as now. I don't quite get the problem.

All your on-tick procs will proc more.

[Zipher]

Guys,

The wording on the changes is pretty straight forward. You can recast any dot at any time and it will reset its duration (if already present). Haste and Crit will benefit all dots. Haste will increase the frequency between DoT ticks, but not alter the DoT duration. Only refresh/reapplys will alter a dots duration to its maximum time.

As far as rotation, there is still a question which, depending on the answear, can make our future rotation mm hunter easy or just a tad easier but, overall, bug free from its current state.

Do Haste, Crit and % Damage buffs (4pt10 for example) get recalculated per tick or on refresh/reapply?

If they end up recalculating stats at every DoT tick, then the only thing to worry about rotation-wise is 1) maximizing filler spells by extending refreshing dots for as long as possible, 2) on fights where dots will definitely fall ( adds swapping) perhaps refreshing dots to its full duration before swapping.

If they end up recalculating stats at refresh/reapply, then it gets a bit more interesting, and provides the means for adding a bit more dps. Essentially, you monitor your haste, crit and % damage and in some cases it may be of benefit to reapply a dot very early to have it benefit from your current stats.

In either case, the rotation per say will be a lot easier. Having to not deal with variable duration corruptions and "stronger spell already active" bugs will make the rotation much more manageable.

I feel the complexity ( and fun for that matter ) will come from the shard management. As planned, we are essentially getting 3 CDs with multiple options. Correct planing and execution of those may be what separates the pros from the noobs per say.

[vkei]

i actually change to demo because i don't like how the DoTs work nowadays , i don't like stack any buff i can have to make the OP corr and don't let it fall off in the fight ,do hope they fix it in cata

since the new DoT system is appling to all classes , i think they will make it run perfectly

[Goatonlb]

The wording on the changes is pretty straight forward. You can recast any dot at any time and it will reset its duration (if already present). Haste and Crit will benefit all dots. Haste will increase the frequency between DoT ticks, but not alter the DoT duration. Only refresh/reapplys will alter a dots duration to its maximum time.

The problem some of us are having is with haste and the number of ticks it gives. As we've mentioned before, you'll still want to refresh DoTs as late as possible to minimize the amount of GCD's you spend doing it (outside of auto-refresh spells like Corruption, or BoA and UA sub 25% via Pandemic). That leaves 2 DoTs that we need to manually reapply for 75% of the bosses HP - with Fel Flame or a direct cast of the DoT.

If we're waiting till the last possible moment to reapply them to max our DPS, what happens when haste benefits you slightly? What happens when the DoT duration has 2 seconds left, but there might be 2.4 ticks left? There's a big ??? in there as to how haste will effect the # of ticks. Again, this would make haste have very specific "spikey" values for DoT classes, as the value of haste would jump dramatically at levels of haste that grant exact #'s of additional ticks, and anything inbetween would seem meh.

[Zipher]

The problem some of us are having is with haste and the number of ticks it gives. As we've mentioned before, you'll still want to refresh DoTs as late as possible to minimize the amount of GCD's you spend doing it (outside of auto-refresh spells like Corruption, or BoA and UA sub 25% via Pandemic). That leaves 2 DoTs that we need to manually reapply for 75% of the bosses HP - with Fel Flame or a direct cast of the DoT.

If we're waiting till the last possible moment to reapply them to max our DPS, what happens when haste benefits you slightly? What happens when the DoT duration has 2 seconds left, but there might be 2.4 ticks left? There's a big ??? in there as to how haste will effect the # of ticks. Again, this would make haste have very specific "spikey" values for DoT classes, as the value of haste would jump dramatically at levels of haste that grant exact #'s of additional ticks, and anything inbetween would seem meh.

I agree with you on that sense if you read the rest of my post. The specific technical question is how will blizzard decide to update/recalculate buffs and stats on DoTs. Will it be per tick or per refresh.

Think about it. If it is per tick, after each dot tick does its damage, the game checks for your current stats and decides when will you next tick hit (haste), its chances of criting (crit) and compute its damage based on your SP + damage modifiers. This scenario will be the easier for us to handle as it allows for pretty much worry free dot refreshes.

The situation you describe assumes they will update buffs on dots on refresh/reapply. In that case, yes there are some interesting decisions to make as to when is the best time to refresh or reapply a dot. I am sure theorycrafters will have a field on this and will take some level of math and complexity to come up with sound rules of thumb. Overall though I personally feel that even with the latter situation we will be in better shape as compared to what we have today.

As i write these, I realize I am assuming that you may no longer be able to roll crit and % damage modifiers on refreshes... probably time to ask the blizzard forums for answers ( Though I thought they were planning on removing/fixing this mechanic for Cata).

[gherkin]

I posted this on the wow forums (warlock cataclysm thread). Let me know if theres anything I should change due to vagueness.

**********************************************

There are several methods for which you could implement dot refresh mechanics. While I use the example of Warlocks in this post, this applies to all classes with any DoT mechanic and no automated refresh ability (i.e. Haunt, Mind Flay). Also, I use the example of UA - I intentionally leave out the Felfire Bolt mechanic to explain my points because it is a simple spell with a standard duration and currently not refreshed automatically. I could use Serpent Sting or Rend in this example, but I don't.

First, the current Corruption refresh mechanic is Refresh Duration = (Base - Haste) - Time to Next Tick. This doesn't have any issues per se but the last tick will always occur when the spell wears off, which is counter-intuitive and causes dot-clipping if you were to overwrite (not refresh). This also forces you to cast more DoTs per minute rather than filler. Overall a good first run.

Second, you could have a Refresh Duration = Base - Time to Next Tick. This allows you to clip without losing damage and has the ticks affected by haste. However, the problem here is that maximum potential damage comes from pushing the refresh to as late as possible. If I had 50% haste on a UA, it would tick 7 times and idle for the last remaining 1s until falling off, dealing no extra damage. Maximum damage would come from refreshing UA at 0.1s remaining so that you could get an extra Shadow Bolt or Drain Soul tick in. If I missed this "late refresh" and the spell dropped off, I would have to start again and lose the 1.0s, effectively clipping my dot, but in reverse. This simply shifts dot clipping from a painful tick-loss to a slightly less painful tick-delay.

This could also lead to haste stacking to set values. For example, to get exactly 7 ticks from UA, I would force myself to 41% haste and not go any higher, and continue to play the game as I normally would.

The reason we would want to push as late as possible is to ensure that we get as many spells in as we can and cast as few DoT's as possible to maintain 100% uptime.
- Joe the Warlock who always refreshes at 3s remaining in a 5 minute fight will cast 25 UA's
- Bob the Warlock who always refreshes at 1s remaining in a 5 minute fight will cast 21-22 UA's
- Both warlocks have 100% uptime

Therefore, Bob did more damage because he had 3-4 more GCD's to spend on Shadow Bolts and Drain Soul.

The best solution I can think of is to allow dot refreshes is Refresh Duration = Base + X, where X is an arbitrary number large enough to encourage a refresh. I suggest 3 seconds. This allows us to refresh the spell anywhere in that last window (in my suggestion, last 3s remaining on the spell) and not lose GCD's by refreshing earlier (2.8s remaining) vs later (0.2s remaining). This way no matter when you refresh, early or late, you sustain the same amount of time casting UA and save the GCD's. Alternatively, X could be the lesser of Time Remaining or 3 seconds, meaning you cap the duration.

Regardless of outcome, the goal should be 100% uptime requiring the least number of player-initiated refreshes without the potential for clipping. At the very least, please reconsider the impact of refreshing as late as possible and the application of latency to that model.