@Ravhin I only had the 245 version so it wasn't too hard, looking at DEP only, to see the huge jump. I really prefer the style of trinket Reign is, but if the numbers tell me something else is better I can't see a reason not to upgrade.

@Zamir. When you choose your trinket on the main page it pulls the avr proc # to calculate your final DPS? Seeing that instead of the DEP value is what confused me the most. My mind is trained to perceive bigger numbers as better numbers and going from +dmg proc to +sp proc saw the number shrink. I don't know how to make it less confusing but maybe that helps you understand why I was confused in the first place.

I'm hesitant to start tweaking away from the pre-set rotation scheme but I also know that with all the movement I do it would benefit me to try and model realistic conditions so my trinket considerations can be more accurate. I'll have to mull this one over for awhile I guess.

Thanks :-)

ZAP! already has two features for modelling movement. The first is the "Movement per LvB" setting which inserts a delay after every Lava Burst cast - you can fill this delay with available instant cast spells if enabled via the UI. The second lets you reduce the damage of any spell by a flat percentage, so if you see in your logs that a particular fight has 20% less Lightning Bolts or something you can reduce LB's contribution to your damage by 20%. These are both in the advanced options thing on the right of the main page.

ZAP! rates heroic Muradin's Spyglass as ~200 DPS upgrade over normal Reign of the Dead but Mr. Robot rates it as ~60 DPS decrease.

Does anyone know if the two model it differently? I figured it would just be rated as a flat 200 spell power because it shouldn't ever fall off and maybe it can be downrated in flat spell power to account for ramp up time.

How are you getting a 200 DPS difference? With my stats the difference in DEP between the two trinkets is only 67. You're not just selecting the trinkets in the Equipped Items section and forgetting their base stats, I hope?

Tee hee, yeah that would do it. Forgot to change the base stats and now the projected DPS only increased by 94.

Still, curious what the difference in modeling is for ZAP and Mr. Robot, because Mr. Robot has it as a 64 DPS decrease.

Found some odd behaviour for you.

With my current setup, 914 haste gives a normal haste DEP value. 915 gives ~12, and 916 goes back down again. As far as I can tell it all hinges on how you calculate your EM uptime. With no T10 2pc I don't think I can expect to get 99 seconds of EM use if I have a 180 second cooldown (no glyph).

What you really need to do is work out the number of times it's used, multiply by 15 (to get a slight under-estimate) and then divide by total fight time to get the real %, which will be close to your 8.33% Realtime figure.

In fact, if you use AO20 as your calculator for the number of casts, you can dispense with extra calculations in AO19 & AO21

Some small improvements to the way trinkets are handled and a new pointless gizmo:

Changes - version 1.2.8

• Added Charred Twilight Scales (normal & heroic) to the trinket list, as well as the ICC healer trinkets, Purified Lunar Dust, Talismans of Volatile Power and Figurine - Twilight Serpent because I get asked about them a fair bit.
• Trinket procs are now modelled based on chance to proc and cast frequency. The "proc every ..." setting is gone.
• The trinket list can now be sorted alphabetically!!
• New tab: Suggestions. Houses the new Trinket suggester as well as the Gem suggester of yore.
  • Trinket suggester has filters for raid size, difficulty and tier/patch as well as hit rating.
  • Returns best two trinkets based on what's filtered.
  • Hit/no-hit is handled by specifying how much hit rating is needed/allowed. e.g. if you need 91 hit rating, Dying Curse would count all 71 of its hit rating as DEP but Maghia's would only count 91 of its total 152 hit rating.
  • Settings used for the filter will affect the main trinket list on the "Trinkets & Procs" tab, effectively filtering the sorting of that list too. This includes the hit rating filter, which affects the "DEP (hit)" column.
• Layout and descriptions of the trinket selection menu on the Main tab, as well as the Trinkets & Procs tab, have been changed slightly to hopefully make it more obvious what's going on.

Changes - version 1.2.8a

• Elemental Focus Stone now correctly filtered as a 10-man trinket.
• Small improvement to the backup Haste DEP thing.

My home internet is down at the moment which has made it rather hard to get this update out. I mainly wanted to have the revised trinket list up in case Ruby Sanctum is opened this week but in the fairly near future I also want to tidy up the way the haste DEP formulator selects which calculation to use.

Let me know if you notice any errors in the trinket data. I got it all from wowhead, checked it with other sources and think it's all OK but it's possible that some weirdness will have snuck in.

Download
Compatibility version

@Zamir,

Thanks for the update; professional as ever.
I don't see any weird stuff in the latest version and the trinket list seems more accurate in terms of ranking at the moment.
I do see one peculiarity; if haste values exceed the 1350-ish mark (for my stats and 0.8s movement and 60ms lag), the DEP of haste drops dramatically and much more then I expected (since only LvB, CL and FlS casts would find no effect of more haste). It seems to drop with more then 25% whereas we estimated some time ago this was more in the regio of 10-15% for haste-points past the softcap.

Can you check whether this is indeed correct, as it may just happen that if one is lucky enough to pick up both haste/sp trinkets this value is exceeded easily. It kinda skews the value of the second haste trinket considerably, and may in fact indicate it is not wise to go for both DFO and Charred Twilight Scale. Please advise.

Thanks!

I just checked out your current stats plus 1350 haste (with Bizuri's on top for 1570 total) and the values were pretty much as expected, haste's DEP dropping by about 20% in absolute terms or 30% relative to spellpower's DEP. With the advanced settings you gave the difference was actually smaller, and in both cases haste's DEP was still above that of crit. The main change we've previously discussed as taking place at high haste levels is that yellow socket bonuses become less attractive to gem Reckless for, and the results I got are consistent with that.

I didn't try with any Chain Lightning use.

At really high haste levels, with EM+Heroism+Normal rotations to account for (+trolls/engineers if relevant), the sheer number of "tip-over" points in the various different haste/rotation formulators could in theory cause both the normal and backup haste DEP to hit inaccurate points at the same time. I've not seen it happen since I released 1.2.7. If it does happen, changing the DEP step would fix it so that at least one haste DEP value returns a sensible number. But yeah, I didn't see anything unexpected using your stats - you could upload a copy of the sheet with the settings that cause anomalies?

@Zamir

Well, the effect was stronger then expected but I don't see anything special that would indicate the haste-DEP values are wrong. I am not looking at specific dips in the haste-DEP curve for this exercise but look at a broader range; so I am not too concerned about the 'anomalies' that might occur with my stats which are caused by the rotation model anyway which will vary a bit in actual fights.

In near BiS gear this lower DEP value does have the consequence however that if you have already one of the 277 haste trinkets, there is no meaningful dps difference anymore between Phylactery 277 and DFO/Charred 277. In other words, our trinket selection has broadened without changing the original BiS list. I regard that as good news; it means I am ok if I just score one of the 2 haste trinkets out there, instead of both.

I can't completely judge for 264 gear sets as I didn't look at it; I suspect there still dfo+charred > dfo/charred + phylactery.

Thanks for checking.

Hey guys, I have a quick question about the ZAP! spreadsheet. I currently have ran both the variety one and the normal one for MS Office. I am using OPEN OFFICE 3.2. It works fine and everything but I calculate all of my gear/stats/buffs/spec everything and where it is suppose to say the final dps it just says DPS : #NAME! Does anyone know my reasoning for this?(probably something dumb I'm missing that isn't calculating it.) Thanks .

Stopping by to make a request. Can you add Potion of Speed to the Calculators section under Cast Time and Spell Information? It would be nice to be able to add that in to mix and match haste buffs to bring LB as close to 1s as possible.

A couple of minor changes that have been requested:

Changes - version 1.2.8b

• "Potion of Speed" is now an option in the Cast Time & Spell Information calculator. Ancestral Spirit is also in there now to balance the display and because it's cool to rez people really fast.
• Pawn export strings now use DEP values rounded to 3 decimal places.

Download links:
Normal / Compatibility

Hey Zamir,

I noticed when I was messing with ZAP that you changed the Sandbox for the 'Snapshot' which does not allow you to create custom rotation scenarios. I was wondering if it was possible to combine these two features? I really like how the Snapshot has the EM / Bloodlust / Normal "modes" but I really want to have the customization of the old Sandbox.

If not it's not a big deal, Thanks!

So far as I know nobody else actually ever used the Sandbox. The rotation formulator used since 1.2 is basically the old sandbox, but automated, tightened up and expanded. It'd be quite easy to copy that over to the old sandbox "choose your spell" style, but the ability to change haste/crit/spellpower variables mid-rotation as in the old sandbox would be a lot of work. Presently if I want to test out a spell sequence/timing tweak I just manually edit one of the rotation formulators.

Ultimately the sandbox really confused a bunch of people and was totally ignored by the rest, so putting it back in is not a high priority, but I will consider it.

Regarding elemental trinkets:

I wanted to find out how much worse HC Phylactery is compared to HC CTS, considering that I can pick up the former for 1 dkp and have to go nearly all-in on the latter.

So I did some ZAP maths, swapping out HC Reign for HC phylactery / CTS. Surprisingly, 172 crit gave me 100 more dps than 184 haste at 3 min fight length, and still 6 more dps at 5 min fight length (mostly owed to relative bloodlust uptime)

In ZAP vhe value of the proc is listed as 232 SP average proc for both Phylactery and CTS; if the proc were actually the same then phylactery would be better than CTS on most fights. But is it really the same? Zap seems to calculate the benefit from the trinket as 232 SP, regardless of fight length.

I decided to evaluate their actual SP value by looking at the best/worst/average case [Disclaimer: mad napkin maths ahead]

For me one full CTS cycle (average time to proc + time for ICD to wear off) is 55,5 sec, one phylatery cycle is 105sec.

Now what's the theoretical value if the boss dies just as the first (or any) full trinket cycle is completed?

Phylactery:
up 20s / 105s = 231 sp (19,1% uptime)

CTS:
up 15s / 55,5s = 232sp (27% uptime)


What's the worst case (trinket procs just as the mob dies) for the trinkets in the 5-6 minute fight range ?

Phylactery:
3* 105 + 5 = 320s fight; uptime 60s
-> 18,75% uptime, 226 sp on average

CTS:
6* 55,5 + 10 = 343s fight, uptime 90s
-> 26,2% uptime, 226sp on average

What's the best case (mob dies just as trinket proc ends) for the trinket in the 5-6 minute fight range ?

Phylactery:
320 + 20 = 340s fight, uptime 80s
-> 23,5 % uptime, 284sp on average

CTS:
343 + 15 = 358s fight, uptime 105s
-> 29,3 % uptime, 253sp on average

What's the "average" case (mob dies when trinket is half way through the time between the end of the proc and the next proc)?

Phylactery:
340 + ((105-20)/2) = 382s fight, uptime 80s
-> 20,9% uptime, 253sp on average

CTS:

358 + ((55,5-15)/2) = 378s fight, uptime 105s
-> 27,7% uptime, 239sp on average




Conclusion:

For both trinkets the worst case is actually just slightly worse than the value of a full cycle. However, when it comes to fight length you can get way more lucky with phylactery (284sp on average); even getting extremely lucky with CTS only nets you the same 253sp as average luck with the phylactery.

Coupled with the seemingly better static bonus (at least for non-super long fights) phylactery is apparently the better trinket to pair with an existing DFO at the highest gear level. Phylactery/CTS only beats out Phylatery/DFO by roughly 4-8 SP.



Edit: Doing the same maths for DFO you get:

Theoretical value:

The proc translates into an average 666sp, up 20s / 55,5s = 240sp (36 % uptime)

Worst case:

343s fight, uptime 120s
-> 35% uptime, 233sp on average

Best case:

343 + 20 = 363s fight, uptime 140s
-> 38,6% uptime, 257sp on average

Average case:

363 + ((55,5-20)/2) = 381s fight, uptime 140s
-> 36,7% uptime, 245sp on average

Unsurprisingly DFO's best/worst/average cases are very similiar to CTS's, on average the DFO proc is between 4 and 8 sp better, while the -14 static haste equals a deficit of about 12 sp.

So if you have DFO and Phylactery already, you might just let someone else have the CTS since it's only a very marginal upgrade at the highest gear level.

Hi,

I'm new using your spreadsheet and I have a question about the "Snapshot" tab.

Why did the Flame Shock has a cast time? And why is it longer than 1 second?
And for the "Elemental Mastery" cast sequence, we don't see when the Elemental Mastery is casted (it seems it's after the la Lava Burst) and no spell is instant (cast time equal to 0).

Thanks for your work on Spreadsheet and Blog, it's very useful.

Theli

The spells we think of as "instant cast" do actually have what we call an effective cast time. That is, though the effect may occur the moment we press the button, we can't press any other buttons for a period afterwards known as the Global Cooldown. The Global Cooldown lasts 1.5s with no haste and is reduced to a minimum value of 1 second at 50% haste, so Flame Shock and the instant-cast spell triggered by EM both have an effective cast time corresponding to this value - which is why you see nothing in the Snapshot page with a "0" cast time, and why Flame Shock's effective cast time would be over a second.

EM itself is actually genuinely instantaneous and doesn't trigger a global cooldown, and neither does Wind Shear, hence there's no need to write an "Elemental Mastery" spell into the data.

By default EM is used on the first Lightning Bolt after FS/LvB, but this can be changed in the Advanced Options. Before I wrote in that option there was a little blue line on the Snapshot telling you when EM was used but with a dynamic use timer that's not possible. Best I could do would be to have a little text note pop next to the place EM's used.

My personal in fight default is to use haste effects with Flame Shock whenever possible. I.e. I have a keybind for Elemental Mastery / Flame Shock, and follow that with a LvB while running in, and then have another keybind for Berserking / Lightning Bolt after EM expired. (Will save Berserking for later on fights where we BL a few seconds in, though)

But of course waiting 14 seconds to use EM midfight just so you can do it right before FS makes no sense, especially since it will still be affecting the FS you cast 14 seconds later even if used right away.

Either way, switching default from LB to FS increased DPS by 30.

I fully support this conclusion, although I have reached it through a slightly different means: I used the ZAP rotations to compare DFO + Phyl vs DFO + CTS and they ended up to be same dps at top-end gear levels.

The reason is the haste of CTS is worth considerably less as we are well over the haste softcap already with full 277 gear and DFO. This brings the CTS trinket on-par with Phylactery (stats+proc). So still an excellent trinket to have, but not as good as pure theorycrafting would suggest when taking isolated DEP values into account.

Wouldn't DFO + CTS + regemming Potent on yellows be better if Haste at high levels are sub-par?

I mean CTS gives 184 Haste vs Phylactery 172 Crit so by using DFO + CTS + regemming, you're gaining more stats in the end (assuming all the procs are roughly equal).

I'm actually curious myself since I'm sitting at an ungodly amount of Haste (1597 with Bizuri up) and is wondering what the best course of action right now is. With Zerking/EM/Glove up, my FS is messing me up like whoa and I feel LB clipping during Lust. The resulting DPS gain also seems minimal.

I think you will find that the dps difference between those configurations is very small, so if you feel it i proves your gameplay it is worth a shot I guess. Note however that the value of haste compared to crit decreases very gradually so if you are just over softcap hsste is still a lot better stat then crit. Personally I would stick with haste