Has any testing been done on PTR as to whether either leeching poison and/or recuperate interact with the RPPM mechanic on some of the new trinkets? Dancing steel currently procs from recuperate which is RPPM. I am not sure as to whether leeching poison can trigger a dancing steel (or indeed any other RPPM proc). Not that I would endorse the use of recuperate as a means of trying to increase dps, rather survivability or throw away combo points.
That haste % in your formula is actually the highest of your attack speed increases. So SnD increases the proc rate of RPPM trinkets by 1.4 for Mut and Combat, and more for Sub. This is multiplicative with other attack speed increases. So mastery increases SnD's strength, which in turn increases RPPM proc rates.
There's no DPS to gain by using Leeching or weaving Recuperate with RPPM mechanics. With a proportionally smaller interval between events comes an equally proportionate chance to proc. This doesn't increase DPS.
Just a reminder that RealPPM now uses your true melee haste or spell haste (whichever is higher) to scale the proc rate instead of attack speed or casting speed. The difference is that things that buff your attack speed (such as the raid buff or Slice and Dice) do not increase the frequency of RPPM procs
Oh dear. Is it likely that the initial 5.2 release of Shadowcraft will factor in this recent change, as Pathal only just stated to the contrary? Blizzard do like to keep the Sim/Theorycrafters on their toes.
Great stuff. The proc scaling with item level is an interesting addition (not sure if this had been announced before) which accounts for the presence of the same proc on each trinket.
"- Proc rate multiplied by 0.1 for non-melee specs. Proc rate multiplied by 0.25 against player controlled units. Proc rate multiplied by 1/(1.15^((528-ItemLevel)/15) for the various Item Level versions of the trinket."
Whether this formula will equate to a proportionally larger proc (which we have been used to, from same trinkets but of different item level), in terms of dps gain remains to be seen.
Quick hack: Pass (haste_multiplier *get_heroism_haste_multiplier() +0.05) into proc_rate method instead of attack_speed_increase. I may have the heroism multiplier in the wrong spot I'm not sure. In the long run that .05 should be dependent on the spell haste buff but as a short hack I just assumed it was present.
I'm not seeing anything off with Bad Juju based on this hack.
EDIT: I'm not adding the .05 (which should actually be '* 1.05' ) until we figure out if the spell casting buff even applies.
No it shouldn't be *1.05, the spell haste buff applies additively not multiplicatively. That said I think its fair to not include it without confirmation since haste mechanics are one place where Blizzard's terminology is somewhat inconsistent.
What is it that makes sub able to play the RPPM Items better than the other specs?
From my understanding RPPM works as follows:
X(ppm) * (haste % + 1) * (time since last chance in sec) / 60 (sec per min)
Is the answer the increase to SnD from mastery, granting more weapon swings and therefore more but smaller chances with each swing? Or is the mastery somehow affecting the haste component in the equation?
Well, mastery won't make a difference anymore. RealPPM is now going by actual haste and ignoring any attack speed buffs.
We changed RealPPM to use your true melee haste or spell haste (whichever is higher) to scale the proc rate instead of attack speed or casting speed. The difference is that things that buff your attack speed (such as the raid buff or Slice and Dice) do not increase the frequency of RPPM procs. True haste boosts, such as Bloodlust or haste rating on gear will improve proc rate. We felt this change was necessary to balance proc rates among different specs.
I don't think this is the root of the bad juju issue however it appears the proc_rate method is using the base ppm rather then the ilvl adjusted ppm.
Actually I'm not sure there is an issue with bad juju. The uptime comes out to just under 20% with 19% haste. Using the formula for expected proc interval (60/(rppm*haste)) we get an expected interval of just over 100 seconds with 19% haste. 20 second uptime with procs every 100 seconds should be around 20% uptime.
EDIT: Maybe the issue identified above is related. When computing dps without any computation of trinket EP values the proc_rate method uses base_ppm. When using the get_other_ep method the proc_rate method appears to be using the lfr ppm.
That's because if I don't remember the base value (which is only used to be changed into the real ppm), we'll keep multiplying that ppm by whatever their ilvl scaling formula generates. IE LFR trinkets go from 0.5, to 0.27285, to 0.148894245, etc due to loops. That's the only time that piece of data is ever called on, and it keeps the loops from letting things spin out of control.
Anyways. I've got work early in the morning. I'll have to sleep on this and hope I figure it out before then. Indz will probably get everything else running so the rest of the engine is working for the patch (it's better than nothing). Trinkets will probably be off/wrong tomorrow, but the rest of the engine should be running full tilt, I'll get right back at this when I get off work if someone doesn't beat me to the punch.
I noticed that other independent theorycrafting has hinted that the trinkets will suck, so maybe I just need to revisit some of the basic assumptions and determine uptime and average stats by hand.