stampy

This is exactly what I came here to take a poke at, and poke at it I will... hopefully it'll actually produce something, but I definitely can't guarantee.

And while prot is definitely not a DPS tree, ignoring it is quite stupid. A complete player is always better than one who isn't, and there are indeed places where a raid may have more tanks than necessary, and for much shorter periods of time than warrant a respec. No sense in standing around with a thumb up ones bum.

Plus, even if we follow your incredibly pro tip to respec, gemming is a question... the best setup of gems for DPS in a prot spec may be quite different from the other trees, or may find some nice synergy (i doubt it -- im guessing that prot will undervalue ap compared to dps specs). For the folks who may aspire to be as pro as you clearly are, and respec frequently for different purposes, a whole lot of different armor gets worn. For the pieces that may have a home in multiple specs, its good to know what gems to fit in them, depending on if you want to benefit one spec the most or get a balance.

 

dr_AllCOM3

Just found this Download OpenOffice.org
OpenOffice with VBA support.
That means, you can use almost all Excel makros with OO.
Direct links, install a language pack if you like.

 

Zadus

okay i have been searching all day to fix this problem and i simply cant find it. I keep trying to use the spreadsheet but it tells me my formatting is wrong or something. Can someone please point me in the right direction?

To be more specific alot of the drop down menus to pick different gear is basically well not dropping down.

 

Apate

Make sure you have enabled Macros. It might help if we knew what software you were using. (OO? Excel 98?)

See you, auntie.

 

Zadus

I'm using excel 98 i believe. Where would i go to enable the macros?

 

Rallik

One big problem with the spreadsheet atm is it doesn't seem to use or carry over the leftover rage in a cycle after HS is included anywhere; so stats that significantly increase rage generation and directly leads to more HSs are undervalued in comparison to those that have less of an influence. It also causes large jumps in DPS when the threshold is reached to allow one more or less HS per cycle. Removing the INT() from around the formulas in the HS column in the DPS sheet changes that to a more fluid value and shows +hit rating being about 10% better than the spreadsheet showed before. It may be a more accurate estimate overall.

Also, and this one seems much more difficult to model to me, execute spam atm is simply considered to be 1.5s cooldown and assumes as long as your RPS is over the cost of execute, you get one off every 1.5s. But, with slow/slow being a fair advantage in dps with both higher WW damage and more flurry uptime, and +hit being considered the weakest stat from a pure DPS perspective, it's quite possible that the choppier rage generation during execute range could lead to plenty of slack between a GCD being used and getting the hit or 2 needed to execute for the next. I currently just swap to Slow/fast to even out that out a bit, but i'd be very interested if anyone has any idea of how to model the actual percentage of times you reach 15 rage in a 1.5s window, and then the average slack afterwards for those GCDs where you don't beat the 1.5s limit in order to give a more accurate average cooldown time. If a formula for that is possible, which I vaguely think it would be from the probability course I slept through, then execute spam would be muuuch more accurately modeled and there could even be a swapped in weapons part of the Gear section to show which give the best execute DPS too. Considering it's 20% of the fight, it seems like the current formula oversimplifies the reality.

Lastly, the rogue DPS spreadsheet has an interesting Upgrade macro to compare swapping in each piece of gear and then tell you which give the largest dps upgrades. I'm not very experienced with Excel macros, so I wasn't sure how to attempt it, but it's a very cool feature that would be nice to see added.

 

Machinator

I have been trying to figure out a way to model weapon speeds properly forever. It may be possible but its beyond my knowledge of statistics. Pretty much you need to find out how many gaps of 1.5s between two weapons, and then somehow fit misses and haste/flurry in it. Hell if we could find a way to model streaks a lot of other things like >100 rage, less than 30 rage for a BT, ect. Might be easier to just make a simulator for that though.

"Information is ammunition."

 

Gruntle

I guess you're using the 9.1.6 version? In the test version the number of HS are computed over the full fight length and the excess rage is added over all of the fight (maybe not for execute phase though). The number of HS over the entire fight is then calculated from this rage. This will be a fair representation of the theoretical number of HS you can do. In reality, you will do even fewer. I agree that the implementation in the previous spreadsheet may have underestimated the HS damage since you never used the extra rage, but in the test version you do. Download that version and see what you think of it. The number is still rounded to integer, but keeping it a float will only add a maximum of 0.5 HS (to a number 50-100 or so).

Ofc, it simplifies reality, it's a model . I agree that this issue is real, and it's the reason you should always use fast weapons (both MH and OH) for maximum damage during execute phase. Furthermore, the same effect can actually also ruin your BT/WW cycle during the rest of the fight if you're unlucky. With a slow/slow setup and lowish +hit (i.e. 9-11% or so), you can get miss streaks that are so long that you will not have the rage to maintain the cycle, and lose a lot of dps compared with the model. With great gear, you'll probably get enough rage anyway, but if you're just starting 25-mans it might be better to stay away from that particular setup.

I don't think you can model this in a closed form, you'll have to run simulations to find it. I doubt you can make those simulations inside the sheet (but sure, I'm not an expert on spreadsheets so maybe it is). One could possible run some simulations with some predefined values for weapon dps AP, crit, hit, haste and so on, something like 20 setups or so, and then make a table with the 15/12 rage per 1.5 sec "uptime" for these setups. The spreadsheet could then use the "uptime" to calculate execute damage from the setup nearest to the one in the sheet.

That would be really nice.

 

dr_AllCOM3

Please test this:
1.0beta1
You'll figure out by yourself, what isn't ready yet.

 

Zasm

Recently on TankSpot.com in the theories and guides section there has been a recent topic come to light upon Protection Spec PvE DPS (0/20/41). It has been known for a while now that Prot DPS has been around since the Devastate talent was introduced but I haven't located any true math craft sessions for it regarding the actual calculations.

Using the warrior dps sheet it's fine for entering in gear for Arms and Fury specs. Is there some way to add in the modified changes to support a prot dps spec such as 0/20/41. Aside from the talent data being added into the sheet the proper rotation would have to be added too. The rotation would consist of a 1 Whirlwind to 6 Devastates, a 1:6 ratio. Devastate does 50% listed weapon damage and an added 35 damage per application of sunder armor currently on the target and costs 9 rage per use when talented for, I'm sure most of the people here are aware of this already...

The reason why It might be a decent idea to add is for the warriors who like to do the off tank role yet stay in to dps on a said boss instead of sitting out. I know on our last few Teron Gorefiend attempts I let another warrior tank him since he has yet to experience the encounter while we were letting others get familiar with the ghost controls. To sum it all up as my full normal tank spec in my DW DPS gear I was pulling a consistent 850 to 890 dps without being grouped with a ret paladin, shaman, hunter, imp bs warrior, or feral druid. This was wasn't with the 1:6 ratio rotation but just going at it with devastate , it wasn't until after the raid that i went into deeper thought on the topic of prot warrior dps.

Summary:
- Possible to add a 0/20/41 Spec to the Sheet?
- Possible to add in a 1:6, WW to Devastate rotation?
- Theory craft on protection spec dps

 

Bronwyn

I'm using 09.16 at the moment and came across something I can't quite explain.

With my setup I get 1487 DPS using the S2 Gladiator 2.6 sword in my offhand. However, just switching to S2 Glad 1.5 OH makes my DPS jump to 1555 DPS. The reason is not rage generation (17.6 vs. 18.1 / s) ... It's simply the OH DPS - and why would the fast OH give me so much more actual DPS?

(Swiftsteel Bludgeon would be 1666 DPS) The difference of almost 100 DPS is unexplainable to me ...

You should have a link to my gear in my signature.

 

Gruntle

Could you try the same setup in the test spreadsheet (i.e. not the beta above, but the version before, posted a couple of pages back). I think it's the same old problem that's been in the sheet all along, that extra rage is just enough to give you one more HS during one cycle (you can check that by looking at the number of HS you get for a certain cycle, and compare the different weapons). The test sheet changes this to calculate the number of available HS over the full fight. Why you get more rage using the fast s2 offhand I can't explain (maybe it's how UBW is modeled).

To drAllcom, I took a look at the beta sheet, but as it's no longer at all compatible with OO, I can't really do anything with it. The 2.3 (go-OO version) can't run the scripts in that sheet. Not sure if it's supposed to.

 

dr_AllCOM3

OO doesn't like Useforms, sorry. Tell me a way to convert them and I'll do it.

 

Kwatah

I've been using this spreadsheet for a while and it has definitely helped me. Thank you.

I am curious though - I've read that using a MH and OH with the same speed grants you 4 flurried attacks. Is this (a) true (b) factored in to the spreadsheet?

 

Gruntle

a) Maybe, you could try to revive the thread by Disquette about flurry mechanics. The testing reported in that thread didn't show any conclusive evidence for the matched speed extra attack. If anything, the testing showed that we don't really understand squat about how flurry works. Well, we do understand the theoretical principle of it, but not how it's coded (i.e. are MH and OH counted separately, when does flurry go off in reality and so on).

b) No. The flurry formula used in the spreadsheet is the one hypothesised by Caesar in this thread a couple of pages back (at least I think dr_Allcom3 changed flurry to that formula).

dr_Allcom3, thanks for making the new version, a pity that it can't be made to work in OO. I know to little about scripting in excel and OO to be of any help in making it work. I will try to make the version before this beta into a fully functional OO version, perhaps you could put a link to it from the first page of the thread when I'm done.

 

dr_AllCOM3

Maybe if I shift some cells around it'll do. Without a nice GUI of course.
Edit: I guess, marking things with a color and locking/hiding everything else will do.

 

Amyannirving

I have a question about the way the spreadsheet models chance on hit enchants on the offhand during a whirlwind. Is this modeled into the spreadsheet calculations, because it could have an effect on the PPM rates of things like mongoose on the offhand, which were previously 1PPM but now could be more with WWs using the offhand weapon too. If so, would that make this enchant better than potency for a slow offhand? The spreadsheet says no according to my setup but that's why I wonder about the WW offhand proc issue.

Also, I noticed that the stats given on the mongoose enchant include the agility boost but not the attack speed increase. What kind of values should I be inputting into the haste portion of the enchant to make it truly accurate?

 

dr_AllCOM3

1.0beta2
Set2 is what it would look like for OO users. At least usable I think. There are no dropdown menus, it's just a preview.
I guess the whole sheet is correct and up-to-date now, since noone gave feedback.

 

Graul

If you always use haste potions (not everyone does) as well as have Bloodlust at 20%, there isn't any real reason to use anything faster than 2.6 speed for Executes. That's not even counting Flurry or if you have it, DST.

 

Malrix

dr_AllCOM3,

I'm trying to test the new spreadsheets but I got to say I'm at a loss for how to use it. I'm using Excel 2003 and the sheet appears to be very segmented. You mentioned that OO doesn't have dropdown menus, so I'm assuming that implies that Excel should show some but I don't see any. Is it just formulas that you want us to look at in this version?

 

dr_AllCOM3

Please be more specific.



1.0beta3
Some more stuff, mostly UI.

 

rayijin

On our first teron kill, I did 1083 dps as 10/5/46:

Wow Web Stats

I didn't have any haste potions to use, but did have an enhancement shaman and a fury warrior (who I barely outdps'd -.-) giving me improved battle shout. I've got around 35% crit / 1700 ap / 10% hit unbuffed, with a s2 mainhand, fury offhand. I also have a DST. I used recklessness + bulwark of ancient kings during an early bloodlust.

Execute seemed like a waste of rage - I used it 6 times in the fight and it only averaged out to about 1k damage each. Most of those executes were probably 30-rage ones, because my slow MH/OH always hit simultaneously. Devastate averaged 390 for 10 rage, whirlwind averaged 1k for 22 rage. Either way, I believe execute is only worthwhile to use without flurry/bloodlust/fast weps if you hit that magic 12-21 rage number.

Which brings us back to the topic of the thread: A spreadsheet option for modeling protection DPS would be excellent, as would a better/easier way to figure out execute DPS with slow and fast weapons. Show prot wars some love

 

• LodeRunner

Hey dr_AllCOM3,
Again, thank you for your continued work on this spreadsheet. However I can no longer use it, as it doesn't want to work with Office 2004 for Mac. When I press "Excel Start" it opens a new gray screen and just hangs there, thinking about it. The last version I used successfully was 0.916. Is anyone else using Mac Office having issues running this?

 

Suesse

We also recently did Teron for the first time. As prot, I did 1125 dps on kill shot without recklessness and did 1.1k-1.3k dps on other attempts (this is without any BT/MH/Kael/Vashj gear). I agree it would be nice to have a good prot model. I've been evaluating gear by picking out a 17/44/0 spec and putting in my gear to get SEP. I know it's probably not totally accurate. I assume agility and crit are rated too high (due largely to flurry).

You bring up a good question about execute. I've always just spammed it during sub 20%, but I think it's worth evaluating.

I don't think dr_AllCOM3 is interested in modeling prot. I figure when he's done with his current spreadsheet one of us can add prot modeling. Hopefully it'll be someone besides me since I really don't understand all this advanced excel stuff, but I'd be willing to figure it out if it came to that.

 

dr_AllCOM3

Haven't tested it on Office 2003 yet.