I just tried this out: I popped a haste pot and activated my haste trinket (resulting in 1562 haste rating) - I got 8 ticks

edit: It seems like corruption falls off the target before it should - at 1562 haste rating it gets removed from the target when there is 2 seconds left. It must be a bug related to the mechanics (1562 haste rating is rather close to '9 ticks' but not quite there, thus 'clipping' corruption short )

edit2: Using the formula from above, we'll be seeing 'haste caps' once this goes live - 17% (to gain 1 extra tick) and 34% (to gain 2 extra ticks). Anything below or in between is basically 'wasted' haste (at least for the scaling of corruption) since you won't gain anything from it. Not to be a negative nancy, but that really sucks; that would be like rounding up cast times on nukes.

A simple illustration at the ex. of Devouring Plague:

I don't see how this would have hastecaps since the next Devouring Plague (Corruption) will be added right after the other - the benefit of decreased time between dotticks through haste is still existent. The only downside I see is at the end of a fight though, but this should not be significant.


Edit: Devouring Plague <-> Corruption

This seems to indicate it always rounds down, unlike devouring plague which (according to Althor above) rounds to the closest whole tick.

This is how it works on live as well - the duration as shown in the UI is always reset to 18 seconds, but the first tick of the refreshed corruption will happen exactly 3 seconds after the previous tick, not 3 seconds after haunt lands. This means the last tick can happen up to 3 seconds before the 18 second duration up, making it look like it "fell off early". Judging from the descriptions I'm seeing, with the glyph it just fits as many ticks as it can into those 18 seconds. This makes sense, and explains why it never rounds up.

This really isn't true, unless they changed the way the dot refreshing works. The way it works on live, you get a corruption tick every 3 seconds regardless of the exact timing of haunt. With the glyph you should get a corruption tick every (3 / (1+haste)) seconds, meaning it'll scale perfectly with haste.

I was worried it would create a pretty low haste cap by shortening corruption's duration by the full haste percentage, but if it turns out it increases the number of ticks, that's not a concern anymore.

In the end it's irrelevant how many extra ticks you get on your Corruption, they are only a convenience to make sure Corruption doesn't drop off before you can refresh it with Haunt. It also prevents you come in situations where you need to cast Haunt on cooldown as ideally you want to cast it a few seconds later to refresh the debuff as late as possible.

The dps benefit of the glyph is in making the time between Corruption ticks shorter, and more ticks per minute means more dps.

Another nice benefit of the glyph is that the haste to dps conversion does not come at the cost of more mana per second usage where it with casted spells does.

It looks like the 3.3 Felguard dps might be under estimated for deep demonology? The 3.3 Simulationcraft numbers currently on the first page have Warlock_T9_00_56_15 with felguard DPS=2104.8. That's lower than what the 3.2.2 SampleOutputT9_Details on code.google.com haave with felguard DPS=2139.8. I ran 3.2.2 and 3.3 simcraft myself, and got felguard dps as about the same for both at 2106.9 and 2107.5 dps.

I'd expect the deep demo felguard dps to be increased in 3.3 because the increased Demonic Pact spell power to the warlock should translate into increased attack and spell power for the felguard.

And comparing training dummy tests on live and PTR shows felguard attack power and dps higher on the PTR. To test this out I wore the exact same pvp gear set (my pve gear sets on PTR vs. live are quite different) and was unbuffed except for fel armor. I did about 35-40 more dps on the PTR. On live, my felguard's attack power with demonic pact and 10 stacks of demonic frenzy up = 4851 (1094 + 3757). On the PTR, it was 5027 (1094 +3933).

So, it looks like simcraft is underestimating 3.3 felguard dps, perhaps due to not factoring in how Demonic Pact's increased spell power to the warlock translates into increased attack and spell power for the felguard.

The difference between felguard DPS in the OP and what you see in SampleOutput is due to a 1.4% miss rate - I believe this started happening when we changed how hit chance is inherited from master to pet. I'm actually pretty sure that's not correct, I think the felguard is hit capped if we are, but as you can see it only makes up about 33 DPS, so the difference isn't exactly major.

I don't know how to explain the difference you're seeing in the felguard's attack power between live and PTR - the Demonic Pact change is a straight 5% damage boost to your own spells, and shouldn't affect the felguard at all.

I just had a real quick question and I am sorry if the idea has already been brought up. For the upcoming 3.3 affliction 55/0/16 build, would it be worth even considering moving the 1 point from Cataclysm into Fel Synergy? As a deep desto lock currently I really like the fact that I don't have to pay that much attention to my imps health, and since the felhunter will have alittle more worth coming up, would it be useful? Not sure what kind of damage the new instances will be bringing. Let me know what you all think.

I'm currently running some tests on the level 83 dummy to figure out expertise and hit scaling. I think I've more or less figured out expertise scaling (And I believe Dedmon is implementing that), I'm testing out hit scaling atm. What I can say though is that with 13% hit I'm seeing 0% felguard miss in raids (so as a temporary measure it might be worth while to just make attacks not miss).

At the moment I have 2 theorys on how it scales;
1] 1 master hit rating = 1 pet hit rating
2] 1% master spell hit = 1% pet melee hit.

Some other things I need to test:
* does suppression effect pet hit/expertise scaling?
* does heroic presence effect pet hit/expertise scaling?

Since the pet avoidance goes from 70% to 90% i dont think it will be need it.

Based on blue posts by Ghostcrawler/3.2 Patch notes, it's the latter. It should take the percentage of hit rating you have vs the amount necessary to never miss, and applies this to expertise and hit rating for your pet. For example, if you have 17% Hit rating, thats 100% of the required Hit Rating so you get 100% of Expertise and Melee Hit for the pet, which works out to (if I remember) 6.25% Expertise and 8% Melee Hit. If you have 14%, thats 14/17 = 82.35% of maximum, so your pet gets 6.25 x 0.8235 = 5.1469% Expertise and 8 x 0.8235 = 6.588 Melee Hit %.

Suppression should factor into this because it affects your tooltip. I'd be really interested in how Misery/Improved Faerie Fire work into this, as well as Heroic Presence.

However, your tests seem to be proving this wrong. It'd be interesting to see the final results as the way it was worded 'officially' may not have worked out in application.

After very extensive testing...

* With 0 hit rating on gear, the felguard misses 8% and gets dodged 6.5%. (As expected)
* You need 17% hit from gear for the felguard to get dodged 0%
* Felguard chance to be dodged is:
* Felguard's chance to miss is:
(The felguard gains 1% chance to hit for every 1% spell hit the master has, or in other words, the felguard gains 32.79 melee hit rating for every 26.23 hit rating the master has.)

I will test the effect of supression and hopefully heroic presence tomorrow (or later tonight).

New numbers are in the OP. The felguard miss/dodge numbers seem to line up pretty well with what I'm seeing in my raid logs.

With suppression: At 5% spell hit my felguard is missing, at 14% spell hit my felguard is getting dodged. Basically the felguard doesn't scale with suppression.

Does Suppression show up in the spell hit percentage on your char sheet in game?

No.

If Suppression does not show up, then we need to move the relevant code from composite_spell_hit() back to warlock_spell_t() constructor.

It also explains why Felguard does not benefit from Suppression. Basically, if a talent doesn't make the paper doll, then 99% of the time it isn't considered a "class-global" effect. Instead it is implemented per-ability. This makes sense given Suppression's history.

EDIT: Done.

This was fixed the day of the patch notes that mention the change. The version you have would be from before the announced change and fix.

The crit multiplier is actually correct; for the brief amount of time that change was applied Shadow Bite was critting for a 2.5 multiplier, but Simcraft was never updated. Basically, as a melee attack, it already crits at 2.0 without pandemic.

I did update SimulationCraft to model the 2.5 multiplier a little while after reading about it on this thread and testing it for myself.

However it wasn't too long after I made the change that Pandemic was removed on effecting it on the PTR and I subsequently *mostly* reverted the change in SimulationCraft. I say mostly because my testing combined with the wording on Pandemic when it effected Shadow Bite showed that the base crit multiplier for a Felhunter's Shadow Bite attack is not 200% like say a Rogue's specials, but is in fact 150%. The difference is made up by an innate 100% crit bonus modifier (100% of the +50% damage that a spell crit does).

For the purposes of a Felhunter I imagine the distinction is meaningless as I'm unaware what effect if any a Chaotic Skyflare Diamond has upon pet criticals. I imagine none. If however some ability was added similar to the CSD meta that effected the pet, then the slight change I made would help model it correctly.

New glyph and change to shadow mastery?

3.3 PTR: Conflagrate: Redesigned. This talent now consumes an Immolate or Shadowflame effect on the enemy target to instantly deal damage equal to 9 seconds of Immolate or 8 seconds of Shadowflame, and causes additional damage over 3 seconds equal to 3 seconds of Immolate or 2 seconds of Shadowflame.

The additional time on Immolate will have no effect on Conflagrate's damage.

Has anyone tried using Glyph of Quick Decay with the 0/56/15_MC demo spec?

Since the uptime of the buff right now is 27% for the spec, basic probability tells us that since we know the trigger chance is 12% every 3 seconds, the number of ticks for each average buff uptime is 'x' in the equation below:

(1-.12)^x = (1 - .27)
x = 2.46

That is to say, the chance of having the buff down for a particular cast (the downtime) is (1 - .12)^x, and it is also one minus the uptime. Thus, if we speed up our corruption ticks by ~ 30%, x jumps up to 3.2, putting our buff uptime at 34%. Wouldn't the extra 7% uptime on MC combined with the extra damage on corruption be worth losing the 1.2% damage buff (18 seconds of 20% damage buff uptime over a 5 minute fight) glyph of meta provides?