maxpowers

The 2.3.2.5 from http://rogue.raidcal.com

Also, I agree with many of Graecus's points. It's like he said what I was thinking, except more articulately.

p.s. I've recently found that waiting longer for a mongoose / executioner / *Haste proc and then spamming hemo, to be a a fun way to theoretically increase damage. I realize it's pretty minor, but at the same time I find it's more fun than mindlessly spamming. Plus without combat potency you don't need to worry about wasting energy. A slightly modified Quartz is very helpful for keeping track and I can post how to change if anyone is interested.

*Haste and then hope for a Warp Spring Coil proc from hemo for additional white damage gains.

 

Kellhus

I find myself in complete agreement with Graecus (post 1574) -- it's long or I'd quote.

I've personally raided in BT as 0/31/30 sword, 17/42 combat maces, combat swords, combat hemo, and currently 41/20/0 combat mutilate, and I have to say my findings on personal dps are never as profoundly different as the spreadsheets or roguethink(tm) would have you believe -- especially highlighting that I've run without the requiste points in assassination that are so frequently used to invalidate builds from use. I don't know that it's the developers that cause us to immediately invalidate specs when visiting the theorycrafting numbers, moreso than it is our ability to get hung up on mathimatical details that simply don't impact us as greatly as predicted outside of the vaccum of a message board raiding environment.

Anymore, and my personal rant, is that it's almost as if theorycraft has become gospel and started to limit the, at times, goofy ingeniuity of the common player to simply spec and try something for no better reason than trying it out of fear of less than optimal spreadsheet dps. This very thread is cited many times per day on multiple rogue sites as the reason to change X in a talent or gear pool, without the suggestor really understanding the impact of the changes made outside of they must happen in order to do more dps on the spreadsheet.

 

Dontmindme

The short answer is "Yes".

The longer answer. Really, the only significant change from the last version was Sinister Calling. Before that, the change from 125% to 110% for Hemo.

Now, before we got too caught up in the spreadsheet numbers...
The changes I made (as concerning this thread) were the following:
1) Added the new Hemo base damage to Sinister Calling.
2) Disabled the Hemo debuff / Hemo debuff estimate double dip.
3) Eliminated the 100% uptime Hemo debuff in favor of an estimated 1/7th (based on 0.5 sec uptime every 3.5 secs) with an option for multiple Hemo rogues in the raid.

What's as important is what didn't change:
1) Shadowstep is modeled exactly as it was.
2) The Hemo Debuff Estimate is modeled exactly as it was.

This is significant because I believe both of these are modeled as estimations. With regards to Shadowstep, it looks to just add 20% of a finisher damage every 30 seconds. Unless your cycle happens to be exactly 30 seconds or some fraction there of, you are probably not quite achieving that. You might make up some of that with increased movement, but the sheet is not intending to model that. So, really Shadowstep in modeled in a best case way. I should maybe point out that I didn't see where Shadowstep modeled in the spreadsheet was taking out the 10 energy used. It may be in there, but my quick run through did not find it. So it might be overvalued by that as well. I plan to revisit that at some point, but until I do, it's stuck with this estimation. But really, Shadowstep itself doesn't help very much; I think it gave me 7 buffed DPS in testing. I think the biggest contributer to DPS going toward Shadowstep is Sinister Calling. This brings your Hemo base up near where it was before this recent nerf and on top of that gives a very nice boost to agility. In many ways, I feel that's what Vitality should have been for being so far down the tree.

I should also point out that if you are using Shadowstep in your finisher cycles every 30 secs, you are probably not also using it to increase your movement for added DPS. You are not going to stop at 4 combo points before having to move out of an aoe or a whirlwind in order to pop back in for your boosted finisher after its over. A typical Combat or Hemo rogue is going to pop a Rupture on the target prior to moving out. Under this scenario, combat/regular Hemo is applying damage during the downtime while Shadowstep is sitting on it. And if you are applying Shadowstep on instants, you are spending 10 energy for 20% of an instant. So unless you are backstabbing (the only 50+ energy instant attack), you are ultimately losing damage applying it to an instant for the benefit of movement rate. Maybe the movement makes up for the lost damage, but this conundrum isn't going to be properly reflected in the spreadsheet.

Now we come to the Hemo Debuff itself. This is where it gets tricky. No one has really figured out how much actual damage it actually provides. This, of course, is partly because of the non-physical damage bug and partly because of it's complexity. As I understand it, when working correctly, the debuff should add 42 base damage. This damage is reduced by armor, but multiplied by crits, skill multipliers (like Murder, Hemo multipliers, Backstab multipliers, other classes surely have similar).

Now, the spreadsheet models this as a raid crit rate of 25%. To be honest, I think that's low. Even at early T5 content I hear of Warriors at 35%, Hunters as high, etc. But, it also seems to be using a rogues own armor values. So it's probably skewed high given Serrated Blades and skews high if one equips a lot of -armor gear. On top of that, I don't believe it takes into account at all the possibility that many of the attacks might be specials which have multipliers that may be greater than 100%. So we have crit rate and special attacks not modeled skewing it low and personal -armor effects that is skewing it high. I chose not to alter any of this yet, because it seems it might come up a wash.

So, how accurate is this estimate? I really haven't a clue. I believe we might have a better idea when the fixed Hemo debuff goes live and people start trying to parse what the real raid crit rate might be (at different gear levels?); how many specials seem to get the benefit vs. auto-attacks; etc.

Hopefully this information helps put into perspective what the spreadsheet is showing.

 

Nurru

I feel like we need more posts from people claiming that the spec they like is the best despite no provided proof or reason and math saying otherwise. It makes for a very enlightening thread.

Did I say enlightening? I mean entertaining.

< Aislinana> Why would it be my job to sleep with vontre? Don't I have standards?

 

Kellhus

Your pointlessly condecending bullshit aside, where was this done?

The only arguement this seems even remotely related to is that people have stated their findings on the dps spreadsheet and in practice are contrasting the idea that combat beats out hemo hands down at top gear levels. I'm not finding where someone says an offspec is the best anywhere -- I can however find them saying that the specs are not, however, as invalid in practice as general concensus makes them out to be.

 

◊ Hanos

Basically what you said is that the math is wrong because you haven't personally observed a difference. From that we can infer that you simply aren't very good and it doesn't matter how you spec, that you play better when you are a spec you think is more fun, or you got some weird RNG luck and got strange results.

This forum is based around hard facts, math, numbers, etc., what you basically said is "Hey guys, based on what I have seen spec really doesn't matter and this math stuff is pointless", give us some numbers that show that raiding as 0/31/30 puts out comparable DPS, explain how you would adjust the cycle, gear, consumables etc. to make it work. Right now the math and everything we know says Combat comes out on top. If you get the full benefit of Hemo, then the net gain from that could be as high as 1% assuming you use every charge, every time, if you don't it is a DPS loss and a 5-7% personal DPS loss.

The people interested in the actual theorycrafting are the ones willing to regem their gear or respec for 0.5% more DPS, if you are fine doing 50-100-200 DPS less then you could do, then this isn't the forum for you and your input would be more welcome on the general forums.

The findings of the DPS sheet will rarely line up exactly with what you get, back in Naxx we had a rogue who should have capped out at 600 DPS break 1000 on Patchwerk do to an obscene crit string, it happens, I have also done fights where my glancing rate is off the charts, my crit rate sucks, and I get put of of combat after blowing cooldowns... yeah my DPS was lower then it could have been, does that mean my spec is wrong? No. The spreadsheet models your DPS in an infinitely sustained fight. The truth is that is more beneficial for Hemo and offspecs, because in reality Combat will typically finish closer due to being able to control your burst damage more.


Everyone please stop posting any of the following: Feelings, Observations, Guesses, Napkin math at doesn't cover anywhere near enough possibilities, or discussing the "not DPS value" of talents (if you want to do that you can start a thread about the value of Shadowstep, Dexterity vs Surefooted vs Cat's Swiftness, Imp Sprint vs Imp Evasion vs Blade Twisting etc.... we really don't care). If you don't have number to back it up, we don't want to hear about it.

 

◊ Lrigatonmai

If the model isn't reflecting the real world then there are two possible problems:

-The model is wrong, either the math behind it is incorrect or there are additional constraints that are not included in the model.

-The player (in this case) is not behaving optimally, which is one of the major assumptions of the model.

I'm inclined to believe that it's the second case. Numbers don't lie, people do. As previously noted in this thread, people respec, try a fight, notice they have higher DPS and attribute that higher DPS to the spec instead of the actual cause. It would take many, many benchmarking runs to actually make a comparison that is already laid out with concrete numbers. I realize there is some issue with modeling the actual contribution of the Hemo debuff. But, when you comment about the assassination talents which yield a very concrete (mathematically) DPS increase, it makes it hard to believe that you were playing optimally in all of your test runs.

Edit: Beaten, and with more verbosity. Oh well.

Note: The statement above is probably a lie.

 

hannigaholic

The particular instance I'm thinking of is when a Coilfang Strider spawned on me and instantly oneshotted me while I was killing the Tainted Elemental. There was literally no time to react. Not limited time, not a lack of focus, but no time at all. It happens now and then and cannot be predicted or planned for.

Sometimes shit happens and it's nobody's fault. It's much rarer than silly mistakes (like dying to Cave-Ins or Flame Patches) but it can happen.

I don't in any way think that speccing Cheat Death is worth the dps, both personal and raidwide, that you lose in the overwhelming majority of cases, but it most definitely has, as others have said, a nonzero value.

 

◊ Kalman

If a strider killed you while you were killing a Tainted, something is very flawed in your strategy.

 

hannigaholic

Strange, I read what he wrote and unless I've misread something it sounded more like he said that the spreadsheets don't back up the claims of the pro-Combat camp.

It does seem that people are far more willing to accept feelings and hunches from those who support Combat than from those who support Hemo, even when there is hard maths (IE the spreadsheets) contradicting the pro-Combat 'feelings'.

I don't understand what you're saying here - it seems like you're saying that Hemo builds have high, uncontrollable burst and that somehow this means that the spreadsheets favour them. I was under the impression that, if anything, the spreadsheets favour deep Combat builds because of the elimination of the real world problem of getting wasted Combat Potency procs (2 or 3 in a row when you already have 70 energy for example, because you were waiting that extra couple of seconds for your rupture to run out).

Basically Hemo cycles are inherantly more stable than bulds including Combat Potency, because they don't rely on as many procs, and as such the real world dps will likely be closer to the theoretical dps than for the deep Combat builds.

 

Stabmaster

This is easy to answer:

In general, people speccing Combat for raiding in the past have done so as a result of careful DPS analysis.

People speccing Hemo/Shadowstep in the past have largely proven to be hysterical fanbois who delight in LEET CRITZ in Battlegrounds.

Now, the balance is changing as the Sub tree is being fleshed out by Blizzard, but it's clear that following the advice of someone who has done math is a better idea than someone who gives unfounded opinion based on a limited dataset or Battlegrounds experience, which, to date, has been rampant among most Sub-tree adopters.

Personally I am extremely interested in recent Sub tree improvements, have raided Combat Dagger/Combat Mut/Combat Sword/Hemo Tri-spec in endgame raiding, and am eagerly anticipating any changes Blizz can bring to provide better raiding variety.

 

• Vulajin

There are two publicly-available spreadsheets and a separate one I make for my personal use. Two of the spreadsheets indicate Hemo builds are roughly on par with combat, provided you are the only one in the raid, and that Shadowstep and Cheat Death builds are inferior to a significant degree (several percent). Only one disagrees.

Hemo builds have fewer cooldowns (assuming you don't go 0/31/30 or something), and they are more stable, thus your DPS is much more likely to line up with the spreadsheet projection. A long-run average is fine for Combat Potency because over the course of a ten-minute fight, you are likely to have plenty of chances to proc it, so that the total will come out near the expected amount. On the other hand, fights are nearly always enough that your cooldowns will have a higher increase on your actual DPS than your theoretical average.

 

hannigaholic

Indeed, and yes I was thinking 'Hemo' as being trispec rather than 0/30/31 or x/x/41.

I just felt the need to point out that Hanos seemed to be jumping on Kellhus for simply stating the same thing, seemingly just because Kellhus didn't agree with Hanos that Hemo was hands down not a viable raid spec at top-end gear levels, despite Kellhus' assertion that his information came from the spreadsheets as well as in practice.

 

Ruqas

Edit: removed because it got me an infraction.

Now how she taketh mine eye.

 

◊ Hanos

The 1% gain on a Hemo Spec assumes every charge is used every time, while true most of the time, it isn't true if you ever have to change targets. Also the way cooldowns work means that you will end up with DPS higher then your theoretical more ofter then not with Combat. I jumped on him to some degree for stating his opinions based on how things felt, not based on any math.

 

Ashere

fixed

Very generalizing, I personally couldn't care less about crits, I'm just fond of the playstyle, and getting sick 'n tired of having to defend a playstyle that works perfectly for me, which I'm more comfortable with then with combat, with which I actually perform better then with combat since it keeps me on edge. And I doubt I'm the only one on that. And guess what, most of my time is spent on PvE, I find PvP nothing but a boring grind for actually overpowered equipment, considering the time invested to get it.

 

Jakani

I'm assuming that first 80% comes from the mantra of keep SnD up 100%, finish when possible. The next 10% I'm guessing comes from using good energy cycles and properly gearing/gemming/enchanting. I'm wondering what tips you might have for reaching that last 10%. My guess is it's things like using strategic use of trinkets to maximize their benefit. I've seen hintings of it mentioned, but since they're things that don't model (I imagine), they don't get brought up much. I am very much interested in hitting the peak of my capability, so anything would be helpful.

 

• Scheme

I'm not sure what significant difference there is between the "playstyles" of hemo (with or without shadowstep) and combat, especially in PVE. Sure, there's a slight difference in cycles, generally 5s/5r for hemo and 2-3s/5r for combat, but otherwise they play the same. Hemo builds generally have slightly more powerful openers, I guess. I don't know, I'm reaching. Explain please.

 

• Aldriana

I think you underestimate just how easy it is to get to 90%. Consider: according to the Rogue Gear Sheet, with my gear and reasonable raid buffs, I should be able to do 1586 sustained DPS. If instead of doing optimal cycles, I just spam SS and keep SnD up - just that, and nothing else - I would do 1480 DPS - aka 93% as much as optimal. Hence, to do less than 90% damage, you need to either a) Let SnD drop or b) Let your energy cap out. You wouldn't think this would be hard... but apparently it is for some people.

Of course, there's also the aspect that a lot of people don't understand gear selection in sufficient detail to pick the best available options - particularly since gear selection is often highly unintuitive (the boots from the 2nd boss in Karazhan are as good as the ones from the 3rd boss in Hyjal? What?), which also hinders their DPS. But I admit it has always puzzled me how people manage to screw these things up as badly as they seem to.

 

• Latito

I could not agree with this more. Really, its been one of the biggest reasons for me switching to a 4s/5r cycle (cutting SnD when energy hits 60-80 or SnD is < 1s left.. whichever comes first) when the sheet recomends 3.1s/5r. Sure I could do 3s/5r MOST of the time.. but there are those occasions when it just doesn't work out and you lose SnD for a few seconds. For ~5 dps in the best case.. why bother? Its NEARLY idiot proof and is more stable.. not to mention it jives well with BT exalted trinket.

I think the mantra of "keep SnD up and never let energy cap out" could even use a third point.. "stay on target". That is.. stay within melee range of the boss as much as you can. This may sound like "wtf.. duhh" but how many times do you see a melee class not in melee range of their target in hyjal? Did those average-dps rouges sprint on FLK when going from one sub-boss to the next (better yet.. did they sprint when going between the 2 successive mobs your guild tanks the furthest apart?) Alar is an excellent fight to gauge a rogues "dps + moving" abilities.. especially if your guild puts melee on adds in p2 as well. One of the rogues in my guild can generally do 25% more dmg than other melee-dps which normally are on par with him... his dps time on WWS is leaps and bounds ahead.

As for "keep SnD up always".. I think too many rogues are attempting to cut corners and use a shorter cycle than they should be in hopes for higher rupture uptime. When I look at a parse of Morogrim and see a rogue with 8 SnD gains and only 1 watery grave.. no wonder their dps is falling off a bit (I know.. often you'll get periods of < 1 second with SnD falling. look at total autoattacks and compare though).


Just looking back at a recent Anetheron kill.. I did 104k dmg in Sinister Strikes over 70 attacks. 4 of those were dodged and cost 8 energy.. total energy spent was 2672. This gives me nearly 39 dmg per energy. Add in the 20k rupture damage I got - 8 ruptures for 8 ticks a piece (all of which are free due to relentless strikes) and I get another 7.5 damage per energy. Thats a little over 45 damage per energy.. plus all that crit from BT exalted trinket, procs off of SS attacks, etc. which all increase white dps. Letting your energy tick over 100 can add up fast. Consider than a single tick of Combat Potency when you are already capped on energy is 675 damage lost. Or worse.. ticking over on an AR tick (likely the most common time to go over energy).. 1800 damage lost for a full tick of nothing. Granted, gaining as much energy as you spend in 2 seconds so as to wind up at 100 energy when the tick happens is already fairly rare (finisher that procs relentless + combat potency ticks + lag). Over the whole Anetheron I did 391k damage.. that 1800 would have been nearly half a percent difference. Thats roughly the same as most modest gear upgrades..
(Latito - WWS if you're interested).


As one of the many many good rogues out there who "get it" when it comes to dps.. it never ceases to amaze me how someone can botch things up so badly as to grab a 1.5 speed dagger and sinister strike the boss with a 2.4 speed offhand mace, no imp SnD, no dual wield spec, no relentless strikes, etc. Not surprising the same rogue has dps times say 20% lower than an AVERAGE rogue, a third the SS's and 2 evis' as the sum of his "non-snd finisher dps". While those numbers and that scenerio may be made up.. its not far fetched. Combat or Hemo - it really doesn't matter if your playing like an idiot. There was a WWS I saw a while ago of a Vashj kill with 5 rogues.. 4 of which we in the ~600 dps or less range. Its expired but yea, bad rogues are out there.

 

Jakani

Fair enough. I guess I was just hoping you might be able to let us in on what you were doing to get that extra 100 DPS on Patchwerk. Maybe that's already been done.

 

• Aldriana

Well, that's the thing. I wasn't doing anything special. The other rogues in my guild were generally hanging close - I may have won a bit more often than expected, but we were all very close. And what was I doing? Simple. I was combat daggers spec, and it was the era before combat potency; hence the optimal cycle was just 5s. So the reason I was beating the other guild's rogues by 100 DPS was literally as simple as 1) Don't let energy cap out and 2) keep SnD up. I said it then, and I'll say it now: I could have written a macro that would outDPS most rogues in that guild (if 5 CPs, SnD; else Backstab). Those two rules are remarkably simple, but would get you 97% of the way there then, and about 90% even now.

In terms of the remaining 10%, the next 3 rules I'd add to the list would probably be:

3) Latito's point about time on target
4) Figure out what your best Rupture-uptime cycle that doesn't violate rules 1 or 2.
5) Use your cooldowns

After that you sort of get into subtle shadings about whether it's better to use cooldowns separately or together (together is usually better), managing procs that you have some control over (i.e. Ashtongue Talisman), and the like - but those are somewhat harder to come up with general rules for and tend not be as large an effect anyway.

I suppose there's also two more pretty fundamental ones that really shouldn't need saying, but I'll say them anyway:

0a) Don't die
0b) Don't do anything massively stupid that risks wiping the raid.

If you master those 7 commandments of roguedom, you'll be realizing the vast majority of your DPS potential.

 

koaschten

I remember taking two rogues from another guild, with which i was running SSC with for the vashj belt, to blasted lands to teach them cycles. They made a jump from lower 700 to 850+ dps unbuffed. They were quite amazed of the 'magic' i had done. Only thing is, all i did was telling them rules to stick to which were a little bit more restricted than those they used. It's a hell lot of a difference between "use slice'n'dice and eviscerate" and the rule-set Aldriana showed up there.

 

Captain Winky

I had someone show up to a Gruul raid with 3/3 Imp Evis, 2/2 Remorseless, and zero points in Malice. True story.

 

Ashere

The main difference between Combat and every single hemo build is (to start with) the fluctuations caused by Combat Potency:
Combat: hit SS when energy >= 40
Hemo: hit hemo exactly once every 3.5 seconds.

Then there's an unobvious quirk with the Dirty Deeds buff. Your damage suddenly gets a very strong boost, how to handle the increased aggro you generate? Over aggroing during the last 35% can be a big issue, you might actually NEED that extra vanish from prep there!

Shadowstep adds something more to the list:
Shadowstep-rupture now, or just rupture and shadowstep after boss did some AOE to get back to him faster?
If I do the last, what attack to use after shadowstepping? Hemo, SS or Ghostly. Which generates most damage/energy, one way or another?

The improved agility from Sinister Calling is also a problem actually. It adds 3~4% to your critrate, and crits are nice, but unreliable when they happen. And crits cause more agro. So you actually have some unreliable mechanics that could raise your aggro a LOT alll of a sudden. The closer you are to the tanks on that list, the better you have to watch it. A boss suddenly turning around to give you a smack isn't that much of a problem with something like Cheat Death. But it's at the cost of the tanks' rage, and he will fight to get control back, wasting rage on attacks because you think it's funny to do the tanking for him.