WotLK DPS spreadsheet - Page 67

Jerem · 02/20/09, 8:41 AM

You don't meet the requirements for your metagem anymore, without the Deadly Topaz.
Replace one of your Perfect green gems by an enchanted Tear, and you should see better numbers.

Ingmar · 02/20/09, 8:46 AM

Hmm I'm thinking it's the meta since you're not using the Valorous gloves / Grotesque Handgrips. You're also using arrows with Envoy of Mortality, not sure if that matters for the spreadsheet but it can't be correct.

Just as a tip, you might want to use less purple gems and just miss out on some socket bonusses.

edit: beaten

Shandara · 02/20/09, 8:51 AM

Also, have you hit the 'Calculate Attributes' button after changing your gear/spec?

OJey · 02/20/09, 9:50 AM

Originally Posted by Jerem

You don't meet the requirements for your metagem anymore, without the Deadly Topaz.
Replace one of your Perfect green gems by an enchanted Tear, and you should see better numbers.

aaaaaaargh ! thx a lot^^ i knew the sheet ist perfect ! the mistake had to be mine ...

i will use an vivid forest emerald to get to the hit cap, so my meta is active again and the Crippled Treads are the boots of choice.

Sapphique · 02/20/09, 4:04 PM

It appears that the spreadsheet isn't factoring in the increased miss chance of Spirit Strike, which seems (from others' and my testing) to use Spell Hit. Versus a level 83 boss, a level 80 pet would need 17% spell hit to never miss, so a hit-capped Hunter's Spirit Beast would miss with Spirit Strike 7% of the time.

It's not an enormous DPS decrease, and in fact the Spirit Beast's ranking in comparison to other Ferocity pets doesn't seem to change with this corrected, but I think it's worth fixing nonetheless.

Please let me know if my analysis is erroneous or if I missed something.

alarge · 02/20/09, 4:19 PM

Originally Posted by Sapphique

It appears that the spreadsheet isn't factoring in the increased miss chance of Spirit Strike, which seems (from others' and my testing) to use Spell Hit. Versus a level 83 boss, a level 80 pet would need 17% spell hit to never miss, so a hit-capped Hunter's Spirit Beast would miss with Spirit Strike 9% of the time.

It's not an enormous DPS decrease, and in fact the Spirit Beast's ranking in comparison to other Ferocity pets doesn't seem to change with this corrected, but I think it's worth fixing nonetheless.

Please let me know if my analysis is erroneous or if I missed something.

I can't speak to Spirit Strike directly. However, hit rating converts differently to physical and spell hit. At 263 hit rating, you have 263/26.23 = 10% spell hit (not 8%). So your pet should only miss 7% of the time.

Sapphique · 02/20/09, 4:49 PM

Originally Posted by alarge

I can't speak to Spirit Strike directly. However, hit rating converts differently to physical and spell hit. At 263 hit rating, you have 263/26.23 = 10% spell hit (not 8%). So your pet should only miss 7% of the time.

Good catch. I corrected my post accordingly.

I've tested in-game and confirmed that Spirit Strike misses when hit-capped. I can think of no other plausible explanation besides the ability using spell hit instead of physical hit.

Argg0 · 02/21/09, 3:05 AM

Any chance we get SP Buffs for pets? As some of them use SP (and Totem of Wrath could maybe change stuff).

Also, I'm not sure, but I think that Moonkin/SPriest/Totem can help with some pet hit (WS, Spirit Beast) and Crit.

Ikuzo · 02/22/09, 7:29 AM

There are a couple of errors in the spreadsheet for PVPers that use it.
Arcanum of the Savage Gladiator is listed on wowhead as 25 resilience, but the spreadsheet has only 20 resilience on it.

Also, Titan-forged Sabatons of Triumph has a yellow socket on it, but the spreadsheet does not reflect this.

Shandara · 02/22/09, 4:49 PM

In anticipation of Uldaur going onto the PTR soon (tm) and hopefully with a slew of hunter changes to go with it, I've just uplodaded a new version of my spreadsheet. It includes a lot of experimental stuff and changes, mostly cosmetic or user-friendliness and I would like to have some feedback on them (and bugs).

The previous version (v84) is available from a link on the first page.

- Spiced Wyrm Burger now +40 Crit Rating
- Fixed Furious Howl not stacking with food and other +AP buffs (i.e. not Bshout/BoM)
- Fixed Arcanum of the Savage Gladiator (now +25 resilience)
- Added yellow socket to titan-forged sabatons of triumph and corrected resilience on shoulder enchant to show up on the gear tab
- Changing pet family will not automatically fill in the family skill and focus dump
- Fixed Boar's Gore to work with Charge (you'll need to put charge in the pet rotation). Charge is assumed to be used every time it's up (not really realistic unless for multi-mob boss fights).
- Made a fix to the rotation test tab not always adjusting the dps on the shot rotation tab if the "Use Shot Rotation Test results" is used.
- You can now sort gems on the settings tab
- Added melee weapon enchants to the gear planner. You can select which enchant is used to compare items. Comparisons between 2-handers and dual-wield are not intended.
- Small fix to PAWN string
- Added setting to not interleave other shots with L&L Explosive Shots on the rotation test tab (exported Settings tab changes won't import for this version)
- Small fixes to base stats calculations and rounding. Also added an 'in-game tooltip' value to the Calculations tab for Hunter RAP to make it easier to verify.

Some notes:
* First of all, selecting different pet families should automatically fill in the correct skills now. Only exceptions are pets that have more than 2 usuable skills. I was considering just removing the whole pet priority queue but it provides a fallback whenever blizzard adds more abilities to pets. Currently the code just fills in the family skills (i.e. Gore for boars) and the focus dump (i.e. Bite/Claw/Smack).
* I've changed the Shot Rotation test tab to mimic actual real behaviour more closely when it comes to Lock&Load and Explosive Shot tick waiting. There's a new setting on the settings tab to toggle this.
* There's a filter option for gems now. You can toggle various things. You can set sorting order with a total of three priorities. You can sort on Color, Alphabetically, Agility, AP, Hit, Crit and Quality. Try it out. I'm not sure myself yet if it actually works how I want it to work.

alienangel · 02/22/09, 5:31 PM

Originally Posted by Shandara

* There's a filter option for gems now. You can toggle various things. You can set sorting order with a total of three priorities. You can sort on Color, Alphabetically, Agility, AP, Hit, Crit and Quality. Try it out. I'm not sure myself yet if it actually works how I want it to work.

The ideal sorting order option would IMHO be sorting by the DPS values worked out by the CalcAttribute button. So if that table works out +1agi is worth 1.5dps, while +1crit is worth 0.5 dps, a 10 agi gem would sort higher (15dps) than a 10 crit gem (5dps).

I'm generally not trying to maximize ap or hit or crit anything, I'm just trying to maximize DPS - so when selecting a gem I'd usually want to pick the one at the top of the list with the highest dps value, and if it's not the right color for the socket bonus, note down how much dps it yields then try the next highest gem on the list that is the right color and compare.

Shandara · 02/22/09, 6:10 PM

Originally Posted by alienangel

The ideal sorting order option would IMHO be sorting by the DPS values worked out by the CalcAttribute button. So if that table works out +1agi is worth 1.5dps, while +1crit is worth 0.5 dps, a 10 agi gem would sort higher (15dps) than a 10 crit gem (5dps).

I'm generally not trying to maximize ap or hit or crit anything, I'm just trying to maximize DPS - so when selecting a gem I'd usually want to pick the one at the top of the list with the highest dps value, and if it's not the right color for the socket bonus, note down how much dps it yields then try the next highest gem on the list that is the right color and compare.

Hmm, good idea.

One problem is that it's not really easy to see whether a gem will satisfy the socket bonus, since that would require generating individual lists for each gem socket on the gear tab.

Sylvand · 02/22/09, 7:16 PM

Alpha sorting for gems sorts them in descending order, in contrast to the initial state which is ascending alpha sort.

The sorting is a good start; I think I agree with alienangel on the calc attributes. However I think an alternative would be to add filtering options such as hiding TBC gems, uncommon gems and purple gems. With those options I think sorting by agility/AP and then hit would be pretty close to the ideal for most folks.

alienangel · 02/22/09, 9:05 PM

Originally Posted by Shandara

Hmm, good idea.

One problem is that it's not really easy to see whether a gem will satisfy the socket bonus, since that would require generating individual lists for each gem socket on the gear tab.

Well I don't think the sort needs to consider socket bonuses if that's what you're saying - just let the sort sort by DPS value of the gem's stats, and leave it up to the user to experiment with different colors to see if any of them can get a socket bonus of greater value than the extra dps from using a better gem.

trunghung · 02/23/09, 12:44 AM

Does anyone else notice a big difference between the estimate that the spreadsheet provides versus the actual dps. The spreadsheet says my dps should be 6400 dps, but the my H naxx dps is typically around 4500dps. I don't think I am missing too many raid buff. The damage by each of the shot are pretty close to what the spreadsheet describe but for some reason I just never been able to come with 500 dps of the theoretical dps from the spreadsheet. It used to be a lot closer in TBC when it comes to my actual dps and estimate.

Ufthak · 02/23/09, 5:38 AM

I must admit when I was playing around with DPS values for gems I would normally end up sorting by DPS & colour or by colour and DPS.

As an idea for the lazy, would it be possible to add a column to the end of the gear tab that would give the DPS for the socket if it was matched? That way you could get an quick estimate of whether it is worth matching the socket.

Shandara · 02/23/09, 8:04 AM

Update:

- Alpha sort for gem is now ascending
- Added 'DPS' sorting option for gems
- Sorting gems on Quality now works
- removed all TBC gems (excepting meta-gems)

Hirgux · 02/23/09, 9:59 AM

Seems to be bugged; I do not get any values at the DPS cells.
The german error says #BEZUG - what I think has to mean reference in this context.

Shandara · 02/23/09, 10:04 AM

Originally Posted by Hirgux

Seems to be bugged; I do not get any values at the DPS cells.
The german error says #BEZUG - what I think has to mean reference in this context.

I've re-uploaded a quick fix for this. Changing all the gemming messed up a reference to one of the meta-gems.

Sapphique · 02/23/09, 10:43 AM

Originally Posted by trunghung

Does anyone else notice a big difference between the estimate that the spreadsheet provides versus the actual dps. The spreadsheet says my dps should be 6400 dps, but the my H naxx dps is typically around 4500dps. I don't think I am missing too many raid buff. The damage by each of the shot are pretty close to what the spreadsheet describe but for some reason I just never been able to come with 500 dps of the theoretical dps from the spreadsheet. It used to be a lot closer in TBC when it comes to my actual dps and estimate.

I've found that relative DPS changes are fairly accurate with the sheet - that is, I can gain approximately x DPS by changing to Survival, or to this talent build, etc. However, I've found the DPS estimates across the board to be higher (in some cases significantly higher) than attainable in game, even with extremely generous latency settings.

Version 84a actually seems to highlight this further, as estimated DPS in all of my configurations went up even more. The gap between BM and SV also closed a little on the new spreadsheet (with my settings), which is surprising because the difference between BM and SV in-game has been greater than even the last version of the spreadsheet predicted.

I've been reviewing the calculations to see where the inaccuracy could be occurring, but so far everything I've looked at appears correct (as correct as we know about). From in-game testing, though, I still think Spirit Beast is being heavily overvalued.

Shandara · 02/23/09, 11:06 AM

Originally Posted by Sapphique

I've found that relative DPS changes are fairly accurate with the sheet - that is, I can gain approximately x DPS by changing to Survival, or to this talent build, etc. However, I've found the DPS estimates across the board to be higher (in some cases significantly higher) than attainable in game, even with extremely generous latency settings.

Version 84a actually seems to highlight this further, as estimated DPS in all of my configurations went up even more. The gap between BM and SV also closed a little on the new spreadsheet (with my settings), which is surprising because the difference between BM and SV in-game has been greater than even the last version of the spreadsheet predicted.

I've been reviewing the calculations to see where the inaccuracy could be occurring, but so far everything I've looked at appears correct (as correct as we know about). From in-game testing, though, I still think Spirit Beast is being heavily overvalued.

There are a few things that will affect the dps predicted:
a) Rotations are tighter in theory than in practice, this is especially important since latency is variable in-game but assumed static for the spreadsheet
b) The standard level 80 raid buff set might or might not be an accurate representation of the average buffs a 25-man raider would have (especially buffs with uptimes or limited ranges).
c) RNG affects crit rates for a non-trivial amount.
d) Some pets are most likely not yet modelled correctly.

That said I generally get results close to in-game for an unbuffed test vs a training dummy with my cat (the cat's unbuffed dps is particularly close to the predicted amount). My own dps varies depending on the crit rates achieved, naturally, but it is usually within a 100 dps of what has been predicted.

Sapphique · 02/23/09, 12:05 PM

Originally Posted by Shandara

There are a few things that will affect the dps predicted:
a) Rotations are tighter in theory than in practice, this is especially important since latency is variable in-game but assumed static for the spreadsheet
b) The standard level 80 raid buff set might or might not be an accurate representation of the average buffs a 25-man raider would have (especially buffs with uptimes or limited ranges).
c) RNG affects crit rates for a non-trivial amount.
d) Some pets are most likely not yet modelled correctly.

That said I generally get results close to in-game for an unbuffed test vs a training dummy with my cat (the cat's unbuffed dps is particularly close to the predicted amount). My own dps varies depending on the crit rates achieved, naturally, but it is usually within a 100 dps of what has been predicted.

I understand. My configured buff set is -exactly- correct for my guild A-team composition. I'm not one of the "take the stock settings and run with them" people. Every setting on my sheet is exactly as it is in game. I calibrate my latency setting based on actual performance data prior to testing. For better or worse, I spend more time working with this sheet than is probably healthy! As others have said, thank you for putting in the time to make the sheet as accurate as possible.

I'm also being reasonable about my assessment with respect to rotation tightness and making allowances accordingly. I'm still seeing BM too close to Survival on the spreadsheet in many cases, and I've found Survival to be the more accurate of the two specs with respect to live (with BM being overvalued slightly). It's not FAR off, but it is consistently off. I don't think there's any harm in questioning the spreadsheet's accuracy sometimes when my data doesn't agree with it. Another pair of eyes studying the calculations can never be harmful.

pheinle · 02/23/09, 1:15 PM

This spreadsheet doesn't work at all for me. Is it because I have a new iMac that uses OS 10.5 even though I have excel or what? None of the buttons seem to work at all or the macros.

Nagisamuro · 02/23/09, 1:37 PM

Originally Posted by pheinle

This spreadsheet doesn't work at all for me. Is it because I have a new iMac that uses OS 10.5 even though I have excel or what? None of the buttons seem to work at all or the macros.

As of the 2008 version, Office for Mac no longer includes VBA (the whole macro editor, debugger, language, etc.).

Microsoft to bring back Visual Basic in Office for Mac | Macworld

You can:
a) find an older version of Office of Mac (I think 2004 was the last to include VBA, there may be legal ways to acquire it today, I don't know).
b) run it on Windows w/ a copy of Office for Windows.
c) try your luck with OpenOffice.org and the instructions for getting the spreadsheet to work in that program, posted in this thread by a third party.

Sorry.

Nageya · 02/23/09, 2:37 PM

It seems that Darkmoon Card: Greatness no longer has a value associated with it in the latest beta spreadsheet.