Incorrect. At one point I had over 160,000 parsed consecutive Sinister Strike hits. Its true that could represent something like 99.99999% chance to hit but I find it unlikely. That was with ~20% to hit total using a yellow attack. A good number of those attacks were also on +3 bosses in Nax/AQ so its not like I was running around smashing critters and grey mobs.
It's not quite that bad. Just for comparison sake the probability of swinging 160,000 times and not missing is as follows
.01% chance to miss or 99.99% chance to hit is .00001% chance to not miss in 160,000 trials
.001% chance to miss or 99.999% chance to hit is a 20% chance to not miss in 160,000 trials
.0001% chance to miss or 99.9999% chance to hit is a 85% chance to not miss in 160,000 trials
So if one were to assume that there was some small chance to miss .001% or .0001% would not be unreasonable numbers assuming your data is correct.
The name of the Ace addon I was using is WindfuryFu. I checked the lifetime stats and since I started using it last week, I had over 19000 melee attacks. My overall WF proc % was 19%. I got a new OH weapon a couple of days ago and used FT on the OH for about 2 hours to grind 300-350 weapon skill. Had I kept using my previous weapon, the WF proc % would have been 20%.
I don't know with 100% certainty how WindfuryFu tracks Stormstrike attacks but assuming it just parses the combat log for its data, I'm guessing it counts it as two attacks.
It certainly looks like the 3 sec cooldown for WF has been removed.
It certainly looks like the 3 sec cooldown for WF has been removed.
I'm sure that I am not alone in saying that if you want to make this declaration, we'd like to see some combat logs on the mobs in the blasted lands and conclusive proof of back-to-back windfuries with a weapon faster than 3.0.
I don't really feel like this warrants its own thread in the UI forums, but I threw together a mod earlier to watch the WF cooldown in anticipation of the 5/4 fix. It's just a green dot that goes red for the duration of the 3 sec cooldown, to help time stormstrike without wasting potential WF procs.
I'm sure that I am not alone in saying that if you want to make this declaration, we'd like to see some combat logs on the mobs in the blasted lands and conclusive proof of back-to-back windfuries with a weapon faster than 3.0.
Like I stated earlier, my declaration is strictly based on analysis of the data that WindfuryFu displayed. If the addon is accurately parsing the WF proc data correctly, it would be statistically impossible with a sample size of 19000+ attacks for me to be getting the results I am getting with the 3 sec cooldown still in place (using 2.6 and 1.9 speed weapons).
I posted what I saw in WindfuryFu in case others want to rerun their own tests to verify that something has changed with WF since the patch.
It's not quite that bad. Just for comparison sake the probability of swinging 160,000 times and not missing is as follows
.01% chance to miss or 99.99% chance to hit is .00001% chance to not miss in 160,000 trials
.001% chance to miss or 99.999% chance to hit is a 20% chance to not miss in 160,000 trials
.0001% chance to miss or 99.9999% chance to hit is a 85% chance to not miss in 160,000 trials
So if one were to assume that there was some small chance to miss .001% or .0001% would not be unreasonable numbers assuming your data is correct.
Maybe, but also consider in all my time raiding from early MC to KT I never ever once saw a yellow attack miss. There is a small chance I would not notice a miss but not very high since I have always used sct. Even back when I only had 5% hit from talents and a couple points from gear, yellow just doesn't miss.
I would hate to guess how many Sinister Strikes, Backstabs and Hemo's two years of raiding involved, but its a very very large number.
I also found some bad results with a different mod - procwatcher. I got procs of over 20%.
(That mod reports hits and crits disaggregated, and doesn't report missed WF hits. Using same-speed weapons and WF4/3, I parsed until I had over 1000 hits. At the end, I added up Hits+Crits+SS hits+SS crits. Then I took WF procs and divided by two. The result of (WF/2)/(Hits+Crit+SS+SScrit) was a proc rate of about 23%.)
I'm very hesitant to do any more parsing until I have a proc tracker that I trust. Once I hit 70, I'll see what's out there and see if I can contribute. (As a side note, it may be a nice feature for the WF-specific proc tracker to check weapon speeds, and use them to calculate the correct proc rates when using different MH and OH enchants. Just knowing # of hits isn't enough - you want to know how many MH hits and how many OH hits. I think the only way to know is (1/MH speed) / (1/MH speed + 1/OH speed).)
After reading a few other threads, I came across somethng I did not consider and did not see mentioned after the graphs, and that's TOTAL damage increase with haste, not just WF damage.
Essentially, you hit the "sweet spot" of WF ppm Nite Moogle mentions at effective speeds .5, 1.0, 1.5, and 3.0. Well, if flurry reduces the speed of a 2.0 weapon to 1.5, wouldn't a haste proc/enchant that reduces that speed to 1.0 or .5 be even better?
Maniq in the raid spec thread menitons that with his Dragonmaw (200 haste proc, 2.7 speed) main hand and blackout truncheon offhand (132 haste proc) and Abacus (260 haste activated) he gets an effective 1.1 speed in his mainhand with 2 procs. This produces only 1 less potential ppm than 1.5 speed, but nearly 30% more white damage from the MH than 1.5 speed.
I don't know the calculations on haste rating, but I am curious to know what flurry + Blackout proc is, and then what the sweet spot speed for the MH is to get 1.0 speed. I would hope the massive increase in white damage would offset the Blackout's lesser damage and then some.
Alright, I did some research, and according to the Rogue DPs thread, 10.5 haste rating = 1% haste.
Either this is wrong for shamans, Maniq is wrong, or I just have my math wrong on Flurry.
592 haste rating = 56.38% haste
2.7/1.5638 = 1.73 speed (procs + abacus)
1.73/1.3 = 1.33 speed (Flurry Multiplied).
I find it much more likely that flurry is added, not multiplied, so...
56.38+30=86.38.
2.7/1.8638 = 1.45
Neither of these are the 1.1 he claimed, so I am curious where he is getting that number.
But to get back to the Blackout Truncheon. 132 haste = 12.6 haste. Multiplied by flurry, it's 46%, added, 42%.
1.5 *1.46 = 2.19 (2.2 is optimal), and it's pretty much the same for additive. Blackout gets very close to 1.0, so maybe the best combo of weapons is Fel Edged + Blackout?
I have only read the first two pages of this really but just wanted to throw some things out there for discussion.
Enhancement shamans biggets weakness has always been there instant mildly controlled burst damage. which has huge ranges. Pre TBC it was the number 1 thing that killed me. Post TBC though I believe that the threat mechanics of a enhancement shaman can be controlled. you bring up that through the current 5/4 it allows a quad wf ( if I understood it right) through duel wield. Now the shaman can control the extreme burst by switching the encahnts on his weapons. But turning this ability on the enemy in pvp tio me seems completely acceptable.
I say this mainly because unlike pretty much everyone else we have no cc so a insain burst dps is what made shamans a threat in pvp. we might not be able to stop you from attacking us but if we get close to you it might be over in 2 secs has always been even in my eyes.
Currently I am using a two hander for instances because it has a high dps out put with wf but with the slower weapon swings I can monitor what I am doing a little better and allow me to cut attacks a little easier. but allows a shorter return on shamanistic rage.
Basically my arguement will stand at. If you want controlled somewhat stable dps in raids and 5 mans you can do what all damage classes do and scale down or buff yourself so that you cant easily cross aggro thresholds. But for pvp we liek every other class should be allowed to throw ever single piece of damage out that we can.
A semi fare trade in my eyes would be if you stamp a manditory 3 sec cooldown on wf procs perhaps we should get a 25% 3 sec stun on SS. maybe reduce the nature boost down to 10%
But then again the change to SS would seem a nerf to some and overpowered to others. so I would just say leave everything the way it is and leave it to the players to control there aggro and not blizz.
WoW shaman forums to the contrary, I don't see much danger of haste effects reducing our DPS. Math follows.
Assume you're autoattacking with a 2.7-speed mainhand weapon, no haste effects, no Flurry. We'll assume you don't have an offhand equipped for the purposes of this demonstration, because (1) it makes the math easier, and (2) it means I don't have to take into account MH procs proccing offhand WF and vis versa. We're also going to assume that Windfury just does a flat 225% weapon damage - in reality mileage will vary depending on your AP, weapon damage, crit, hit, etc, but it's a decent approximation.
Each time you swing and proc WF, you swing again 2.7 seconds later and can't proc WF off of that second swing. On one in six attacks you lose an opportunity to proc WF, which brings your WF proc rate from 20% (one in five), to 16.67% (one in six). Windfury thus gives you an additional 16.67% of 225% of your weapon damage. By putting up WF you're dealing 137.5% of your base damage.
Now say you have Flurry up. You're swinging 30% more often which means you have 30% more chances to proc WF, and you haven't pushed yourself to the point yet where you're making two swings inside the WF cooldown. Your attack damage is now 137.5% *130%, or 178.75% of your normal damage.
So what about a weapon where the 30% haste increase would push you to the point where you're losing an additional chance to proc WF? A 2.0-speed weapon for example.
With a 2.0-speed weapon and no Flurry, you lose one in six opportunities to proc WF. Windfury still gives you 137.5% of your unbuffed DPS, just like with the other, slower weapon.
With Flurry, your WF proc rate per swing starts going down. Flurry pushes your attack speed below 1.5, so each time you proc WF you make two subsequent attacks while under the 3-second cooldown. Your WF proc rate drops from 16.67 (one in six) to about 14.3% (one in seven). From Windfury you now gain only 14.3% of 225% of your weapon damage, which puts you about 132% of base. However, you're also flurried, making 30% more attacks and getting 30% more opportunities to proc WF at your lowered proc rate. Factoring in Flurry, your damage is 130% of 132% of base. With both WF and Flurry factored in, you're up to 171.6%.
With the slower weapon you get more out of WF and Flurry in combination (178.75% weapon damage vs 171.6%) but the difference isn't huge and with either weapon you deal more DPS with Flurry than without. The point is, even if your weapon is fast enough that Flurry ir another attack speed increase costs you an extra chance to proc Windfury, haste effects still increase your DPS considerably. This holds true for Flurry, it's going to hold true for Bloodlust, it's going to hold true for Dragonspine Trophy and the Abacus of Violent Odds and the BoP hammersmithing weapon. It's going to hold true for two or more of those haste effects stacked on top of each other.
The only time haste would decrease your DPS would be a tiny increase to swing speed that pushed you from just over to just under the brink of losing an extra chance to proc WF. I'm not sure even it's possible with any combination of gear in the game right now. Unless you're rocking the old +attack speed to glove enchant and a 3.1-speed 2hander with a counterweight on it, you have nothing to worry about.
The only time haste would decrease your DPS would be a tiny increase to swing speed that pushed you from just over to just under the brink of losing an extra chance to proc WF. I'm not sure even it's possible with any combination of gear in the game right now. Unless you're rocking the old +attack speed to glove enchant and a 3.1-speed 2hander with a counterweight on it, you have nothing to worry about.
See there is the sticking point.
It is absolutely undisputable that at some times haste will result in lower total dps. When it results in higher dps it will be trading white damage for yellow damage. White damage is flat out worse than yellow, white glances, yellow doesn't, white misses at 24%, yellow misses at 5%, etc. Against a 73 mob the gap where haste is a detriment will be larger.
Almost completly useless, but an interesting looking demonstration of why people generally don't generate images with javascript and tables, with over 6 megabytes of html generated (126 times the size of the PNG version, and 18 times the size of the bmp). X coordinate is weapon speed (.5 to 3.5, Offhand 90% MH speed), Y coordinate is haste (2% to 60%). Brightness is WF damage. Light-colored streaks are optimal haste/weapon speed combinations, dark streaks are places where the haste is reducing WF damage -- there's actually a lot when you look at haste in .2% steps.
Generating a list of situations using currently accessable gear/haste buffs where the haste would result in a net loss of damage would probably be doable if that'd be of interest.
When it results in higher dps it will be trading white damage for yellow damage.
You're not necessarily trading away yellow damage. In my example with the 2.0-speed weapon, you're actually gaining more Windfury procs by having flurry up.
Without flurry, the 2.0 speed weapon gives you +37.5% damage.
With flurry, it gives you 130% of +32% damage: +41.6% damage from windfury. That's not including the white-damage increase from flurry, that's the damage you gain from windfury alone. The additional chances to proc WF more than make up for the reduced proc rate.
While this might not be true for every combination of weapon speed/haste rating, at the worst you're going to trade a substantial increase to white damage for a minimal loss of WF. That is, unless we're talking about very small haste increases (1-3%), and I can't think of any source of haste from spells or lv70 gear that gives less than 10%.
This isn't the point, the point is that because of the heavy dependence on Windfury a haste item is nowhere near as good as it would seem at first blush on a Shaman. The fact that such a situation could conceivably exist is stupid beyond expression. You're probably going to get better results out of a comparable-level trinket that improves AP or crit than a haste trinket.
From what I understand about the current Windfury each hand has a hidden 3 second cooldown when using different levels of the spell: i.e. MH should have 3 seconds between procs, and OH should have 3 seconds between procs when using different ranks.
If this is correct, then I had an interesting event when playing around on my shaman today, and I thought somone here might be able to clear this up for me.
You will notice that this seems to be two mainhand procs (this is a standard hit for WF for me at the moment) within one second. There is also a stormstrike in between. This leads be to believe one of three things:
1. There is no (3s) cooldown. (Unlikely)
2. There is a bug with your OH weapon giving MH damage.
3. Stormstrike resets this 3s hidden cooldown.
I really can't see any way of easily proving any of these, so I was hoping somone knew more about it. (Or maybe this means nothing and my understanding of WF is incorrect.)
Option nr2 is known; quadprocs have a way of sometimes giving 4x mainhand, or 4x offhand dmg. This screenie would suggest there's also a cross polination without a quadproc.
Alternatively, Blizzard have ninjaed in a lower cooldown. Sadly it would take several thousand swings with all kinds of addons running to prove that, and then we still wouldn't know if it's intended ;-)
So since I am planning to change to elemental in the mid/late 60s, I've been playing around with this alternate WF on weapons while DW'ing to see how it performs. I'm seeing some odd behaviour though. Now I've not done a big analysis of combat logs, but I have looked over mine while fighting as well as looking at my SCT.
I see these 4 in a row WF (either offhand of mainhand) and it is nice, but there are two weird things I see that I didn't see when going with same level.
1) A lot more misses, dodges, parries than I did before. This could be just that its proccing more often and thus more of that, but I'm not sure. This one is just more of a feeling though.
2) The big one. Whenever I see these 4 in a row hits with MH or OH. None of them ever seem to crit. They're all just regular hits. It seems to me that splitting the WF is giving me weird behaviour on crits. Where before I'd get my double proc WF and quite often I'd get one crit or even 2, now I just get 4 hits time after time. Anyone else notice this?
2) The big one. Whenever I see these 4 in a row hits with MH or OH. None of them ever seem to crit. They're all just regular hits. It seems to me that splitting the WF is giving me weird behaviour on crits. Where before I'd get my double proc WF and quite often I'd get one crit or even 2, now I just get 4 hits time after time. Anyone else notice this?
No you're just noticing patterns where there are none. I've defintely gotten 4 MH procs with 1-2 crits, I'll end up doing in the area of 3500-4200 dmg from it.
03/12/07, 2:59 PM
Similar Threads
