Shadow Priest haste

[konway]

Oh, let's completely forget the fact that 1% haste is a 1% increase in DPS for all non-CD spells for all caster classes. Yep. Haste is absolutely atrocious if you're not hitting a breakpoint.

The fact that you state that as if you're actually proving a point just shows how little you know. How many non-CD, non-DoT spells do we cast? Mind Flay? So you'll increase the DPS of our worst spell by 1% whenever you go for haste in between breakpoints? Brilliant, I'll start re-gemming right away.

Do you know the DPS % increase for 1% of crit or mastery for comparison? Or are you only focusing on haste because it's what all the cool sim kids are doing? Just FYI, 1% crit is also a 1% DPS increase... for ALL spells, and you don't need to worry about reaching breakpoints to see that increase.

1% mastery is also at least a 1% DPS gain for most of our spells, plus you get more instant Mind Spikes and Mind Blasts because of the additional SW:P and VT ticks. Again, no breakpoints to worry about before you start seeing the benefits. Every single point is worthwhile no matter how much you currently have.

---------- Post added 2012-11-07 at 03:52 PM ----------

I'm no theorycrafter, but it's not rocket science to note that a dot gains a tick ans lasts longer, giving you time to do something else.

It may not be rocket science, but it definitely isn't that simple either. Each 1% of haste does not grant an additional tick to all of your DoTs. If you read the rest of this thread, you'd notice that you gain zero DoT ticks between ~2020 haste and ~8060 haste. That's a pretty significant gap considering the amount of reforging and gemming that needs to be done to hit the latter number in average gear right now.

no Haste will always be our best secondary stat. The only current discussion is whether crit > mastery or mastery > crit and nether has really confirmed anything yet. they are both so incredibly close that it honestly wont change anything yet. SO there really isn't anything to worry about it. Go haste> crit=mastery.

In the thread you linked, the 250PP value seems to be derived from a 1% difference between sim results. Maybe I'm being overly cautious, but I think it's a little naive to assume a sim created by non-Blizzard programmers with no access to the WoW source has a <1% margin of error.

[Aica]

Each 1% of haste increases the DPS of your DoTs by 1%. Each 1% of haste increases the proc chance of (and subsequent DPS of) DI and FDCL by 1%. Each 1% of haste increases the DPS of MB and DP by a small amount.

Haste speeds up your entire rotation except for MB, SW:D (which still speed up a small amount) and DP (which improves at breakpoints, and speeds up at the same rate as MB and SW:D). It costs 425 rating for 1% increased dps on most of your spells.

Crit increases the damage of your entire rotation. It costs 600 rating for slightly more than 1% increased dps for all your spells.

Mastery increases the damage of Mind Flay, SW:P, VT, DP and Mind Sear. It increases the proc chance of FDCL and DI. The FDCL benefit is marginalized due to MF's damage getting increased while MSP's is not. FDCL and DI are also bad choices for a decent number of encounters. It costs 333 rating for 1% increased dps on a bunch of your spells.

So, a high rating required for all spells, a medium rating required for most spells, and a small rating required for some spells. It mostly balances out. Haste comes out on top...for now (not even taking into account breakpoints). Next tier, once we're sitting comfortable on enough haste, it could be that the other ratings will rise.

The fact that you are not mentioning how haste affects dots outside of breakpoints speaks volumes. Haste's strength has barely anything to do with breakpoints.

[konway]

Each 1% of haste increases the DPS of your DoTs by 1%.

This is 100% wrong and basically invalidates the rest of your analysis. Your DoTs gain no benefit from haste in between breakpoints aside from the GCD reduction.

You would be correct if we were still pre-Cataclysm. Back then, every point of haste would reduce the duration of your DoTs while still maintaining the same number of ticks -- so they would esentially tick faster. This is obviously no longer the case, hence all the math to determine break points.

[Arlee]

This is 100% wrong and basically invalidates the rest of your analysis. Your DoTs gain no benefit from haste in between breakpoints aside from the GCD reduction.

You would be correct if we were still pre-Cataclysm. Back then, every point of haste would reduce the duration of your DoTs while still maintaining the same number of ticks -- so they would esentially tick faster. This is obviously no longer the case, hence all the math to determine break points.

What do you mean it obviously is not the case anymore? You are basing that on what?

[Aica]

This is 100% wrong and basically invalidates the rest of your analysis. Your DoTs gain no benefit from haste in between breakpoints aside from the GCD reduction.

You would be correct if we were still pre-Cataclysm. Back then, every point of haste would reduce the duration of your DoTs while still maintaining the same number of ticks -- so they would esentially tick faster. This is obviously no longer the case, hence all the math to determine break points.

You have no idea how dots work. Go do some research and come back.

[Twintop]

This is 100% wrong and basically invalidates the rest of your analysis. Your DoTs gain no benefit from haste in between breakpoints aside from the GCD reduction.

You are 100% wrong, since we're speaking in absolutes.

If I have a DoT that ticks every 3 seconds for a 5 minute (300 second) fight, that gives you 100 ticks (assuming you don't let it drop off).

If I have a DoT that ticks every 2.5 seconds for a 5 minute (300 second) fight, that gives you 120 ticks.

How does Haste not increase the damage from DoTs again?

You would be correct if we were still pre-Cataclysm. Back then, every point of haste would reduce the duration of your DoTs while still maintaining the same number of ticks -- so they would esentially tick faster. This is obviously no longer the case, hence all the math to determine break points.

If I have to recast a DoT every 16sec because of haste, then I'll be casting it at (roughly) 16sec, 32sec, 48sec, 64sec, etc. Over a 5 minute fight, that means I spend 20 GCDs (19.35) casting that DoT. If I push my haste far enough to get an extra tick and thereby increasing the duration to 17sec, then to get the same number of ticks I only spend 18 GCDs (17.65) casting that DoT. I gain 2 GCDs over the course of a fight that I would have been spending recasting DoTs, but instead I can use another ability. THIS IS WHERE THE HASTE BREAKPOINT DPS GAIN COMES FROM, NOT FROM THE DOT.

Seriously. Math isn't hard. You refuting these claims only exhibits your ignorance toward high school level math. This isn't calculus. This isn't even statistics.

[konway]

What do you mean it obviously is not the case anymore? You are basing that on what?

I can't give you a link to patch notes or anything, but the easiest way to see the new system in action would be to look at the duration of your DoTs with haste gear/buffs active. An 18 second SW:P would be cut down to almost 14 seconds with ~20% haste prior to Cata.

Here's a quote from wowpedia:

As of patch 4.0.1, all DoTs can crit, and all spell DoTs and some physical DoTs innately scale with haste. Unlike prior implementations of DoT haste scaling, DoTs do not lose duration when affected by haste. Rather, the game engine will add additional ticks to the DoT to keep the total duration of the DoT as close to the original duration as possible, plus or minus up to half of the (hasted) time between ticks.

If the old system were still in place, there would be no talk of haste break points because they simply wouldn't exist.

[Frmercury]

Referring to the sim, which seems to be the main focus of theorycrafters these days.

Actually, now that I think about it, there really doesn't seem to be much old-fashioned theorycrafting going on anywhere. As soon as there's any discussion about gearing or playstyle, people run straight to their sims and then cling to the results for dear life. It's as if everyone has lost the ability to think for themselves, so they're relying on some third party program to do it for them.

There's one thing we know for certain about the sim: It's not 100% accurate. We don't know how inaccurate it is, but we know for sure that it can't mimic the game exactly. They don't have access to the WoW source code, so it's not their fault. The problem is that no one seems to take this into account when making gear decisions based on sim results. Everyone just states the sim results as if they're undeniable facts and then starts calculating gem and item values based on those numbers.

You could have just said "No, no I don't. I think I have an idea what it is and I don't like that but I've never actually taken the time to actually understand it."

See, that was much easier.

[Aica]

I can't give you a link to patch notes or anything, but the easiest way to see the new system in action would be to look at the duration of your DoTs with haste gear/buffs active. An 18 second SW:P would be cut down to almost 14 seconds with ~20% haste prior to Cata.

Here's a quote from wowpedia:


If the old system were still in place, there would be no talk of haste break points because they simply wouldn't exist.

Seriously, man. Go test it in-game for yourself. Dot duration decreases with haste. At a haste breakpoint, the dot's duration resets. The only thing you get out of a haste breakpoint is two spare GCDs per dot cast.

[konway]

You are 100% wrong, since we're speaking in absolutes.

If I have a DoT that ticks every 3 seconds for a 5 minute (300 second) fight, that gives you 100 ticks (assuming you don't let it drop off).

If I have a DoT that ticks every 2.5 seconds for a 5 minute (300 second) fight, that gives you 120 ticks.

How does Haste not increase the damage from DoTs again?

Each 1% of haste is not a 1% DPS gain on DoTs. Between break points, the only gain is GCD reduction.

If I have to recast a DoT every 16sec because of haste, then I'll be casting it at (roughly) 16sec, 32sec, 48sec, 64sec, etc. Over a 5 minute fight, that means I spend 20 GCDs (19.35) casting that DoT. If I push my haste far enough to get an extra tick and thereby increasing the duration to 17sec, then to get the same number of ticks I only spend 18 GCDs (17.65) casting that DoT. I gain 2 GCDs over the course of a fight that I would have been spending recasting DoTs, but instead I can use another ability. THIS IS WHERE THE HASTE BREAKPOINT DPS GAIN COMES FROM, NOT FROM THE DOT.

Seriously. Math isn't hard. You refuting these claims only exhibits your ignorance toward high school level math. This isn't calculus. This isn't even statistics.

You're willing to sacrifice thousands of points in other stats to gain 2 GCDs in a 5 minute fight (roughly 700 DPS), and I'm the one who's ignorant of math? That's rich.

---------- Post added 2012-11-07 at 06:19 PM ----------

You could have just said "No, no I don't. I think I have an idea what it is and I don't like that but I've never actually taken the time to actually understand it."

See, that was much easier.

Is it too difficult to back your post up with some substance? Or are you just suffering the effcets of cognitive dissonance because you know you're wrong but you can't stand to give up your precious sim figures?

[Arlee]

You're willing to sacrifice thousands of points in other stats to gain 2 GCDs in a 5 minute fight (roughly 700 DPS), and I'm the one who's ignorant of math? That's rich.

Who said anything about sacrificing 1000's of points of other stats for haste?

[Elmi]

Dot duration (respectively the time between ticks) scales linear with haste. Reaching a breakpoint adds [tick time] to your dot duration. This does not directly increase the dps of the dot but its dpe/dpet and therefore your total dps.

[konway]

Seriously, man. Go test it in-game for yourself. Dot duration decreases with haste. At a haste breakpoint, the dot's duration resets. The only thing you get out of a haste breakpoint is two spare GCDs per dot cast.

As mentioned in the quote, DoT duration can be altered by up to one half of the time between ticks. That amount can either be added or subtracted.

---------- Post added 2012-11-07 at 06:24 PM ----------

Who said anything about sacrificing 1000's of points of other stats for haste?

If you want to reach the 8085 haste breakpoint without full heroic BiS gear, you're going to be neglecting a lot of other stats.

[Aica]

As mentioned in the quote, DoT duration can be altered by up to one half of the time between ticks. That amount can either be added or subtracted.

Dot duration is reduced until the duration is 1 GCD below the unhasted dot duration, at which point the duration resets to 1 GCD above the unhasted dot duration. Regardless, the time between dot ticks decreases with every single point of rating you gain, and as such, the DPS of the dot increases with every single point of rating you gain. Hitting a breakpoint changes nothing about this.

[Twintop]

You're willing to sacrifice thousands of points in other stats to gain 2 GCDs in a 5 minute fight (roughly 700 DPS), and I'm the one who's ignorant of math? That's rich.

Who said anything about sacrificing 1000's of points of other stats for haste?

Pretty much this right here.

It's obvious you've not done any investigation about this topic and are spouting off your opinion while disguising it as fact. You may recognize this tactic from the elections last night, because that strategy worked so well for all of the Nate Silver haters last night with the electoral projections, claiming their gut was telling them one thing but Statistics proved another. Interesting parallels here...

If you had bothered to read anything I've posted about this subject in numerous places--or had read the thread on H2P that was linked earlier in this thread where Dyre lays out the exact same point--you'd already know that I've been telling people since before the expansion hit that if you're sacrificing more than a couple hundred Intellect to reach a breakpoint you're hurting your DPS. Isn't that rich? Seems like math is still winning here (along with consistency).

[konway]

Dot duration is reduced until the duration is 1 GCD below the unhasted dot duration, at which point the duration resets to 1 GCD above the unhasted dot duration. Regardless, the time between dot ticks decreases with every single point of rating you gain, and as such, the DPS of the dot increases with every single point of rating you gain. Hitting a breakpoint changes nothing about this.

The duration alteration is based on half the time between ticks of the DoT. And if you reduce the duration, you'll spend more GCDs keeping DoTs up, which is one reason people are so focused on the 8085 break point. 2 GCDs in a 5 minute fight seems like a pretty marginal gain to me...

---------- Post added 2012-11-07 at 06:58 PM ----------

If you had bothered to read anything I've posted about this subject in numerous places--or had read the thread on H2P that was linked earlier in this thread where Dyre lays out the exact same point--you'd already know that I've been telling people since before the expansion hit that if you're sacrificing more than a couple hundred Intellect to reach a breakpoint you're hurting your DPS. Isn't that rich? Seems like math is still winning here (along with consistency).

/sigh

If you had read the thread, you would see that I had already addressed that exact post and the point it linked to:

In the thread you linked, the 250PP value seems to be derived from a 1% difference between sim results. Maybe I'm being overly cautious, but I think it's a little naive to assume a sim created by non-Blizzard programmers with no access to the WoW source has a <1% margin of error.

Perhaps we should consult Nate...

[Frmercury]

The duration alteration is based on half the time between ticks of the DoT. And if you reduce the duration, you'll spend more GCDs keeping DoTs up, which is one reason people are so focused on the 8085 break point. 2 GCDs in a 5 minute fight seems like a pretty marginal gain to me...

You seem to not have any facts behind... well anything you've posted. Just that everyone else is wrong because that's your feeling.

Feelings != Facts, just an fyi.

[Twintop]

If you had read the thread, you would see that I had already addressed that exact post and the point it linked to:

Yes, where "addressed" means "here's my opinion with nothing to back it up." You're argument is a strawman at best.

Perhaps we should consult Nate...

So you trust 538's predictions. That are a model of a system. Where inputs and outputs are known and results can be verified for their correctness...

...yet all the same things can be said about SimulationCraft? Tell me the difference between the two then, please. All of the interactions are documented and behaviors can be and have been verified.

If you want to continue wearing your tinfoil hat and screaming conspiracy, be my guest. Excuse the rest of us while we point and laugh.

[konway]

Yes, where "addressed" means "here's my opinion with nothing to back it up." You're argument is a strawman at best.



So you trust 538's predictions. That are a model of a system. Where inputs and outputs are known and results can be verified for their correctness...

...yet all the same things can be said about SimulationCraft? Tell me the difference between the two then, please. All of the interactions are documented and behaviors can be and have been verified.

If you want to continue wearing your tinfoil hat and screaming conspiracy, be my guest. Excuse the rest of us while we point and laugh.

When it comes to the accuracy of SimulationCraft, I believe the WoW devs over the sim devs. It has been stated by blue posters on numerous occasions that the sim is not accurate. Ghostcrawler is especially fond of pointing this out.

[Burchoid]

If you want to reach the 8085 haste breakpoint without full heroic BiS gear, you're going to be neglecting a lot of other stats.

Dude, are you trolling? Because it sure is starting to seem like it... unless you are getting no gear with haste as a base stat, your statement is quite inaccurate. I was able to reach the haste breakpoint in less-than-ideal 480ish gear without sacrificing stats as you suggest. Are you confused about stat weights perhaps?