Page 3 of 3 FirstFirst
1
2
3
  1. #41
    Re: haste plateaus

    In New's guide I posted about this and New & Aica replied to clarify a few things.

    Basically when you reach a plateau the extra tick itself does not increase DPS (as dot ticks are equally spaced apart so it doesnt matter if you have 2 ticks 2 seconds apart or 80 ticks each 2 seconds apart the Damage Per Second is constant. Damage per mana/casttime is a different matter) however when you gain an extra tick the spells length is reset to the default length (+half a gcd - thanks Aica) which as New pointed out can free up some GCD's over the length of a fight which of course increases DPS.

  2. #42
    Link to your armory and he's, I bet difference is huge or else you're mate is bad player.

  3. #43
    Stressing on the haste plateau on the dots is a sign of total misunderstanding of the new dot mechanism.

    First off, all dots last for fixed amount of time now.

    Say in a 15min (900 seconds) fight, with no temporary haste modifier proc, there is a dot that ticks once every 10 seconds, and the dot lasts 45 sec. So one cast of the dot gives 4 ticks plus 5 seconds. If the dot is refreshed at exactly the same time as it ends, you will get the first tick 5 seconds after it is refreshed instead of the 10th second. In another word, no time is wasted if a certain haste rating puts the end of a dot between two ticks. So, in a 900 sec fight, you will have 90 ticks. Now due to a haste increase, this dot ticks at 9 sec interval, then there will be 100 ticks in total in 900 sec.

    According to the "haste plateau" theory, until I can push my dot tick timer from 10 sec to 9 sec, the relationship between the haste and dps benefit is non-proportional and threfore I should stack other stats until there is enough item points to get the haste milestone. This is ICC mindset and it is no longer valid. See this table (assuming the dot is always refreshed when it ends):

    In cata (fixed dot duration, tick timer modified by haste)
    Haste 10% 8% 5% 3% 0%
    Tick timer 9 9.2 9.5 9.7 10
    # of ticks 100 97 94 92 90
    Tick # gain 3 3 2 2 --

    In ICC (fixed # of ticks per dot cast, dot duration modified by haste)
    Haste 10% 8% 5% 3% 0%
    DoT duration 40.5 41.4 42.75 43.65 45
    # DoT refresh 22.2 21.7 21.1 20.6 20
    # of ticks (4 per)88 86 84 82 80
    Tick # gain 2 2 2 2 --

    So, in conclusion

    A) Always refresh a dot before it ends. Refresh early in most cases should be better than let it drop.

    B) In cata, haste is constantly increasing our dot ticks. The higher the haste is, the more ticks we are gaining. Since shadow priests has 4 dots, haste is inarguably the best secondary stat at any time, even after the GCD cap. There would be still a so-called "haste plateau", but it exists only between two ticks, not between recasts of dots. And the plateau range will be much smaller. Due to the different duration and # of ticks for each dot, it makes no sense try to find the optimal point among these plateaus.

    C) It is easier to play a shadow priest in Cata since dot refresh will not become more and more frequent as haste inflates. Also noting the # of ticks difference at the same haste level, we are gaining a decent # of free ticks in cata as no dot time is wasted. Let alone the theorycrafting and memorization work for people who wants to go uber.

  4. #44
    Quote Originally Posted by Neliah View Post
    I honestly find this whole subject confusing, specially since they list a BiS trinket pre-heroic raids (On Shadowpriest.com), that has a 1900 something mastery proc.. -_-
    After gemming, enchanting, and grabbing the right pieces of gear the most important difference between good shadow priest dps and bad shadow priest dps is "rolling/weaving" your procs into your dots. Let me explain and it will become clear how this relates to trinkets and your mastery confusion.

    When we cast Devouring plague, Vampric touch or when sw:P is refreshed by mindflay, wow looks at your spell power, haste, and crit you currently have and applies them to these dots for their duration. So if for example your a tailor(shadow weave cloak), engineer(synapse springs on gloves), have power torrent on your wep, darkmoon card volcano, and Theralion's Mirror you have a lot of procs you can roll into your dots to make them more powerful.

    Lightweave embroidery - 580 int
    Synapse springs - 480 int
    Power torrent - 500 int
    DMC Volcano - 1600 int
    T's Mirror - 1926 Mastery

    Now I haven't done the math on when its worth it to refresh dots when x procs, however it does seem pretty clear that when you have 1600 int proc you should refresh dots also DMC volcano shares an internal cd with power torrent. This means they usually proc around the same time with some rng mixed in there.

    To answer your question as to why a 1926 proc on a trinket is so good when its not weighted that high in our stats. Our mastery directly affects our dots with empowered shadows buff. So when we are looking at trinket procs we need to weigh the benefit these stats give to our dots over the benefit to other spells like mf mb and sw:d because we will be using our dots during these procs. This changes stat weights and priorities for trinkets. Dots are even more heavily weighed when it comes to trinket procs due to the fact that we can refresh dots when the trinkets first proc and then refresh them again when the proc is about to wear off. This increases the time we gain the bonus from the trinket by an extra whole dot time. Big dps increase.

    So this is where theralion's mirror gets its awesome rating. If any of you shadow priests are looking to push the envelope of your dps and really shine this is probably an area you can improve. I would consider weaving your procs into dots to be the "Art of Shadow"

  5. #45
    Quote Originally Posted by Vaeoth View Post
    After gemming, enchanting, and grabbing the right pieces of gear the most important difference between good shadow priest dps and bad shadow priest dps is "rolling/weaving" your procs into your dots. Let me explain and it will become clear how this relates to trinkets and your mastery confusion.

    When we cast Devouring plague, Vampric touch or when sw:P is refreshed by mindflay, wow looks at your spell power, haste, and crit you currently have and applies them to these dots for their duration. So if for example your a tailor(shadow weave cloak), engineer(synapse springs on gloves), have power torrent on your wep, darkmoon card volcano, and Theralion's Mirror you have a lot of procs you can roll into your dots to make them more powerful.

    Lightweave embroidery - 580 int
    Synapse springs - 480 int
    Power torrent - 500 int
    DMC Volcano - 1600 int
    T's Mirror - 1926 Mastery

    Now I haven't done the math on when its worth it to refresh dots when x procs, however it does seem pretty clear that when you have 1600 int proc you should refresh dots also DMC volcano shares an internal cd with power torrent. This means they usually proc around the same time with some rng mixed in there.

    To answer your question as to why a 1926 proc on a trinket is so good when its not weighted that high in our stats. Our mastery directly affects our dots with empowered shadows buff. So when we are looking at trinket procs we need to weigh the benefit these stats give to our dots over the benefit to other spells like mf mb and sw:d because we will be using our dots during these procs. This changes stat weights and priorities for trinkets. Dots are even more heavily weighed when it comes to trinket procs due to the fact that we can refresh dots when the trinkets first proc and then refresh them again when the proc is about to wear off. This increases the time we gain the bonus from the trinket by an extra whole dot time. Big dps increase.

    So this is where theralion's mirror gets its awesome rating. If any of you shadow priests are looking to push the envelope of your dps and really shine this is probably an area you can improve. I would consider weaving your procs into dots to be the "Art of Shadow"
    I'm pretty sure you are wrong, I was under the assumption that every dot tick benefits from your stats at that time, there is no need to recast your dots.

  6. #46
    Quote Originally Posted by Töasty View Post
    I'm pretty sure you are wrong, I was under the assumption that every dot tick benefits from your stats at that time, there is no need to recast your dots.
    Plz provide some evidence to your claim. After being an avid shadowpriest for all 3 expansions and following the community I have never seen this theory conradicted and have seen it as a dps strategy in more than one place. I am ready to admit that I may be wrong but I wont believe it until I see some links or theorycrafting. I myself see the benefit in my dps charts.

    Also are you sure you are not referring to SW:P because this gets refreshed by MF all the time and at that point would refresh crit and haste values

    ---------- Post added 2011-02-16 at 01:05 PM ----------

    Sorry editing post wasn't working,

    I have found a shadowpriest(dot)com posting where they discuss this and how VT and DP do not refresh crit and haste values in real time. They also discuss how haste plateaus do not save mana but plz stay away from a plateau vs non plateau argument. If your stuck in your ways fine by me the gain is minimal and who has mana problems as a sp.

    Post is 4 down
    shadowpriest(dot)com/viewtopic.php?f=61&t=29778

  7. #47
    Quote Originally Posted by Vaeoth View Post
    Dots are even more heavily weighed when it comes to trinket procs due to the fact that we can refresh dots when the trinkets first proc and then refresh them again when the proc is about to wear off. This increases the time we gain the bonus from the trinket by an extra whole dot time. Big dps increase.

    So this is where theralion's mirror gets its awesome rating. If any of you shadow priests are looking to push the envelope of your dps and really shine this is probably an area you can improve. I would consider weaving your procs into dots to be the "Art of Shadow"
    Darkmoon Card: Volcano
    Binds when equipped
    Unique-EquippedTrinket
    Requires Level 85
    Item Level 359
    Equip: Increases your mastery rating by 321 (1.79 @ L85).
    Equip: When dealing damage with spells, you have a chance to deal 900 to 1500 additional Fire damage to the target and gain 1600 Intellect for 12 sec.

    So casting dots when trinket procs, and recasting them again before the proc wears off, all in a 12s window makes you a good spriest????

    I call BS on that.

  8. #48
    You are right to an extent, I usually dont refresh dots twice with volcano because I'm spending some time getting empowered shadows up, I do however put dots up towards the end of volcano. VT lasts 15 secs, DP 24 secs. Thats benefiting from volcano for at least twice as long with both those spells. That my friend does make you a good shadow priest. Right now I am running with Heart of ignacious not Mirror as my second trinket which adds 1600 haste for 20 secs on use so I do refresh my dots twice when popping that baby. 44 secs of hasted DP and 35 secs hasted VT are huge dps increases. If you dont want to take the time or effort to do these things feel free to be below me on dmg charts I dont mind. Just trying to give some advice to people ready to take it to the next lvl. No need to hate.

  9. #49
    Not hating. Just pointing out oddities in the advice you are providing.

  10. #50
    When it comes to DoTs and refreshing... A Dot (and shadowfiend) takes a 'snapshot' of your stats at the time it is cast or refreshed, it DOES NOT update in real time. In order to benefit from a proc, it would need to be recast, this includes empowered shadows.

    Keep in mind:
    1) SWP updates with all active buffs whenever mindflay refreshes it.
    2) Recasting DP is almost never a dps loss, (though highly mana ineffecient) its usually worth it to recast over casting mindflay. Though if you have a really nice proc effect going, make sure you don't over-write it with a smaller one, it can happen.
    3)VT is usually the main concern, most of the time, the extra gcd will cancel out any benefit of recasting... though if you get something really good, like the volcano buff and the mirror enchanced empowered shadows at the same time, recast VT early before either of those two fall off.

  11. #51
    Quote Originally Posted by adilina View Post
    Stressing on the haste plateau on the dots is a sign of total misunderstanding of the new dot mechanism.

    First off, all dots last for fixed amount of time now.
    no point in reading any further, your wrong right there.

  12. #52
    Quote Originally Posted by adilina View Post
    Don't forget each flay crit reduces the CD of our pet by 10 sec. So theoretically crit scales faster than mastery. Not saying crit is better than mastery but do keep in mind that shadowy appartition cannot hit boss in the air.
    I'm not sure that is evidence to suggest crit scales better than mastery. Firstly mastery is a direct % damage increase on a large portion of our damage and % damage increases always scale in a linear fashion.

    Crit on the other hand never scales in a linear fashion, the more crit you get the less % impact it has on your DPS. Going from 0% crit to 1% crit will increase your DPS by 1%, but going from 50% to 51% crit will do significantly less. You are quite correct in stating that you will get higher fiend uptimes but you have yet to prove that the damage increase from higher fiend uptimes is enough to make up the diminishing effect crit has on your dps %.

    All I can say is I would be rather skeptical that crit will scale better than mastery but I'm not stupid enough to declare it as fact based on what seems logical to me.

    Lastly, just to murky things even more, stats synergise togther.. in particular crit and haste will have a compound effect on fiend uptime. With higher haste there are more tick chances that can crit, so which stat is truely responsible for the higher uptime. Sometimes I can be just about impossible to disentangle and understand which stat is doing what, it's the combination of both that is giving the effect. Haste will also increase ES uptime which will have a positive effect on mastery as well...

    In short these sort of presumptions don't really help people make informed decisions. You need to research and come up with data to support a particular gear strategy.

    ---------- Post added 2011-02-17 at 11:03 PM ----------

    Quote Originally Posted by ralphie View Post
    Darkmoon Card: Volcano
    Binds when equipped
    Unique-EquippedTrinket
    Requires Level 85
    Item Level 359
    Equip: Increases your mastery rating by 321 (1.79 @ L85).
    Equip: When dealing damage with spells, you have a chance to deal 900 to 1500 additional Fire damage to the target and gain 1600 Intellect for 12 sec.

    So casting dots when trinket procs, and recasting them again before the proc wears off, all in a 12s window makes you a good spriest????

    I call BS on that.
    Keep in mind that recasting a DoT that is already present doesn't cause any DoT clips, all you are doing is locking in a short term powerful buff for a longer period of time. Also keep in mind that VT and DP do roughly 60K damage for their GCD and have generous SP coeffecients (particularly VT its double what it should be)... this means they will scale far greater than any nuke spell you could cast.

    I'm not 100% sure its a no brainer in all situations, its probably best to get an undestanding for the DPeT of your DoTs with standard stats and compare them to the DPeT with temp buffs present. That way you should be able to identifty if and when a DoT is worth recasting to prolong a buff.

  13. #53
    Quote Originally Posted by Chrispythegull View Post
    I think it is you who is not understanding the mechanics of how dots tick now. DOTS now tick continually and must be refreshed BEFORE they fall off. The reason why they must be refreshed before they fall off is because the mechanics of dots has changed from WOTLK in that they can now be 'refreshed'.

    Previously you would have to refresh dots just after they fell off. The reason why is because the entire effect of haste was baked into each individual DOT. The tick rates were recalculated and then neatly packed into each haste-shortened DOT. In Cata this is not the case any longer. Now what happens is that the DOT is shortened until it reaches a 'haste plateau' at which point the DOT gains an extra tick and reverts back to the default DOT time. So ostensibly what could happen is that your haste might reduce an 18 second dot down to 16.5 seconds but you would still only get the same 9 ticks at slightly lessened intervals. You would have to recast the dot just before it fell off so that you would get your 'bonus' tick at the beginning of the next DOT, if this makes sense.

    ***THIS ILLUSTRATION IS NOT DRAWN TO SCALE, IT'S FOR ILLUSTRATION PURPOSES ONLY***

    Normal 18 sec DOT -----> (CAST_____TICK_____TICK_____TICK_____TICK_____TICK_____TICK)
    New 16.5 sec DOT -------> (CAST___TICK___TICK___TICK___TICK___TICK___TICK__)

    Notice the appearance of empty space at the end of the second haste-influenced DOT. It is for THIS reason that you must recast the DOT before it expires. Doing so allows this to happen

    Normal 18 sec DOT -----> (CAST_____TICK_____TICK_____TICK_____TICK_____TICK_____TICK)
    DOT + (X)Haste -------> (CAST___TICK___TICK___TICK___TICK___TICK___TICK__)
    DOT + (X+Y)Haste -------> (CAST_TICK___TICK___TICK___TICK___TICK___TICK__TICK__)

    Now when you're talking about 'plateaus', these are theoretical numbers that would essentially allow you to not have to recast DOTS early, you could theoretically cast them as you used to in Wrath. This is because the DOT at certain points of haste gain the extra tick and revert back to the original length of 18 seconds. This is just likely never going to be the case so this is why the plateaus are theoretical.

    In summary: Haste is our best scaling stat because it scales pretty linearly due to the way DOTS now work. The DOT shortens in length until it hits a 'plateau' where the DOT gains an extra tick from haste. At that point the DOT reverts back to it's original length but with an added tick in it. This cycle is repeated ad infinitum (or until the hard cap of haste which is far too high a number to be achieved)

    Hope this helps.
    I do not believe the dead space after the final tick you have on the chart (in red) exist anymore. I remember this being the thinking during the Cata beta. But, while I was trying to hit my next haste plateau I found that the dot debuff on mob immediately falls off after the final tick. But what does happen (or is suppose to since I haven't actually tested this) is that if you recast the dot before the final tick, it won't clip it any longer and you'll get another full series of dots.

    Yes, I am a firm believer in haste plateaus since you gain ~2.3s between each VT cast going from say 2550 haste to 2589 haste (my situation at the time.) This is significant in multi-dot situations.

  14. #54
    Quote Originally Posted by Worshaka View Post
    I'm not sure that is evidence to suggest crit scales better than mastery. Firstly mastery is a direct % damage increase on a large portion of our damage and % damage increases always scale in a linear fashion.

    Crit on the other hand never scales in a linear fashion, the more crit you get the less % impact it has on your DPS. Going from 0% crit to 1% crit will increase your DPS by 1%, but going from 50% to 51% crit will do significantly less. You are quite correct in stating that you will get higher fiend uptimes but you have yet to prove that the damage increase from higher fiend uptimes is enough to make up the diminishing effect crit has on your dps %.

    All I can say is I would be rather skeptical that crit will scale better than mastery but I'm not stupid enough to declare it as fact based on what seems logical to me.

    Lastly, just to murky things even more, stats synergise togther.. in particular crit and haste will have a compound effect on fiend uptime. With higher haste there are more tick chances that can crit, so which stat is truely responsible for the higher uptime. Sometimes I can be just about impossible to disentangle and understand which stat is doing what, it's the combination of both that is giving the effect. Haste will also increase ES uptime which will have a positive effect on mastery as well...
    Depending on your perspective... crit scales in a nearly linear fashion, with the exception of the effect from shadowfiend. Crit's effect on shadowfiend is somewhat exponential, as more crit = short fiend cd, and in turn increases the fiend's chance to crit.

    If you figure you have a base DPS of 5000, adding 5% crit will, on average, increase your DPS by 250. Adding another 5% will continue to increase the DPS by 250.
    In this sense...
    Haste is slightly more than linear, mainly due to the fact of DoT scaling, and its effect on the rate of shadow orb procs.
    Intellect is slightly more than linear, because adding int also increases crit, thus increasing the value of all intellect.
    Mastery is completely linear.

    However... if you're going to say crit is non linear, I assume you're looking at crit from this perspective:

    Base 5000 DPS, 5% crit increases DPS by 5%. If you gain another 5% crit, then your new Base DPS of 5250 is now increased by 4.76%. Looking at stats from that perspective, nearly all stats have diminishing returns. Mastery, Intellect, Crit, and Haste will all suffer the same effect, haste and intellect to a lesser degree.

    I'm stressing the distinction between these two viewpoints because they are both equally valid. However, the first viewpoint is the one most commonly used by shadow theorycrafters. So, in order to avoid confusion of other readers, I'm clarifying.

    When in comes to mastery and crit, again, the main reason most will see mastery scaling better right now, is they've most likely been actively avoiding mastery, and now that mastery is only slightly under crit, at 372 BiS... they're seeing an increase because the optimum gearing would want you to try to split the stat points you have evenly between crit and mastery. So if you have 15% crit, and all of 2 mastery points over the base, mastery will scale better than crit until you get that balance closer to 8.5% crit, and 8.5 mastery points over the base.

    Most likely when the next tier of raiding is accessible, the effect of higher haste will begin to push mastery over crit. Of course, that all depends on what Blizzard has in store for us between now and then.

  15. #55
    Deleted
    I'd like to know if I should cast Shadowfiend BEFORE Heroism or later. I mean, is shadowfiend affected by Heroism once he is already attacking a target or once the caster is affected by heroism, and then he casts Shadowfiend?

    Excuse my poor english. Thanks

  16. #56
    Shadowfiend before heroism.


  17. #57
    Quote Originally Posted by Worshaka View Post
    Crit on the other hand never scales in a linear fashion, the more crit you get the less % impact it has on your DPS. Going from 0% crit to 1% crit will increase your DPS by 1%, but going from 50% to 51% crit will do significantly less. .
    It is for this reason that simulationcraft should be run to determine which stat is more valuable at a given time with each specific gear set up. Given extraordinarily high crit levels, you're right it would cease being as valuable as it is currently. It is for this reason that Blizz reset all the secondary stats to much lower levels when cata was launched. But the key point is that at current levels crit DOES scale approximately the same as mastery on the basis that millions of damage will be done by the player to the boss, thus drastically eliminating the RNG factor that crit can provide in more short term settings. With much higher mastery and crit levels, mastery will almost certainly begin to outscale crit for the reason you gave, but that point is a long ways off.

  18. #58
    Deleted
    I've "craftsimulated" some different combat types, with and without Dark Intent. With my current gear (356 item lvl) and in all of them mastery is the worst stat. And in a couple of results, crit was even more important than haste, specially if I didn't count with DI. So Crit > Mastery nowadays.

  19. #59
    Quote Originally Posted by Eldarc View Post
    I've "craftsimulated" some different combat types, with and without Dark Intent. With my current gear (356 item lvl) and in all of them mastery is the worst stat. And in a couple of results, crit was even more important than haste, specially if I didn't count with DI. So Crit > Mastery nowadays.


    Correction, Crit > mastery best for you.

    These stats are so 50-50, it can be better for someone else with higher or lower gear level.

  20. #60
    Deleted
    Quote Originally Posted by nagina View Post
    [/B]

    Correction, Crit > mastery best for you.

    These stats are so 50-50, it can be better for someone else with higher or lower gear level.
    Ok, you're right. I agree with you. But other simulations I've read say exactly the same about crit > mastery.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •