Future of a DPS Warrior - Page 53

Gruntle · 09/27/07, 5:12 AM

Originally Posted by Garithras

Let me try spelling it out a little differently. Total threat = (damage + bonus) * modifiers. We can agree on this, yes? HS threat = (damage + 176 [bonus damage] + 196 [threat bonus]) * mods. Cleave threat = (damage + 70 [bonus damage] + 130 [threat bonus]) * mods. The bonus damage is reflected directly in the combat log; there is no need to add additional bonus threat.

Yes, you are right that HS produces more threat than Cleave. But you do not have to add TWO bonuses onto the damage the ability causes to get the actual threat created, which is what you are arguing for.

Sorry you must have misunderstood me. I'm not adding the damage twice to the threat. I agree it might look like that in my post (will try to edit to make it clearer). I'm saying the same thing as you, the threat from HS is (damage+bonusthreat) *modifier (although the modifier is not very interesting in this case, since it's the same for both Cleave and HS).

What both me and Natural agree on (I think

) is that the extra damage a HS or Cleave does is not only the bonus damage (176 or 70). The abilities also do extra damage (on average) because they miss less and don't glance. This is what I try to use in the Expected damage differences.

Natural · 09/27/07, 11:26 AM

Originally Posted by Gruntle

Sorry you must have misunderstood me. I'm not adding the damage twice to the threat. I agree it might look like that in my post (will try to edit to make it clearer). I'm saying the same thing as you, the threat from HS is (damage+bonusthreat) *modifier (although the modifier is not very interesting in this case, since it's the same for both Cleave and HS).

What both me and Natural agree on (I think

) is that the extra damage a HS or Cleave does is not only the bonus damage (176 or 70). The abilities also do extra damage (on average) because they miss less and don't glance. This is what I try to use in the Expected damage differences.

My calculations show which is more threat efficient in terms of the wasted "extra" threat from the ability itself.

If you want to calculate TOTAL threat efficiency, as you said, you certainly have to include the threat caused by the damage. But that isn't necessary in order to decide which attack is more threat efficient because the raw damage-to-threat ratio is a constant. The only variables between the two abilities are the wasted threat and the expected damage.

Jeho · 09/27/07, 1:54 PM

Personally, I feel the problem with threat caps (for various classes) has less to do with their innate skills or abilities, but moreso with the fact that tank threat scales very poorly. (You can't "crit" a sunder.) So while most classes can put upgrades to use in their bottom line (DPS numbers) a tank rarely sees anything substancial that will increase their rage/threat generation. (Not to mention the flawed mechanic by which the more damage you mitigate the less threat you generate.) As a tank gets upgrades, their threat output tends to increase very little.

However, what I would like to see when the level cap is raised (and talents are reworked) would be a talent/ability to reduce threat in battle/zerker stance by X. (Say 5%, to a max of 10% for 2 talent points, something that would be 'on use' - so that it would not impact the ability of a DPS warrior to tank an instance.) I think DPS warriors are too hard on themselves since they do a great job of doing DPS in a situation where the threat ceiling is low. (IE, 110% vs 130% for ranged) I think the only class more screwed over by threat are the shamans.

Natural · 09/27/07, 3:06 PM

Originally Posted by Jeho

Personally, I feel the problem with threat caps (for various classes) has less to do with their innate skills or abilities, but more so with the fact that tank threat scales very poorly. (You can't "crit" a sunder.) So while most classes can put upgrades to use in their bottom line (DPS numbers) a tank rarely sees anything substantial that will increase their rage/threat generation. (Not to mention the flawed mechanic by which the more damage you mitigate the less threat you generate.) As a tank gets upgrades, their threat output tends to increase very little.

There are counter-argument to this. Advantages of "mostly static" threat with gear:

1) Boss tuning can be done more easily since the threat caps are more constant
2) TPS is more skill-based than gear based, letting lesser-geared tanks put out enough TPS for raiding
3) The Dev's don't have to worry about threat scaling out of control for highly geared warriors, which could potentially make threat meaningless if nobody had to worry about it.

However, what I do NOT like is when tank threat actually decreases with gear due to mitigation. Better gear leads to more mitigation, more mitigation reduces damage, less damage produces less rage, and finally less rage produces less threat. This chain of events has our tank actually downgrading his tanking gear or purposely not putting up debuffs like Tclap in order to get hit enough to build rage. Frankly, this mechanic sucks.

Garre · 09/27/07, 4:47 PM

Has anyone seen a clear drop in their DPS post-patch, due to the Haste nerf?

I've also heard rumors of a proc rate nerf to [Dragonstrike], and considering the 150+ drop in my own DPS that I've seen, this might be part of the cause.

Bouncey · 09/27/07, 5:25 PM

WoW Forums -> So Dragonstrike Sucks Now?

Looks like a huge hit to me.

Hozz · 09/27/07, 5:25 PM

Originally Posted by Garre

Has anyone seen a clear drop in their DPS post-patch, due to the Haste nerf?

I've also heard rumors of a proc rate nerf to [Dragonstrike], and considering the 150+ drop in my own DPS that I've seen, this might be part of the cause.

There was definitely a proc rate drop on Dragonstrike.

20.6% for me personally from Hyjal last week to Hyjal this week.

Xylik · 09/27/07, 5:47 PM

Would you say the hit to Dragonstrike is worth changing to the Claw off Al'ar or the T3 axe?

Xy.

Chadw · 09/27/07, 6:09 PM

All thoughtout Hyjal yesterday I didn't notice a change in the proc rate of dragonstrike. Although I wasn't looking very closely.

Graul · 09/27/07, 7:31 PM

In other news, Onyxia does Deep Breath more. Don't use the WoW forums as your empirical evidence.

Emeraude · 09/27/07, 7:42 PM

At the end of the PTR testing phase I compared the live Dragonstrike to the PTR one, the weapon itself didn't seem to have a lower proc rate, but it might seem like it procs less because haste overall has been nerfed.

Machinator · 09/28/07, 12:14 AM

[quote=Jeho;490533
However, what I would like to see when the level cap is raised (and talents are reworked) would be a talent/ability to reduce threat in battle/zerker stance by X. (Say 5%, to a max of 10% for 2 talent points, something that would be 'on use' - so that it would not impact the ability of a DPS warrior to tank an instance.) [/QUOTE]

I had the idea of the next level of talents in fury would reduce threat by some amount, 10% or so to be close to the base of cat/rogue, while rampage was up. This would give a little more to the non-scaling rampage and be an 'active' ability.

Gruntle · 09/28/07, 4:40 AM

Originally Posted by Natural

The only variables between the two abilities are the wasted threat and the expected damage.

Well, the difference in expected damage means that also the threat from damage produced by the two abilities are different. You're correct in that the only two variables are bonus threat and expected damage, but the expected damage difference enters in two places in the efficiency. If this difference wouldn't make a difference the resulting threat efficiency should be the same regardless if you use only the bonus threat or the total threat. This is not the case.

Anyway, I guess we just disagree, not much point discussing it further

Chadw · 09/28/07, 5:04 AM

Originally Posted by Hozz

There was definitely a proc rate drop on Dragonstrike.

20.6% for me personally from Hyjal last week to Hyjal this week.

Chadw - WWS

I had 'Haste' proc 12 times in the 5+ minutes it took to kill Naj'entus. Now I can't find any WWS of previous weeks I am almost positive that is on par if not better then previous weeks.

Hozz · 09/28/07, 9:41 AM

Originally Posted by Chadw

Chadw - WWS

I had 'Haste' proc 12 times in the 5+ minutes it took to kill Naj'entus. Now I can't find any WWS of previous weeks I am almost positive that is on par if not better then previous weeks.

Do you have a DST? Anything else that would be causing 'Haste' in the logs, like Haste pots? I dont think there is an internal CD on the Dragonstrike but my proc rate is sucking, its definitely been less than pre-patch.

Natural · 09/28/07, 12:40 PM

Originally Posted by Gruntle

Well, the difference in expected damage means that also the threat from damage produced by the two abilities are different. You're correct in that the only two variables are bonus threat and expected damage, but the expected damage difference enters in two places in the efficiency. If this difference wouldn't make a difference the resulting threat efficiency should be the same regardless if you use only the bonus threat or the total threat. This is not the case.

Anyway, I guess we just disagree, not much point discussing it further

To put an end to this discussion, I've created a spreadsheet to model the threat efficiency of Heroic Strike vs. Cleave. It has a few variables you can use including crit rate and mob armor %.

I've used both my calculation (Bonus dmg / ability threat) and your calculation (Bonus dmg / Ability threat + bonus dmg threat) to prove that they give you the same result. Both calculations have Heroic strike & Cleave efficiencies intersecting at the same point. As I've said, the more efficient attack is the same regardless if you use the bonus threat or the total threat.

cleave_vs_hs_v002.xls - FileFront.com

Please share your comments. Thanks.

Notes (included in spreadsheet)

-Heroic strike will always be more rage efficient
-The two abilities have very similar threat efficiency for the 100-200 range
-There are two graphs here to prove that the intersection of threat efficiency (where one becomes more efficient than the other) is independent of the raw +damage bonus provided to the abilities (e.x. 176 for HS). You can change this number and it will move the intersection point, but both calculations result in the same intersection
-You may adjust the armor, crit rate, ability bonus, and ability threat in cells B64-B69.
-The X axis represents the bonus obtained from converting your main hand attack from a white attack to a special attack. A special attack gains the following bonuses:
1) Cannot be a glancing blow
2) Base miss rate of ~8.6% rather than ~29% for dual wield
3) Gains +10% increased damage on critical strikes from impale

-You must calculate the expected damage based on your attack table and weapon damage. This spreadsheet does not show these calculations. You should figure them out yourself.

-As your gear gets better, you move farther to the right on the X axis

Chadw · 09/28/07, 3:25 PM

Originally Posted by Hozz

Do you have a DST? Anything else that would be causing 'Haste' in the logs, like Haste pots? I dont think there is an internal CD on the Dragonstrike but my proc rate is sucking, its definitely been less than pre-patch.

Nope the only item I have/use that causes 'Haste' is Dragonstrike.

Although I looked at our Hyjal WWS logs the day before and it was considerably lower so I guess I just got lucky?

And there definitely isn't an internal cooldown on dragonstrike.

Grymm · 09/28/07, 3:41 PM

I don't quite get the results of the spreadsheet. It's not constructed similarly to how I would evaluate the use of one versus the other.

Here is how I evaluated it:

[editted out my explanation as it is now redundant.]

You have a 100 dps, 2.5 speed weapon with 3500 AP. The boss takes 85% damage (15% mitigated), you hit 85% of the time (miss+dodge = .15) and you crit 35% of the time. Glancing blows are 24% at 75% damage.

White Hits: 26% - .85*(3500*2.5/14+2.5*100) = 744 damage
White Glance: 24% - .75*.85*(3500*2.5/14+2.5*100) = 558 damage
White Crit: 35% - 2*.85*(3500*2.5/14+2.5*100) = 1488 damage
White Miss: 15% - 0
Average White: .26*hit+.24*glance+.35*crit+0=580 damage

HS Crits: 35% - 2.2*.85*(3500*2.5/14+2.5*100+176)=1965 damage
HS Misses: 5.6% - 0
HS Hits: 59.4% - .85*(3500*2.5/14+2.5*100+176) = 893 damage
Average HS: .35*crits + .594*hits = 1219 damage

Damage returned on HS = HS - White = 639 damage
Aggro from HS = damage*.8*.7+196*.8*.7 = 467
Efficiency = Damage returned/Aggro Returned = 1.366

Cleave Crits: 35% - 2.2*.85*(3500*2.5/14+2.5*100+70)=1767
Cleave Misses: 5.6% - 0
Cleave Hits: 59.4% - .85*(3500*2.5/14+2.5*100+70) = 803
Average Cleave = .35*crits + .594*hits = 1096

Damage Returned on Cleave = Cleave - White = 516
Aggro from Cleave = damage *.8*.7+130*.8*.7 = 361
Efficiency = Damage returned/Aggro Returned = 1.426

Same result but easier for my small brain to follow.

Natural · 09/28/07, 3:46 PM

Originally Posted by Grymm

If HS and Cleave had 0 WD and never crit, an average HS/Cleave would be more damage than an average MH swing because of the loss of glance and miss. You spread sheet does not seem to count this return as a gain. This would inflate the value actual expected damage from using either of these moves.

Please take a closer look at the spreadsheet. The scaling portion of the damage bonus is the X-Axis. I've been advocating the understanding of these scaling bonuses in many of my recent posts in this thread and mentioned it in both the spreadsheets and my previous post.

Grymm · 09/28/07, 3:58 PM

Sorry, I started my post as I was going through the spreadsheet and my own math. I've been editting for the last couple minutes and not quite done.

Natural · 09/28/07, 4:03 PM

Originally Posted by Grymm

Adding an example.

You have a 100 dps, 2.5 speed weapon with 3500 AP. The boss takes 85% damage (15% mitigated), you hit 85% of the time (miss+dodge = .15) and you crit 35% of the time. Glancing blows are 24% at 75% damage..
.
.
.
.

Same result but easier for my small brain to follow.

I believe the logic behind your results (Calculating expected damage & comparing vs. white is correct. My spreadsheet is different in that it pulls out the bonus +damage bonus for HS/Cleave so that you can play with the numbers and potentially use it in the future as the abilities change. Plus it gives a graphical representation of how things will change with gear.

Grymm · 09/28/07, 4:08 PM

One note: In your calculation of HS bonus damage, you do not use the impale (2 crit multiplier instead of 2.2). Maybe since you are starting from expected returns already and not from base crit/hit/glance mechanics it doesn't matter.

Second note: My efficiencies will assume .8*.7 threat reduction as if I don't have salve I generally jump on top of the light paly (who will usually give me my salve) until he does so.

I originally found it difficult to use the spreadsheet as it does not start with player stats but rather the expected return of cleave/HS that you would already have needed to calculate from your stats. Looking at the X axis, I don't readily follow where I am on the graph.

Natural · 09/28/07, 4:14 PM

Originally Posted by Grymm

One note: In your calculation of HS bonus damage, you do not use the impale (2 crit multiplier instead of 2.2). Maybe since you are starting from expected returns already and not from base crit/hit/glance mechanics it doesn't matter.

Thanks, it should be counted. I'll fix it. =)

Second note: My efficiencies will assume .8*.7 threat reduction as if I don't have salve I generally jump on top of the light paly (who will usually give me my salve) until he does so.

My goal is to compare HS vs. Cleave. I've neglected threat multipliers since it will affect both equally, but you're correct that it's necessary if you want to calculate the final threat created.

I originally found it difficult to use the spreadsheet as it does not start with player stats but rather the expected return of cleave/HS that you would already have needed to calculate from your stats. Looking at the X axis, I don't readily follow where I am on the graph.

Yeah, I chose to skip this calculation at first. I'll throw a simple calculator in there for the next revision.

Thanks for taking a look at it. The spreadsheet was as much to prove the point to myself as it was to share with you guys.

Natural · 09/28/07, 4:45 PM

Originally Posted by Grymm

White Hits: 26% - .85*(3500*2.5/14+2.5*100) = 744 damage
White Glance: 24% - .75*.85*(3500*2.5/14+2.5*100) = 558 damage
White Crit: 35% - 2*.85*(3500*2.5/14+2.5*100) = 1488 damage
White Miss: 15% - 0
Average White: .26*hit+.24*glance+.35*crit+0=580 damage

I think your calculator had a typo here. Your result should be 848.

Edit: Here is a link to the updated calculator. cleave_vs_hs_v003.xls - FileFront.com

It now uses the impale bonus when calculating the damage bonus for heroic strike and cleave. It also adds a pseudo-attack-table calculation to determine your expected damage of a white & a special attack. It calculates the "special attack bonus" that you can use to find your place on the X axis. Please let me know if you find any errors.

Thanks.

Edit2: I am going to move this post to the DPS warrior spreadsheet. We can continue the discussion there since it is probably more appropriate.

Grymm · 09/28/07, 6:27 PM

You are correct. Something there doesn't work out. I set up to fall through in excel so I can play with the base values and see how things behave. I closed that spreadsheet without saving though as it was just scratch work so I'm not sure if I copied something over wrong or if I tinkered with the base values while copying stuff over thus changing what everything should add to. I probably won't reconstruct it. It was enough to just see the mechanics. The actual numbers don't mean much provided the behavior is the same.