The values in the first post are now significantly different then any individual test. With 1 crit rating being exactly equal to 1 point of Strength.
I'm not sure how much the disparate ratios will stand up to a straight average like that, it seems to skew the ratio significantly. It has a tendency to make generic the contribution across the small range of values, which will likely have less then desirable results when multiplied out to the hundreds.
Edit: Ahh, Pater's numbers had them as equal. But Pater's post was more along the line of specific valuations at different existing levels of gear, so I'm not sure how valid that is overall. Basically, things like Disquette's 1.51 crit rating/AP is very different from saying 2/AP
The values in the first post are now significantly different then any individual test. With 1 crit rating being exactly equal to 1 point of Strength.
Pater's test showed 1 crit roughly equal to 1 strength. Tornhoof's model showed 1 crit roughly equal to 2 AP (100 AP = 21.5 DPS increase, 50 crit = 21 DPS increase). Disquette's sim was the outlier, and he told me to disregard the numbers from his sim as he knows it isn't modeling flurry correctly, so I did and we're left with Pater and Tornhoof's work -- which came up with very similar conclusions. So that's what we've got in the first post now.
First, I'm not familiar with Pawn (I don't use it since I plan out gear upgrades ahead of time), so would someone tell me if it's possible to add haste rating to this string?
(Malan, looks like the formatting for that part of the post got screwed up when I sent it to you; I'd probably take Disquette's out for now as well, since Pater's and Tornhoof's results are overlapping so nicely.)
And secondly, has anyone modeled the effect of "ignores X armor"? Obviously, I would expect this to vary between mobs, and based on debuffs present (sunder, faerie fire, CoR, etc.) but there is probably some work done on this. Maybe rogues or DPS warriors have done something here?
And a statement --
Although not particularly melee-friendly, Leotheras was a really fun fight. And I really like having a warrior specced Blood Frenzy -- forget Fury as a DPS tree
I'm not familiar with Pawn (I don't use it since I plan out gear upgrades ahead of time), so would someone tell me if it's possible to add haste rating to this string?
My numbers there are about the same as in the first post, with slightly less weight given to Kings. He has DPS values for weapons, but not speed or more importantly, top end, so I zeroed it out for now.
For the Thottbot thing, make sure you go to the link, modify/whatever, then hit save. And then when you search, check "Score top 50", but you need to put a level range or something on it. I was getting level 60 caster cloaks until I put in 68-70 for the level.
Edit: The cloak was because it thought it was a weapon for some wierd reason, and I put a high weight on weapons we coudl actually use. Just hitting item type and the top 50 box seems to work.
And a statement --
Although not particularly melee-friendly, Leotheras was a really fun fight. And I really like having a warrior specced Blood Frenzy -- forget Fury as a DPS tree
Leo isn't too bad for melee, you can go balls to the wall nuts during demon phase if your tank is worth his shit. We rotate in shaman during demon phase to bloodlust the melee group on every demon transition so that we can max the DPS on him. I also start running back in to the tanking area as soon as 2 seconds are left on whirlwind so I arrive just as he stops.
Edit - fixed the formatting issue on the first post
Yeah, we were able to go nuts during demon phase as well, although we didn't have Heroisms rotated in -- we saved those for 15%. Once our lock's Curse of Doom goes off in demon form (he had it going off within 3 seconds every time unless it was resisted), he's got aggro solidly for the duration. DPS still wasn't really a problem once we got people to not be stupid about whirlwind;, on the first kill we only had two (stupid) deaths before 15%, so DPS was fine. At 15% we did the kind of freak out and zerg him down, I think next time I'll probably be more cautious, in terms of avoiding the WW.
On that subject, I threw in all the best of slot mail pieces I could find into excel and calculated the difference between non-set pieces and Skyshatter including the 4 piece bonus. What do you guys think? If you take the formula at face value, it's better to get the non-set pieces and forget about Skyshatter. However, when I plotted it out like this, I noticed that the non-set pieces, being obviously hunter gear (they even look like hunter gear from what I gather) they have very little +hit. Now I know +hit has a relatively low value, and I've proven earlier in this thread that dropping really low isn't bad as long as you replace it with other stats. On the plus side though, going with non-set means more agility vs. crit rating, which means you will gain dodge too. Also, the non-set pieces are easier to get as they are sooner in progression and you would have to compete with less people (hunters only vs. warriors and other shaman specs)
So what do you all think? Is it better to go with the non-set? Or is there some reason to try and get the Skyshatter at least to the 4 piece (probably go with non-set legs still, as they are the most gain)?
BTW, I used epic gems in the gem slots for the calculations, so 10 str for red, 5str/5crit for yellow, and 5str for blue.
Edit: Sorry, forgot the conversion I used. Str=1.1, Agi=.97, CR=1, HR=.7, Atk Pwr=.5
Taking your numbers as listed, if we swap the t6 legs to bow stitched, we get non set = 592.57 and set = 590.44, the non set then needs to get a second blue gem to activate a meta, as you have only one blue slot in the listed section. (Optimal BT/Hyjal gear for other slots will also not give blue slots) As such we can then count off set as only 587.57, with set as 590.44.
The set gear also has quite a bit of stamina and looks way cooler than the off set. However, the set gear needs the set bonus to compete, gloves and hat are both worth picking up when you can imo, but chest should be left until after you have shoulders too, as its basically no upgrade from t5 without the 4 piece. Whatever you go for though, bow-stitch are the best leg option. Overall 'non set and set are very similar, non set is better before 4 piece, with 4 piece set is slightly better and looks way cooler, avoid t6 pants no matter what' seems about right.
Both are BoP leathrworking patterns in BT. With this and the belt patterns... aksdnalsdnalsdnasdnklasd I guess I have to drop Herbs and start grinding up Leatherworking.
Up until 1.5 weapon speed, the completed attack power equivalences are:
Haste Rating = 2.2
Strength = 2 (2.2 w/Kings)
Crit Rating = 2
Agility = 1.8 (2 w/Kings)
Hit Rating = 1.4
Attack Power = 1
Looking at these Values, i am not sure if you thought of Unleashed Rage in your calculations.
Since its permanent up we could count on the 10% extra AP, when we even count Blessing of Kings.
Strength would be 2,2 (2,42 w/Kings)
and
Attack Power = 1,1
Rest would stay the same.
I hope you could understand what i am trying to say. Maybe you already got it in your calculations and i missed it.
(Sorry for my poor English)
Yeah, we were able to go nuts during demon phase as well, although we didn't have Heroisms rotated in -- we saved those for 15%. Once our lock's Curse of Doom goes off in demon form (he had it going off within 3 seconds every time unless it was resisted), he's got aggro solidly for the duration. DPS still wasn't really a problem once we got people to not be stupid about whirlwind;, on the first kill we only had two (stupid) deaths before 15%, so DPS was fine. At 15% we did the kind of freak out and zerg him down, I think next time I'll probably be more cautious, in terms of avoiding the WW.
The trick is to NOT go nuts at 15% - the idea is to get him to that point with plenty of time to spare so that you can slow/steady your way through all the whirl winds. Otherwise people get stupid and die because they're trying to nuke through a whirlwind or they yank. You also don't have enough time between whirlwinds to really use your bloodlust to its full potential since the timer comes up much faster. During demon phase you can get the full 40 seconds of bloodlusted DPS time.
Up until 1.5 weapon speed, the completed attack power equivalences are:
Haste Rating = 2.2
Strength = 2 (2.2 w/Kings)
Crit Rating = 2
Agility = 1.8 (2 w/Kings)
Hit Rating = 1.4
Attack Power = 1
Looking at these Values, i am not sure if you thought of Unleashed Rage in your calculations.
Since its permanent up we could count on the 10% extra AP, when we even count Blessing of Kings.
Strength would be 2,2 (2,42 w/Kings)
and
Attack Power = 1,1
Rest would stay the same.
I hope you could understand what i am trying to say. Maybe you already got it in your calculations and i missed it.
(Sorry for my poor English)
UR is already included, what you are doing would be adding it a second time.
No its not included at all, he's very much correct. If you assume that UR is always up when you value itemization, than strength and AP would gain larger weight values.
As far as I remember, the sims and tests from which the weights were derived were done using shaman enhance abilities. As such unleashed rage had an uptime based on the persons crit rate (and a real uptime for in game tests) and the value of crit includes the fact it procs flurry.
My own in game tests which have (bar haste at 1.5 speed) broadly agreed with the weights from the sim have most definately had UR included as it is part of my spec. Kings and the like were not included as they are outside buffs, and therefore need adding as a multiplier (defined by an individual depending on how often they can assume they have kings).
The trick is to NOT go nuts at 15% - the idea is to get him to that point with plenty of time to spare so that you can slow/steady your way through all the whirl winds. Otherwise people get stupid and die because they're trying to nuke through a whirlwind or they yank. You also don't have enough time between whirlwinds to really use your bloodlust to its full potential since the timer comes up much faster. During demon phase you can get the full 40 seconds of bloodlusted DPS time.
I would definitely re-iterate this. Use Bloodlust/Heroism in a demon phase. During the last 15% play it safe: wait to go in until after the first whirlwind and don't push it for the rest of them. The last 15% isn't a dps fight, it's a control yourself fight. Keep your head and you'll pull through it a lot more reliably.
The trick is to NOT go nuts at 15% - the idea is to get him to that point with plenty of time to spare so that you can slow/steady your way through all the whirl winds. Otherwise people get stupid and die because they're trying to nuke through a whirlwind or they yank. You also don't have enough time between whirlwinds to really use your bloodlust to its full potential since the timer comes up much faster. During demon phase you can get the full 40 seconds of bloodlusted DPS time.
We have plenty of time on the enrage timer, that's not the issue at all. We are slightly paranoid though about the demon stacking Chaos Blast debuff high and one-shotting the warlock tank, which is why we DPS zerged phase 3. We will probably not do that in the future.
Originally Posted by Malan
No its not included at all, he's very much correct. If you assume that UR is always up when you value itemization, than strength and AP would gain larger weight values.
No, UR is included in both Pater's simulation and Tornhoof's closed-form model. So these AEP values/stat weights/pawn strings/lootzor coefficients do already account for UR, just like they account for crit and hit improving Flurry uptime. The only possible change would be that STR might technically be worth 2.22 instead of 2.2 if you have Kings; considering this is 1% error I'm not terribly concerned as the rounding errors in Hit, Crit, and Agi are all certainly greater than 1%.
Is this now the "how to maximise shaman dps on leotharis" thread?
That aside UR should always be factored in, on the last BT raid I went on UR was only applied once on each boss fight when I was always attacking the boss.
On that subject, I threw in all the best of slot mail pieces I could find into excel and calculated the difference between non-set pieces and Skyshatter including the 4 piece bonus. What do you guys think? If you take the formula at face value, it's better to get the non-set pieces and forget about Skyshatter. However, when I plotted it out like this, I noticed that the non-set pieces, being obviously hunter gear (they even look like hunter gear from what I gather) they have very little +hit. Now I know +hit has a relatively low value, and I've proven earlier in this thread that dropping really low isn't bad as long as you replace it with other stats. On the plus side though, going with non-set means more agility vs. crit rating, which means you will gain dodge too. Also, the non-set pieces are easier to get as they are sooner in progression and you would have to compete with less people (hunters only vs. warriors and other shaman specs)
So what do you all think? Is it better to go with the non-set? Or is there some reason to try and get the Skyshatter at least to the 4 piece (probably go with non-set legs still, as they are the most gain)?
BTW, I used epic gems in the gem slots for the calculations, so 10 str for red, 5str/5crit for yellow, and 5str for blue.
Edit: Sorry, forgot the conversion I used. Str=1.1, Agi=.97, CR=1, HR=.7, Atk Pwr=.5
Strange, if I use the exact values you listed I get slightly different numbers, the main problem here though is you've only included mail items, and you've missed a few +haste items. If you're maxing dps for each slot you'd use these items:
Cursed Vision of Sargeras - 170
Swiftstrike Shoulders - 103 (mail version scores 100)
Grips of Damnation - 101 (mail version scores 99.2)
Bow Stitched Leggings - 137
Midnight Chestguard - 146
Total - 657 (652.2 using mail versions of similar items)
T6
Helm - 125
Shoulder - 92.1
Gloves - 83.5
Legs - 109
Chest - 114
Set Bonus - 35
Total - 558.6
The difference is pretty dramatic.
Also, to include haste rating in pawn open up it's localization.lua and add {"^Equip: Improves haste rating by (%d+)%.$", "HasteRating"}, to the "Strings that represent statistics that Pawn cares about" section. You can then add "HasteRating=2"(or whatever value) to your pawn string.
Also, to include haste rating in pawn open up it's localization.lua and add {"^Equip: Improves haste rating by (%d+)%.$", "HasteRating"}, to the "Strings that represent statistics that Pawn cares about" section. You can then add "HasteRating=2"(or whatever value) to your pawn string.
Ooh, thanks. I thought this involved more then just messing with the localization. Please add this to the first post as well Malan.
The ratios are 1 CR = 1.43 HR = 2 AP = 0.9 Haste. This means, for example, 1 crit rating is worth 1 'point' while 1 hit rating is worth 0.7 'points'. We just use this ratio to assign each stat on an item a value and then add those values up to get a total that we can compare against other items. A higher total is better.
Skyshatter Grips (assuming +10 crit gem)
+41 Strength (45.1 with kings?)
+45 Stamina
+30 Intellect
Equip: Improves hit rating by 13 (9.1)
Equip: Improves critical strike rating by 31 (31)
Total: 85.2
Grips of Damnation
+27 Agility (26.25)
+38 Stamina
Equip: Improves haste rating by 37 (40.7?)
Equip: Increases attack power by 76 (38)
Total: 104.95
Is my math correct Seb? I've been drooling over Skyshat gloves for a while. but it seems the grips dominate it.
First, I'm not familiar with Pawn (I don't use it since I plan out gear upgrades ahead of time), so would someone tell me if it's possible to add haste rating to this string?
(Malan, looks like the formatting for that part of the post got screwed up when I sent it to you; I'd probably take Disquette's out for now as well, since Pater's and Tornhoof's results are overlapping so nicely.)
And secondly, has anyone modeled the effect of "ignores X armor"? Obviously, I would expect this to vary between mobs, and based on debuffs present (sunder, faerie fire, CoR, etc.) but there is probably some work done on this. Maybe rogues or DPS warriors have done something here?
And a statement --
Although not particularly melee-friendly, Leotheras was a really fun fight. And I really like having a warrior specced Blood Frenzy -- forget Fury as a DPS tree
Choker of Serrated Blades has been netting me about 15 dps (on average) more than Vile Intent. On some fights (low armor targets) it is worth way more, although it would be pretty hard to model it, given the unknown status of armor on any given npc.
Ooh, thanks. I thought this involved more then just messing with the localization. Please add this to the first post as well Malan.
Way ahead of you there.
Regarding the armor deduction procs - do we really need to know how much armor a boss has, or do we just need to somehow understand how the curve functions in terms of their mitigation? Do we know if mob armor functions in the same way as player armor and how the damage curve is plotted for high vs low armor?
07/27/07, 11:26 AM
Similar Threads
