WotLK DPS spreadsheet - Page 13

sapan · 10/11/08, 6:33 AM

Nice thread and spreadsheet you have here..
I was working on a level 70 spec on this spreadsheet and noticed that when I put 1 point in "go for the throat" the dps goes up by 216 and when I put the second point it goes up by 5.
Something should be wrong I guess right?

Shandara · 10/11/08, 6:38 AM

Originally Posted by sapan

Nice thread and spreadsheet you have here..
I was working on a level 70 spec on this spreadsheet and noticed that when I put 1 point in "go for the throat" the dps goes up by 216 and when I put the second point it goes up by 5.
Something should be wrong I guess right?

If your pet has enough focus to sustain his abilities with 1 point (or very near enough) then the 2nd point will of course have reduced effect.

And that's what happens with 2/2 GftT (if you have 2/2 Bestial Discipline).

Indora · 10/11/08, 8:03 AM

You should put 2 points in it though, as fokus regen via crit isn't constant.

KraxisSingular · 10/11/08, 8:27 AM

Originally Posted by Indora

You should put 2 points in it though, as fokus regen via crit isn't constant.

Not as important any longer. Our pets don't regenerate focus in big lumps every 4 seconds, but rather a small amount all the time, leading to more steady focus and less of a chance of our pet running out even in a critdrought (running out of focus, then waiting for the big regen, now it will take longer to run out of focus due to the constant regen).

Miya Mirage · 10/11/08, 12:12 PM

I found something I just cannot believe, maybe someone can explain to me why it works out to the results I get.

For my gear on the calculations tab Ashtongue Talisman of Swiftness averages out to 139,9 AP (50,87% uptime) and Blackened Naaru Sliver averages out at 137,88 AP (38,28% uptime). Now when I keep all my gear the same and just switch between them on one trinket slot I get 45dps (about 30 hunter and 15 pet) more using the Talisman. I just can't understand how that comes to be when they average out to about the same and the Sliver even has a bit of haste on it for the ever so slightly more Auto Shot damage.
Can someone enlighten me as to why this happens? Is it just the difference in uptime?

Sebudai · 10/11/08, 12:42 PM

I've never trusted the way either spreadsheet models the Sliver. It'll rank high on the Gear Planner tab, but actually equipping it to an empty trinket slot provides only a marginal dps increase.

Indora · 10/11/08, 2:42 PM

That's because the spreadsheet doesn't factor in the trinket's procc. (However, the Gear Planner does.)
It should be added to the "On Proc +AP Effects" (Calculations!D58)

Shandara · 10/11/08, 5:34 PM

Originally Posted by Indora

That's because the spreadsheet doesn't factor in the trinket's procc. (However, the Gear Planner does.)
It should be added to the "On Proc +AP Effects" (Calculations!D58)

Good catch. Fixed next release. I recently redid practically every trinket's stats behind the scenes for the Gear Planner, so there might be others that work correctly on the gear planner but fail to have effect in the actual calculations. I'll give it a check myself, but if you find any more, let me know.

Devastor · 10/11/08, 11:08 PM

I noticed some things, eventually bugs - please check or explain me ;-D

First thing:

Originally Posted by KraxisSingular

Not as important any longer. Our pets don't regenerate focus in big lumps every 4 seconds, but rather a small amount all the time, leading to more steady focus and less of a chance of our pet running out even in a critdrought (running out of focus, then waiting for the big regen, now it will take longer to run out of focus due to the constant regen).

I just tested some things myself.
Lvl70, 4x T6 as a gear-indicator, 41% crit, lvl70 Buffs, 50-11 BM Specced (no cobra strikes)

1 point in GftT adds 160DPS
2nd point adds: exactly 0DPS

Can this really be true? A cat was my Pet in this test. So aren't there Pets with better Focusdump abilities?
This really makes me wonder why 2/2 GftT adds absolutely nothing for me.

Second observation
Using Lvl70 Buffs Preset:
Regarding "Time to OOM",
Viper usage is "Viper - Regen"

with 0/2 Invigoration: 368sec / 3429 DPS
with 1/2 Invigoration: 759sec / 3447 DPS

Maybe i missunderstand something, but seems unlikely that Time to OOM nearly doubles with 1/3 Invigoration, though DPS increases only by 18DPS?

Notice:
Using "None" Buffs Preset, Time to OOM:
0/2 = 77sec
1/2 = 82sec
This looks "true" - what makes me wonder is the EXTREEEM increase while Lvl70-buffed.

Third thing:
First off, I have absolutely no clue about internal calculations etc...

What i just saw while looking at the "Pet Calculations" sheet:
In the smaller/left Rabid block, it states Proc Duration = 30sec and Uptime = 92%
In the bigger block to the right, it states Proc Duration = 20sec (which represents actual ingame duration)

30sec Uptime just wrong? Or what is it i oversee?

Thanks in advance

Shandara · 10/12/08, 2:56 AM

@Devastator

The Rabid thing is a 'bug', or rather it's a relic of some old calculations. The big block on the right side of the page are the actual calculations, the small one on the left isn't used, except to see if Rabid is used at all. It was still stuck at the old duration. Should be fixed next release.

GftT doesn't add anything beyond the first point (if you have 2/2 Bestial Discipline) _on paper_. With high enough crit rate your pet should never run out of focus _on paper_. Unfortunately reality isn't so reliable. You will have streaks of crits or dearths of crits and your pet will run out of focus. So in general 2/2 GftT is recommended. But on paper the second point doesn't add DPS. Because your pet can't spend taht amount of focus fast enough.

With invigoration it's really easy to explain why the time to OOM doubles. Say you are losing 20 mana per second. Now you get 1/2 Invigoration which adds 10 mana per second. You are now losing half the mana from before and will last _double_ as long. It's not a linear scale.

Chul · 10/12/08, 4:51 AM

Remember that Claw is actually firing faster than the 1.5s GCD, so you need more than the theoretical value (until they change it

)

Dalinoth · 10/12/08, 12:48 PM

Hi. I'm new to Elitist Jerks and need a little help please. This is my first time using the spreadsheet whatsoever (thanks Shandara - great work!) and since I'm neither a math wiz nor an excel wiz, I'm having some difficulties. Is anyone available over vent or skype to spend 5 minutes helping me sort things out? Thanks in advance.

Shandara · 10/13/08, 8:18 AM

Update:

- Disabling averages now also turns off AotV calculations
- Blackened Naaru Sliver's AP is now correctly used in the calculations
- Scorpid Poison should now be correct in its DPS calculations. Scaling changed to match tested results.
- added Minor Glyphs (no functional implementations since none of them affect DPS)
- Misc fixes to various trinkets
- You can now select where to export to, on the Settings and Results tab

messcaline · 10/13/08, 9:03 AM

Unfortunatelly never seen this spreadsheet working (Office 2008 for Mac) so maybe little OT:
Anyone tried maybe to rewrite this document in google docs? Platform and software independent. Would be great to have such one.

Shandara · 10/13/08, 9:26 AM

Originally Posted by messcaline

Unfortunatelly never seen this spreadsheet working (Office 2008 for Mac) so maybe little OT:
Anyone tried maybe to rewrite this document in google docs? Platform and software independent. Would be great to have such one.

They seem to have a size restriction of 1mb for Spreadsheets when I last checked and I couldn't convert the existing sheet. I'm not sure whether the 1mb restriction is only for importing or if it's overall but it wouldn't be enough unless I'd severely trim some functions from the sheet.

Not to mention a web-based application is probably quite a bit slower.

veyn · 10/13/08, 9:57 AM

The Cobra Reflexes calculation (Pet Calculations!F9) references wrong cell for tenacity pets (Pet!K3 instead of Pet!M3).
Also, in IAotH calculation, shouldn't formula for shots in initial proc (Calculations!N87) be (IAotHDuration-IAotHBaseSpeed)/IAotHShotSpeed + 1 instead of (IAotHDuration-IAotHBaseSpeed)/IAotHBaseSpeed + 1?

Miya Mirage · 10/13/08, 10:12 AM

Did you change something about Pet calculations in general? Comparing last version and this version my pet lost 50 dps on autoattack, 20 dps on claw and 20 dps on rake.

messcaline · 10/13/08, 10:13 AM

Unlucky me... Unfortunatelly Office 2008 for Mac has VBA removed :-(
Numbers from iWork of course doesn't have VBA ;-) so as far as I found it's possible to use Automator or ActionScripts instead VBA for Numbers...
Do you have some IM? Maybe I'll be able to simulate something like this xls spreadsheet...

Shandara · 10/13/08, 10:13 AM

Originally Posted by veyn

The Cobra Reflexes calculation (Pet Calculations!F9) references wrong cell for tenacity pets (Pet!K3 instead of Pet!M3).
Also, in IAotH calculation, shouldn't formula for shots in initial proc (Calculations!N87) be (IAotHDuration-IAotHBaseSpeed)/IAotHShotSpeed + 1 instead of (IAotHDuration-IAotHBaseSpeed)/IAotHBaseSpeed + 1?

Yes, it shouldn't count the initial shot that procc'ed it but that's fired at the pre-hasted speed and the amount of further shots in the initial proc is determined by the hasted speed.

I won't pretend to understand the calculations that follow though.

@Miya:
I did toy with the buff profiles a bit because I worked on saving profiles so it might have a different buff set than the last version. So if that's not it might be something unintended.
EDIT: After checking V73f and V74 and equalizing the buff profiles my pets did nearly the same dps in both (I made some fixes to some trinkets I had equipped in my gear set which made up for the differences).

Miya Mirage · 10/13/08, 10:29 AM

Ah yes, I thought when exporting, it does include the Buff Profiles I made for our raid setup. Should've thought of it myself.

A second thing that did have a small influence on hunter dps actually was the fight length and the viper settings. All things I thought to be included in the export file. Now that I know, I shouldn't bring up things like that again.

Shandara · 10/13/08, 10:38 AM

Originally Posted by Miya Mirage

Ah yes, I thought when exporting, it does include the Buff Profiles I made for our raid setup. Should've thought of it myself.

A second thing that did have a small influence on hunter dps actually was the fight length and the viper settings. All things I thought to be included in the export file. Now that I know, I shouldn't bring up things like that again.

It should export the buff profiles properly, but it usually breaks when I add/remove a buff from the available ones since I don't provide conversion from version to version if I do change them.

As for the AotV settings, I hadn't added them to the exporting yet. The sheet is quite beta in the user-friendly department, sorry. Once things settle down in WotLK I'll probably find some time (if I'm not busy leveling) to organise these things.

EDIT:
Things I'm currently not saving in profiles:

AotV settings (Shot Rotation tab)
Default Gem Choices (Settings and Results)
Gathering Skill (Settings and Results)
Control Settings (Settings and Results) - exception Custom Target level/armor
Filename to Export to (obviously)
Player name/realm/server set
Target selected
Gear Planner selections
Custom Gear
Sniper Training range selector (Talents tab)
Boss fight length settings (Shot Rotation)
Boss HP % selector
Latency
Misc settings on the calculations page (Lock&Load modifiers, etc..)
Pet GCD (Pet tab)
Pet Level/Happiness

I think that's all of it. Most of these things are pretty much static in between comparisons.

Miya Mirage · 10/13/08, 10:45 AM

That's fine, I'm just fighting a flu so I'm not 100% concentrated and things like these slip by.

The thing with the AotV Setting was that switching from Viper-Regen to Viper-oom led to 1 AP more being transferred to the pet in the pet calculations tab somehow and it changed the numbers slightly (less than 1 dps but I noticed it when comparing after equalizing the buffs).

I for one am happy with the functionality in the spreadsheet already, so don't worry about small things like that. 99% transferabilty from version to version is superb enough for me! I'll happily check the Buffs and Shot Rotation Tab from now on (if I get rid of that stupid flu at least).

Chul · 10/13/08, 6:18 PM

If you have 2 or more points in Longevity, Rake will only tick twice, not three times.

Shandara · 10/14/08, 6:22 AM

Originally Posted by Chul

If you have 2 or more points in Longevity, Rake will only tick twice, not three times.

Cheers. I had the day off so I spent some time reorganising the UI and moving stuff around. I've moved all the 'options' for various things (spells, range checks, etc..) to a separate tab and only left the really specific toggles where they were (e.g. Latency on the Shot Rotation tab).

I've also cleaned up the Settings and Results tab and moved most of the informational windows there. No doubt this will break stuff (although I didn't get many errors myself and I fixed those), so let me know.

It should also save ALL the custom settings with each profile now. There's an option to override them on the 'Settings' tab in which case it will use predetermined defaults for all of them. The defaults are also used if you don't specify a custom setting for an option.

- Cobra Reflexes now references the correct cell for Tenacity Pets
- Small fix to IAotH
- Redesigned UI, moved all custom settings to a separate tab. Cleaned up the Settings and Results tab
- All custom settings will now save with the profile (and export); existing exports won't load correctly (hopefully the last time I have to make changes)
- Rake will only tick 2 times if you have 2 or more talent points in Longevity
- Fixed a bug with the Gear planner involving +hit and trinkets

Stealth update:

- Rapid Killing 1/2 leads to a Rapid Fire downtime of 127,5 seconds with Readiness; 0/2 leads to a 157,5 cooldown
- You can now name talent specs and these names are exported/imported properly

veyn · 10/14/08, 8:01 AM

Small correction regarding Readiness and Rapid Killing. When you put 1 point into Rapid Killing, you consider Rapid Fire cd to be equal to 180 sec - which (I assume) means you reset Rapid Fire cd immediately when Readiness becomes available. If you would wait for RF cd to end naturally instead, and you'd have 2 RF's every (240+15) sec, which leads to greater uptime (effective cd of 127.5 sec).

edit: In fact, even if you spend 0 points in Rapid Killing, waiting for natural cd of RF leads to effective cd of 157.5 sec, as opposed to resetting RF cd with Readiness immediately (180 sec cd).