In my eyes, it does really invalidate it. Your theoretical max dps is never going to be achieved anyway, and there's no rules or ideas as to which is the best for practical dps. What if you have to move during the Mjolnir proc? Then full ArP wins. What if you get stunned or knockbacked during Mjolnir proc? Most fights are even so short now that its mostly down to basic luck what kind of dps you do. Also, with the introduction of the kind of gear we see in ToC there aren't even any choices for what gear you want to equip, except 2-3 pieces. No 2-3 pieces will make up for a great DPS loss or gain. All you have to do is look at the theoretical numbers and see how close to each other they are. People nitpicking about sets when in THEORY they could POSSIBLY gain 100 dps or less. Fantastic.
I miss more hands-on comparison of DPS. People doing the fights and sharing experiences. What works where, etc. Maybe I'm on the wrong forum for that tho
Skill is of course a huge part of reaching maximum dps, no doubt about that. Alot of the finer points of how to dps on a certain fight gets somewhat lost when everything evolves around BiS lists and stat weights. FBN is (if i understood correct) coded around a specific rotation, as is Rawr and other calcs(atleast when calculating BiS lists). This and that might not always hold true on every fight. You would still need some sort of mathematical approach to these situations though, to be able to back up claims of what is more and less dps. Instead of saying "i did it this way on that fight and did xxxx dps". That situation can be a good starting point for what to actually input to a sim like FBN, but it wont stand alone.
I've not seen any tests with Banner of Victory+Death's Choice yet, how come?
Even after the ArP nerf I'm still able to hit 85% passive ArP, which is where the breakpoint is supposed to be. Again tho, all the tests showing 10500 dps and above are totally utopic. I'm very close to having the BiS gear for hardcapping ArP and I barely do 9400 dps on Jaraxxus (not changing target). Somehow I dont see 1 trinket, a cloak and 1 245 > 258 upgrade giving me 1.1k extra DPS. So my question is.. how much of that is down to logical flaws in the program and how much is the players fault? I've never used Rawr, FBN or anything else; I've always used my own logic. Thats solely because I dont trust Rawr, FBN or whatnot because the numbers generated by those programs are ridiculous in all respects.
Actually you did rely on them all the time.
The pure knowledge of your hit cap, the fact that ArP is your best DPS Stat, etc. ...
All that is derived from theory crafting.
I also though it was weird that you write the sims can't be trusted and then ask for a sim about BoV + DC.
Of course those numbers always leaves room for personal preferences.
Programs like Rawr try to take the mathematical calculatable part of the game to a field of possible obejective discussion.
If we were all just to assume that all theorycrafting is too "theoretical" and as such useless then we would be at a position where we could just stop all calculations and everyone writes what he or she thinks is the best, then posting their WWS logs.
This of course might have his place. But it should never be the center or only focus of DPS discussion.
Since those would be extremely biased due to personal performance, etc. ... the results would be pretty pointless.
Theorycraft does not dictate your playstyle. It's just giving you a well analysed basis for your descisions.
Originally Posted by Abbichum
What programs like Rawr and FBN calculates is Maximum theoretical dps. In-game will always be less than that due to a number of reasons(lag, movement, buffs not being present 100% of the time and what not), and there might be a few flaws in regards to how trinkets and other procs are calculated, since we don't have full access to game-code. That does not invalidate the stat values or BiS gear list though, and it will most certainly give a better picture than common-sense logic.
They can be both higher or lower.
Think of all the "gimnick" - encounters like Hodir.
Or all the encounters where AE phases matter, like Freya, wich does not get calculated in the outcome of sims in tools like rawr.
Or even target switches.
But I totally agree that this does not invalidate theorycrafting or sims in general.
@Furial
[Stupid comment deleted due to my brain waking up ]
Found the mistake on my part.
Though I still wanted to mention that I am confused on your numbers.
Looking at your Hardcap + 4xT9 + 3x ivl258 pieces (Neck, Bracers, Ring1) sim, wich I used as a basis for my own hardcap sim you get: 68 ArP (Shoulders) + 68 ArP (Feet) + 61 ArP (Hands) + 50 ArP (Ring 2) + 49 ArP (Back) + 94 ArP (weapon) = 390 ArP from Gear.
Paying no attention to the metagemslot and the headslot where your nightmare tear is gemmed you have 18 Gemslots.
18 x 20 = 360
360 + 390 = 750
Even considering you would be BS and JC this would be 390 ArP (from the gear) + 16 x 20 ArP (regular ArP gemslots minus two for 2 x fractured Dragon's Eye) + 3 x 34 ArP (the two gemslots from the gear + one from BS for the Hands), wich would end up being = 812 ArP Rating.
You could get an additonal 40 from bufffood ending up at 852. But I really do not understand how you get the 892?
Now the only way I found to get even more ArP would be to use an elixir of armor penetration over the Flask of Endless Rage. But that would lower my overall DPS and end up at 897, again not matching your numbers?
I will most likely have done some mistakes on my math here, but I just wanted to point this out.
Last edited by Curan : 09/30/09 at 7:53 AM.
Reason: @Furial
I miss more hands-on comparison of DPS. People doing the fights and sharing experiences. What works where, etc. Maybe I'm on the wrong forum for that tho
I have redone my simulations with the great input I got here (special thanks to Furial, Odas and Jodanu) and wanted to post the numbers for some feedback and thoughts:
I did not include gains from professions to keep this general since I did not want to assume a certain orofession for the sims.
I did not include the 1% hit aura from the draenei as it is restricted to the aliance only and again I wanted to keep this general.
I assume a fully supported Raid-Cat.
I did not incluede the following Buffs that could alter your DPS outcome, due to their very special nature: Hysteria, Tricks of the Trade, Shattering Throw, and Heroic Presence (as mentioned above).
I assume a lag variance of 200 ms.
I simulated a 300 seconds / 5 minutes long encounter.
Gemming: Glinting Ametrine x 4(Head, Shoulders, Finger2, and Feet), Fractured Cardinal Ruby x 13, Nightmare Tear (Weapon blue slot), Delicate Cardinal Ruby x 3 (Weapon, 2 x Waist) and Relentless Earthsiege Diamond (Meta)
Attack Power: 13178
Agility: 2681
Strengh: 371
Crit Rating: 599
Hit Rating: 258
Expertise Rating: 142
Haste: 557
ArP Rating: 744 DPS: 10270
Some things I wanted to point out as rather interesting:
Berserking always ended up as the best enchant for the weapon in any given situation in terms of maximum DPS.
The Hardcap and Softcap simulations almost come out with the same BiS Gear. With the biggest difference being the Mjolnir Runestone getting traded for a dual Death's Choice / Death's Verdict and being able to take some of the agiliy – socket boni on the softcap version.
The four main DPS maxed Gearsetups vary only around 63 DPS, wich really isn't that much. This kind of sums up for me to all of them being pretty viable in term of endgame DPS, even though the softcap does come out best.
Taking the numbers I have here and putting them into Toskk I get complete different resulsts. In example my highest DPS sim here was Nr 2 with 10428. Toskk tells the DPS Output to be 10946.3281 DPS with the exact same numbers. Does anyone has an explanation here for me?
You lose 158 DPS according to the sims if you prefer stable rotation over just maxed out DPS and want to go for the caps.
The impact of the professions on the current lategame DPS:
Again I took my simulation Nr 2 as a basis. Especially in terms of jewelcrafting this numbers are by no means meant absolute.
Of course it is commonly known that Jewelcrafting is the best boost for DPS as a profession and that they are about even. I also realise that synergy plays a great role on profession gains like Blacksmithing + Jewelcrafting making a great combination. I just wanted some clear numbers and thought I might as well post them here when I take a look at them.
Profession
DPS Increase
Alchemy
+58
Enchanting
+58
Herbalism
0
Mining
0
Skinning
+57
Inscription
+58
Leatherworking
+58
Blacksmithing
+36*
Engineering
+79
Jewelcrafting
+94**
Tailoring
+61
Jewelcrafting+Blacksmithing
+130
Jewelcrafting+Engineering
+174
*One additional Socketslot on the Hands with 1 Delicate Cardinal Ruby
**3 x Fractured Dragon's Eye (Back, Chest, Waist), changed 2 x Fractured Cardinal Ruby in the Chest and Waist to 2 x Deft Ametrine to get agility socket boni due to having extra ArP from the 3 x Fratured Fragon's Eye.
Last edited by Curan : 09/30/09 at 10:13 AM.
Reason: Underlined the DPS resulsts
Could we please not post any more Rawr "simulations" in this thread? It makes too many illogical assumptions, in my opinion, to get an accurate read on itemization. The major points I have against Rawr's numbers are that bleed and SR uptimes were pretty much assumed to be 100%, and it finds the total number of extra combo points you'll have over the course of the whole fight, and uses them all up on 5cp bites, which I'm sure is inaccurate. It also assumes that every SR use is at 3/4/5 CPs when in actuality, it varies during every fight.
Really, it's not a fault of the program or its coders, but in the program's use of a mathematical model used to estimate feral DPS versus repeated simulations with tools such as FBN and Simulation Craft.
In my eyes, it does really invalidate it. Your theoretical max dps is never going to be achieved anyway, and there's no rules or ideas as to which is the best for practical dps. What if you have to move during the Mjolnir proc? Then full ArP wins. What if you get stunned or knockbacked during Mjolnir proc? Most fights are even so short now that its mostly down to basic luck what kind of dps you do. Also, with the introduction of the kind of gear we see in ToC there aren't even any choices for what gear you want to equip, except 2-3 pieces. No 2-3 pieces will make up for a great DPS loss or gain. All you have to do is look at the theoretical numbers and see how close to each other they are. People nitpicking about sets when in THEORY they could POSSIBLY gain 100 dps or less. Fantastic.
I miss more hands-on comparison of DPS. People doing the fights and sharing experiences. What works where, etc. Maybe I'm on the wrong forum for that tho
Not the wrong forum, but the wrong thread. This is the FBN threat. There's both a general cat thread and a WWS thread.
Having read your DPS guide (as the only insight into your play-- I might look for parses later), you do have a bit of error in your playstyle-- at least at the time, you didn't FB, which can be a significant DPS increase. I also imagine that even if you've started since then, you may not be doing it optimally. I imagine you execute what you think is right close to perfectly, but RNG randomness may mislead you as to when you are playing well versus when you are playing poorly.
Moving onto more concrete things. You don't take Feral Aggression-- you can say the DPS isn't worth the talent points, but it will be another 100 DPS. You also didn't max out SotF, so you're missing strength and agi-- more DPS you've discarded for Nurturing Instinct, swipe, and improved mangle-- all things with utility, all tradeoffs not done for the single-target simulations you are comparing to. If you're responsible for keeping mangle up, that's another few hundred DPS lost.
In sum: 100-150 DPS in talents. 250 DPS from mangling. 3 upgrades, which at this gear level are 50-100 DPS each-- let's put that at 200 DPS. Add all that to your Jaraxxus total, and that's 600 DPS-- enough to get you within 5% of the estimates, and we still haven't even accounted for playstyle/user error. I'd say the estimate is pretty accurate.
And that's without getting into relative error-- if your criticisms are for the damage cycle rather than assumptions of always having buffs up, or damage estimates of attacks when those buffs are up, then you're dealing issues that affect every stat. In some cases you're trading a small portion of damage from one yellow attack to another, which might add some small error; but overall an overestimate of absolute DPS is usually a slight overestimate.
In other words: saying ArPen is 10% better than Agi (for instance) is something you can claim reliably. Saying ArPen will produce exactly 1.61 DPS per point may not be entirely true, but general-case accurate.
Finally: you're right in that there isn't a lot of difference no matter what you choose-- that is Blizzard idiot-proofing the gear metagame. But figuring out why these things are true simply verifies that the sim is doing the right thing overall. Although, remember all the times that you missed downtime on your trinket during a knockback/run-out period/etc? Those are much more likely to occur, and since trinket procs are more likely to happen at the start of an attack sequence, trinket uptimes are actually higher in most interruption-based fights.
*sniped out huge post of meaningless gear lists and numbers*
*One additional Socketslot on the Hands with 1 Delicate Cardinal Ruby
Firstly Blacksmithing gives two additional sockets over the belt buckle everyone can use, hands and wrists.
Secondly this really isn't the thread to be posting Rawr results in, this thread is about the FBN Simulator. Rawr is not a simulator, it uses a model and spits out a result. Given the limitations of a model it comes up with surprisingly similar results to the simulations which indicates that the model is fairly good.
So your Rawr results aren't simulations and the huge lists aren't really of any use as anyone, who isn't lazy, can fire up the program and choose the highest item in each slot and get similar results. It isn't adding anything of value.
Note. For balancing sets around the hit and expertise caps the [Solar Bindings] from Algalon is the best piece to use as you end up only one expertise over the cap. They are also the smallest DPS drop from the BiS item for the three places you can easily do this at about 45dps less than the BiS bracers. Using [Soul-Devouring Cinch] is about 120dps less and [Acidmaw Treads]/[Acidmaw Boots] is about 100dps loss but probably the easiest piece for most of us to get to balance the stats.
Tun the problem with a discussion like what you are after lies in the variance of the players involved, not just the cats too as the raid DPS affects the kill times which affects how much of the fight is spent under large buffs like bloodlust/heroism.
As for why there is no listing of [Banner of Victory] and one of the Deaths trinkets, it's probably because the author of this simulator has been busy with real life and nobody else has spent the time to test it. Having said that there is nothing stopping you from downloading the code and running the simulator to get the answers you want, if you do decide to do this and post the results you get it might be of interest to other people.
Edit. Allev you are making assumptions that his current spec is what he is always raiding with, last time he brought up getting to the hard cap for ArP I had a look at his gear choice on armory and he was running a build with five points wasted in Feral Aggression. I call them wasted as I really don't think they are worth the damage gain over the utility talents, if Survival Instincts saves you once a raid it is better than Feral Aggression.
I agree with his current build being strange though, just adding the missing point from SotF in Rawr ups the damage by 80 points. Only two points in Furor would drive me up the way, I still don't like only having three points in it but I think 10% bonus incoming heals is more useful than a bit of frustration about starting energy which is mostly noticed while solo.
Last edited by Cluey : 09/30/09 at 12:49 PM.
Reason: Allev had replied as I did.
A problem which you might not have factored in is that the main dps gain from Engineering (Rocket Glove enchant on hands) shares cooldowns with on use trinkets.
Whilst this won't be an issue with bis gear because they are passive procs, I'm currently using [Mark of Supremacy] to maintain hit caps and I've noticed a severe reduction in the rocket gloves effectiveness when the trinket and glove cooldowns clip each other.
A minor note I know, but one worth mentioning incase others were not aware.
Not the wrong forum, but the wrong thread. This is the FBN threat. There's both a general cat thread and a WWS thread.
Having read your DPS guide (as the only insight into your play-- I might look for parses later), you do have a bit of error in your playstyle-- at least at the time, you didn't FB, which can be a significant DPS increase. I also imagine that even if you've started since then, you may not be doing it optimally. I imagine you execute what you think is right close to perfectly, but RNG randomness may mislead you as to when you are playing well versus when you are playing poorly.
Moving onto more concrete things. You don't take Feral Aggression-- you can say the DPS isn't worth the talent points, but it will be another 100 DPS. You also didn't max out SotF, so you're missing strength and agi-- more DPS you've discarded for Nurturing Instinct, swipe, and improved mangle-- all things with utility, all tradeoffs not done for the single-target simulations you are comparing to. If you're responsible for keeping mangle up, that's another few hundred DPS lost.
In sum: 100-150 DPS in talents. 250 DPS from mangling. 3 upgrades, which at this gear level are 50-100 DPS each-- let's put that at 200 DPS. Add all that to your Jaraxxus total, and that's 600 DPS-- enough to get you within 5% of the estimates, and we still haven't even accounted for playstyle/user error. I'd say the estimate is pretty accurate.
And that's without getting into relative error-- if your criticisms are for the damage cycle rather than assumptions of always having buffs up, or damage estimates of attacks when those buffs are up, then you're dealing issues that affect every stat. In some cases you're trading a small portion of damage from one yellow attack to another, which might add some small error; but overall an overestimate of absolute DPS is usually a slight overestimate.
In other words: saying ArPen is 10% better than Agi (for instance) is something you can claim reliably. Saying ArPen will produce exactly 1.61 DPS per point may not be entirely true, but general-case accurate.
Finally: you're right in that there isn't a lot of difference no matter what you choose-- that is Blizzard idiot-proofing the gear metagame. But figuring out why these things are true simply verifies that the sim is doing the right thing overall. Although, remember all the times that you missed downtime on your trinket during a knockback/run-out period/etc? Those are much more likely to occur, and since trinket procs are more likely to happen at the start of an attack sequence, trinket uptimes are actually higher in most interruption-based fights.
I have 3/3 SotF, of course. Armory must be buggy. I've got 1/2 imp lotp and 3/3 sotf. I dont weave in FB because its gonna give a lot of downtime on Rip. Honestly, I've just not tested if FB is better or not. I doubt it tho, seeing as my crit is severely gimped with all out ArP socketing, making CP generation a pain as it is. Until I see you prove me wrong, I dont believe using FB in my gear is a DPS increase. Thats what my experience tells me, anyway. I dont mangle on raids either, but since I dont take FA there's nothing else to spend the points on. My "concern" is simply that you cannot trust these. Even with insane luck the theorized maximum DPS these programs generate aren't realistic, and I just dont see how you can trust them if they aren't accurate. Dont get me wrong, I'm not saying I play perfectly, I'm merely remarking on the fact that I have very close to the BiS gear on most peoples lists and I'm still 1k+ dps behind these simulations. As a matter of fact, I dont know of a single druid in the entire world who has been able to prove that the simulations aren't totally off the mark. That, in turn, hurts the simulations credibility in my eyes.
Besides, I still haven't seen a parse with Banner of Victory. All the people claiming to be testing "hardcap ArP" are at best around 80% passive ArP, which isn't good enough.
I miss more hands-on comparison of DPS. People doing the fights and sharing experiences. What works where, etc.
That's called anecdotal evidence and it's basically crap. There's far too much variation from fight to fight based on randomness and nobody runs these fights hundreds of times to average it out. That's why we theorycraft.
I have 3/3 SotF, of course. Armory must be buggy. I've got 1/2 imp lotp and 3/3 sotf. I dont weave in FB because its gonna give a lot of downtime on Rip. Honestly, I've just not tested if FB is better or not. I doubt it tho, seeing as my crit is severely gimped with all out ArP socketing, making CP generation a pain as it is. Until I see you prove me wrong, I dont believe using FB in my gear is a DPS increase. Thats what my experience tells me, anyway. I dont mangle on raids either, but since I dont take FA there's nothing else to spend the points on. My "concern" is simply that you cannot trust these. Even with insane luck the theorized maximum DPS these programs generate aren't realistic, and I just dont see how you can trust them if they aren't accurate. Dont get me wrong, I'm not saying I play perfectly, I'm merely remarking on the fact that I have very close to the BiS gear on most peoples lists and I'm still 1k+ dps behind these simulations. As a matter of fact, I dont know of a single druid in the entire world who has been able to prove that the simulations aren't totally off the mark. That, in turn, hurts the simulations credibility in my eyes.
Besides, I still haven't seen a parse with Banner of Victory. All the people claiming to be testing "hardcap ArP" are at best around 80% passive ArP, which isn't good enough.
So you are sacrificing all your agi for arp and you are arguing that FB is not worth using? Seems kind of counter-productive to make such a statement. If you were gimping yourself to socket all ArP why would you purposefully sacrifice abilities that take advantage of all that ArP in favor of bleeds that do not? Sounds like you're shooting yourself in the foot. Also, are you suggesting that the Banner of Victory is what makes or breaks a static ArP set? I don't really understand where you are getting your theories from, though I do agree 80% passive ArP is not good enough to warrant gearing past the softcap. I believe it's in the range of 85%.
I'll believe the armory is buggy-- it's currently down, which means they probably have something wrong. But at the very least, you're playing suboptimally-- which is reinforcing your idea that the DPS is beyond possible. What evidence would convince you? My experience with people in esteemed positions like yours (DPSer in top raiding guild) is that you won't believe it until you do it, and you don't do it because your crit is low-- yet it's been a DPS increase since T7 gear, let alone close-to-BiS T9.
You could look at a lot of history to see that there's a lot of math done from the lowest level to multiple independently-coded theorycraft tools (FbN, SimulationCraft, Rawr, Toskk, various spreadsheets) who have all determined that there's a DPS increase in FBs. It's not just slaves to a single outlier tool who are claiming it. I'll separate this for emphasis:
Everyone who has done the math has found a damage increase with FB. No one who has come to this conclusion has had reason to change their mind based on their experience.
From the short napkin math (i.e. it clearly does more damage than a shred) to the longer napkin math (that damage exceeds three ticks/six seconds of rip downtime), from more basic estimations (every player gets crit streaks after a finisher at least once in a while) to as-realistic-as-they-come simulators (and like I said, independently coded and verifiable-- you can even look at the exact attack sequences they generate if you feel like exploring that much). Every single one agrees.
Now, you can make a variety of claims to mitigate this point-- like, it's a DPS loss as you're learning because you're likely to mess up, or it creates variable DPS streaks, or the DPS gain is small (like 5%), makes the cycle unnecessarily complex, or that collisions in various timers while mangling can minimize the benefit. But it's tangible, provable (and proven), accurately simulated and verifiable to be more DPS to use FBs.
My justification previously is that, given you're about 10% off right now, 5% can be chalked up to talents and gear, and much of the rest can be chalked up to playstyle and execution. Nothing you've said negates that.
Are there sources of error still? Yes. Assumed 100% uptime on buffs from other raid members is one of them-- just like most don't maintain 100% mangle uptime. I'm pretty sure FBN assumes you can refresh rip/rake as soon as they expire to the hundredth of a second, whereas players couldn't (I think SC has a corrective factor here as well, but I'm not positive).
I could go on and on about where you can attribute these errors (it's mostly in the execution rather than the formulas), and what kinds of errors the over-estimate or "upper bound DPS" will mean in terms of actual stats, but I don't think that will convince you either. I could make a reasonable approximation of your attack cycle/talents/gear and drop it into SimulationCraft and show you that you're closer than you think.
But really the point is, you don't care enough to figure out why it's wrong, so why should I care enough to show you that you're also wrong? That hurts your credibility.
Here is a parse you'd like-- the player has 2 pieces of 258 loot and did 8900 DPS. He's still using 2t8, and is using an Executioner enchant on his weapon. He got a little lucky on Shred crits, but unlucky with Rake applications. Note also that even he doesn't have full raid buffs all the time-- only 91% on haste buff, for instance.
That's called anecdotal evidence and it's basically crap. There's far too much variation from fight to fight based on randomness and nobody runs these fights hundreds of times to average it out. That's why we theorycraft.
So what's the point in basing theories on something you wont achieve anyways? If you have to average it out over 100 tries to get the results you're simulating, it's in my eyes the same as saying you shouldn't argue about the +/- 100 theoretical DPS between the BiS sets. Since, after all, you wont do a fight enough times for one of the sets to statistically be better than the other. Or did I miss something? You're simulating it 100 times in a program so you don't have to do it ingame? Any given attempt on a boss could be better, and as you say yourself, you dont get a lot of kills on a given boss, so who knows what set would've actually won in your game. We only know which set won in your program.
Originally Posted by ithecho84
So you are sacrificing all your agi for arp and you are arguing that FB is not worth using? Seems kind of counter-productive to make such a statement. If you were gimping yourself to socket all ArP why would you purposefully sacrifice abilities that take advantage of all that ArP in favor of bleeds that do not? Sounds like you're shooting yourself in the foot. Also, are you suggesting that the Banner of Victory is what makes or breaks a static ArP set? I don't really understand where you are getting your theories from, though I do agree 80% passive ArP is not good enough to warrant gearing past the softcap. I believe it's in the range of 85%.
I am sacrificing the extra agi from socketing so I can get the 87% passive ArP I have now. At this point my crit is pretty low tho, so Im saying that if I start using FB as well I will have a lot of trouble keeping SR and Rip up meanwhile. Losing out on too much Rip damage is an overall DPS loss, and I don't have a lot of time between rotations as it is, so if I weave in FB now I will certainly go down in overall DPS. That combined with the fact that a lot of the times you're sitting with 10 sec left on SR and ~10 on Rip as well. You end up there quite a lot, at least I do with my current CP generation. Try getting out of that situation with 55% raidbuffed crit.
Anyways, with Banner I'm at 87% passive. Without Banner I'm at 80%. People say that you need 85% passive before its worth more than soft cap, yet NOBODY has so far tried a set with the needed 85% passive ArP. Begs to question the quality of simulations.
I'll believe the armory is buggy-- it's currently down, which means they probably have something wrong. But at the very least, you're playing suboptimally-- which is reinforcing your idea that the DPS is beyond possible. What evidence would convince you? My experience with people in esteemed positions like yours (DPSer in top raiding guild) is that you won't believe it until you do it, and you don't do it because your crit is low-- yet it's been a DPS increase since T7 gear, let alone close-to-BiS T9.
You could look at a lot of history to see that there's a lot of math done from the lowest level to multiple independently-coded theorycraft tools (FbN, SimulationCraft, Rawr, Toskk, various spreadsheets) who have all determined that there's a DPS increase in FBs. It's not just slaves to a single outlier tool who are claiming it. I'll separate this for emphasis:
Everyone who has done the math has found a damage increase with FB. No one who has come to this conclusion has had reason to change their mind based on their experience.
From the short napkin math (i.e. it clearly does more damage than a shred) to the longer napkin math (that damage exceeds three ticks/six seconds of rip downtime), from more basic estimations (every player gets crit streaks after a finisher at least once in a while) to as-realistic-as-they-come simulators (and like I said, independently coded and verifiable-- you can even look at the exact attack sequences they generate if you feel like exploring that much). Every single one agrees.
Now, you can make a variety of claims to mitigate this point-- like, it's a DPS loss as you're learning because you're likely to mess up, or it creates variable DPS streaks, or the DPS gain is small (like 5%), makes the cycle unnecessarily complex, or that collisions in various timers while mangling can minimize the benefit. But it's tangible, provable (and proven), accurately simulated and verifiable to be more DPS to use FBs.
My justification previously is that, given you're about 10% off right now, 5% can be chalked up to talents and gear, and much of the rest can be chalked up to playstyle and execution. Nothing you've said negates that.
Are there sources of error still? Yes. Assumed 100% uptime on buffs from other raid members is one of them-- just like most don't maintain 100% mangle uptime. I'm pretty sure FBN assumes you can refresh rip/rake as soon as they expire to the hundredth of a second, whereas players couldn't (I think SC has a corrective factor here as well, but I'm not positive).
I could go on and on about where you can attribute these errors (it's mostly in the execution rather than the formulas), and what kinds of errors the over-estimate or "upper bound DPS" will mean in terms of actual stats, but I don't think that will convince you either. I could make a reasonable approximation of your attack cycle/talents/gear and drop it into SimulationCraft and show you that you're closer than you think.
But really the point is, you don't care enough to figure out why it's wrong, so why should I care enough to show you that you're also wrong? That hurts your credibility.
Here is a parse you'd like-- the player has 2 pieces of 258 loot and did 8900 DPS. He's still using 2t8, and is using an Executioner enchant on his weapon. He got a little lucky on Shred crits, but unlucky with Rake applications. Note also that even he doesn't have full raid buffs all the time-- only 91% on haste buff, for instance.
What I'm simply saying is, based on the CP generation I'm seeing in my current gear, it's not possible to weave in FB's unless I sacrifice 10 seconds of Rip. I did not know the exact number of Rip ticks was 3. However, getting an actual number of 6 seconds into my head just makes me believe even more that it's not viable weaving in FB with full ArP gemming/gearing. That's not me being stubborn, that's me looking at what happens ingame. I haven't seen the math youre refering to (big surprise there), but I'm fairly sure it's based on more crit than what I have. If you look at pure numbers its easy enough to figure out that FB is better than a little downtime on Rip, yeah, thats fine. But when you mention the actual breakpoint, I say its not practically possible to stay within your parameters. 6 seconds without Rip.. Thats 6 seconds of energy ticking in an empty bar, and I've got 55% crit, and I need to crit on 3 of my next specials...
About your reply to me-- your first statement is wrong, and you only need to look at the parse I linked to verify this-- he's not playing perfectly, and trades 4 bites for his 34 seconds of downtime... You say "that's almost 9 seconds each", but 6 is unavoidable due to startup, he probably FBed/shredded at the end instead of renewing rip for another 6 seconds, and 1s intervals between Rip applications is another 5s, so we're at 17 seconds total skipped for 4 FBs, which is 4.25 seconds per FB.
And he has worse gear than you, with similar crit.
That's not me being stubborn, that's me actually looking at what happens ingame, when people try it.
Originally Posted by Tunz0r
Anyways, with Banner I'm at 87% passive. Without Banner I'm at 80%. People say that you need 85% passive before its worth more than soft cap, yet NOBODY has so far tried a set with the needed 85% passive ArP. Begs to question the quality of simulations.
Search the first post for "Banner" for the result that you claim doesn't exist. And then, realize most of the posts you've probably looked at show better numbers than that, meaning that 45 or 58 item levels does, in fact, make a difference in your second trinket slot.
About your reply to me-- your first statement is wrong, and you only need to look at the parse I linked to verify this-- he's not playing perfectly, and trades 4 bites for his 34 seconds of downtime... You say "that's almost 9 seconds each", but 6 is unavoidable due to startup, he probably FBed/shredded at the end instead of renewing rip for another 6 seconds, and 1s intervals between Rip applications is another 5s, so we're at 17 seconds total skipped for 4 FBs, which is 4.25 seconds per FB.
And he has worse gear than you, with similar crit.
That's not me being stubborn, that's me actually looking at what happens ingame, when people try it.
Search the first post for "Banner" for the result that you claim doesn't exist. And then, realize most of the posts you've probably looked at show better numbers than that, meaning that 45 or 58 item levels does, in fact, make a difference in your second trinket slot.
Right. 70% crit on Shred. Very useful parse. Anyone who read that weaving in FB is more DPS can try it, that guy just got lucky on CP generation and he pulled off some pretty good damage. We've all tried something similar I suspect.
Just goes to show how much you can use DPS simulators for, am I right?
He only crit 36.3% on Rake applications, leaving his net at 61%. Raw, that's 30/49. 27/49 is 55%. You can't convince me that those 3 CP would mean the loss of 4 FBs.
Edit: Here is me and my guild playing terribly -- I get ~50% crit on CP generators. 5 bites in 3:41, only losing roughly 27 seconds for 5 bites, although it's much lower. True it's a low bite rate and 4.3% of my damage, but assuming 1s downtimes between rips and 6 seconds at the start, I'm losing only 13 seconds-- 2.6 seconds per bite.
As for damage, there are lots of reasons for me being at about 7k-- 85%-95% on some key raid buffs, our Ret pally died early taking some buffs/debuffs with him, and I'm undergeared. I also didn't play well (see the low mangle uptime.)
Yeah you definitely have the numbers on your side, I guess. Can't really argue for or against something I've just never done. I just dont understand your parse. How can you have 23% Shred damage done? There must be something secret that I've totally missed. You seem to want to do FB's at the expense of Shreds. The only DPS parse I have at hand is from our first Jaraxxus hardmode kill (which was 4 minutes and 47 seconds), and it breaks down like this:
That was before I even tried hardcapping ArP. I dont get how you can do so few Shreds? I'll try getting some more parses tho, I think Nopher has them stored away somewhere.
Remember tho, the first parse you linked was a 2½ minute normal mode kill with lust. Under similar circumstances I push around 11k DPS if I get a little RNG with me. Also, you might wanna get some epic gems and a new enchant on your staff. I guess.
Here is a parse of myself on Normal Jaraxxus from the first week of ToGC opening. My gear was mid T8 at the time due to being a tank first and foremost. World of Logs - Real Time Raid Analysis
As you can tell just looking at the numbers, I played very sloppy, only having 72% rake uptime and 78% rip uptime. The fight was 93 seconds long, and rip was only up for 76 seconds (38 ticks). That leaves 9 ticks of missing rip. Even if they all crit (~4600 dmg), it would have been 37k damage added. I did 78k damage from FB alone during that fight. I barely maintained a 42% crit rate on my combo point building abilities as well. In fact, mangle was my only saving grace as everything else was critting at roughly 30%. Using FB is easily a DPS increase.
ramenchef, don't forget that you're losing shred damage when you FB as well-- that energy doesn't disappear.
But you also only gave up about 3 ticks of Rip if you consider that you have time between rip expiring and reapplying which you don't get back even full-time ripping, and you can't possibly apply a 5CP SR-buffed rip in the first 6 seconds of a fight (and only after 6 GCDs if you're lucky and saving Berserk).
09/30/09, 3:34 AM
