The State of Assassination

[Deleted]

100% bot procc, when u cast kb. problem solved.

[milkbonez]

It is completely irrelevant how many useless Assass rogues there are. The display of the statistics is bound to parse brackets. In the 99% bracket you will have no rogues that have a 60% rupture uptime, if there even is such a rogue outside of LFR.

Point I'm trying to make is about sample size, which does in fact matter. Your point about brackets is true, but there's also a reason they show us the number of parses. You can't do any kind of real statistical analysis when the sample sizes are so skewed. Comparing 10k parses to 800 or 500 and then declaring them as "facts" is a bit misleading...I really wonder what outlaw and sub would look like at 10k parses.

[ScrubRouge]

75% https://www.warcraftlogs.com/statistics/11#
99% https://www.warcraftlogs.com/statistics/11#dataset=99

Am i interpreting the data correctly — assa is the WORST rogue spec on live atm, right?

Technically.... yes? but Sin is very consistent (for performing at 75%) . Outlaw's TB RNG kinda got fixed with the new gold trait. Subt has always been solid with minimal rng element.
Assuming best legendary for all specs and perfect execution. Subt pulls ahead of Sin and outlaw. However, if RNG is amazing for sin or outlaw, they pull ahead.

At this current moment Sin is still the best out of the 3 specs for burst phasing, allowing them to pull out huge burst even without BoT proc.

If you want a very consistent high dps output through the whole duration of the fight, then Subt is the way to go. (assuming you can perfect subt -- not saying its some impossible task)

[Faildevil]

I burst a lot higher with Assa then with Sub. Sub is super consistent, with better cleave and AOE options.

[Won7on]

I agree with pretty much everything you said, though I don't think Poison Bomb on-use is the necessarily the best solution. Too bad the devs won't see it because it's on the EU forums. (Also I can't upvote/reply for the same reason)

Also, in general, when you're communicating feedback to developers, I don't think it's worth it to give them solutions - coming up with solutions is their job. Tell them the problems, how they make you feel, why X and Y are not fun/not satisfying/feel weak, etc.

But yeah, I feel you. Assassination was awesome in 7.1.5. I felt like I was in full control, and the rotation felt elegant and satisfying. 2MA + Boots + CoF lined up Vendetta and KB almost exactly and that felt good. The new gold messed all of that up though (and the other new traits didn't help), and I just don't feel like the spec is as fun as it used to be in 7.2.

[Umbrex]

I agree with pretty much everything you said, though I don't think Poison Bomb on-use is the necessarily the best solution. Too bad the devs won't see it because it's on the EU forums. (Also I can't upvote/reply for the same reason)

Also, in general, when you're communicating feedback to developers, I don't think it's worth it to give them solutions - coming up with solutions is their job. Tell them the problems, how they make you feel, why X and Y are not fun/not satisfying/feel weak, etc.

But yeah, I feel you. Assassination was awesome in 7.1.5. I felt like I was in full control, and the rotation felt elegant and satisfying. 2MA + Boots + CoF lined up Vendetta and KB almost exactly and that felt good. The new gold messed all of that up though (and the other new traits didn't help), and I just don't feel like the spec is as fun as it used to be in 7.2.

Thx for the input

I read some of your other posts and u were in the same boat as me.

Yday i cracked it. I have spent alot of hours on dummy trying to optimize my rotation and getting back to orange logs consistently.

I was going through logs when i noticed, the ones with high logs and low procs, were the ones holding kb for vendetta ...this indtantly helped parsing 95 thereabouts consistently for our HC alt split runs.

Do yourself a favour, smack a dummy and compare ...for me it was 50-100k dif in fav of holding kb for vendetta.
It also solves the rotation issue since pooling for kb doesnt make it feel so clunky. Solution, yes - but almost completely negates our golden trait.

The only way it makes sense to me is in tos without cof and 1 ma relic and timing vend with kb Every other.

[Won7on]

I've seen top logs that do both - some hold KB for Vendetta too and play with 3 MA relics for min Vendetta CD, some just allow the two to desync and use them whenever. Honestly, I think it might just depend on what relics you get. It just so happens that I've gotten 905/905/915 Vendetta relics, so I believe holding KB for a few seconds to line up with a ~40 sec Vendetta CD is the best option for me.

[Zabatakis]

Let's look at the 75th percentile, that's generally agreed upon as good way to look at how a class is performing.

Eh I do not agree with this. 75th percentile is either mediocre play or really bad luck with procs. 90th percentile is either great play without insane procs that would put you 95-99, or mediocre play with insane procs. So that is where I look. I feel Sin will get a boost since it probably performs better than most specs when played at a mediocre level.

That said, even at 90% where sin drops on all those parses they are more than fine, and well within the dps of top markers.

I do want to point out something tho; You should not look at boss damage on fights like Chrono and see yourself X% above the next guy and think to yourself "This is why they bring me, this is my niche". Plenty of classes would have around that % (or more) if they were told to sit on the boss, look at starbro dmg to boss logs for example. Your RL lets you stay on boss for the various fights not because you are so amazing at it, but because you are so absolutely bad at anything else from a dps standpoint. This is what you should be concerned about. If Sin is going to be the freaking worst by miles at constant target swaps and quick burst on priority adds then it needs to be equally that much better at sitting on the boss imo.

[shaunika123]

bro you are linking 99% parses and drawing conclusions about the whole class based on smallest percent of parses. Let's look at the 75th percentile,.

let's not

nobody cares about mediocre performance
balance should not be done around mediocre performance
it should be done around the best

the fact of the matter is
that no class should be able to even come close to assa single target because no class has worse AOE than us, in fact, most classes can out AOE us in like -40 ilvl deficit

if we're good for one thing and we're not the best at even that, than our spec is mediocre at best.

[Deleted]

Point I'm trying to make is about sample size, which does in fact matter. Your point about brackets is true, but there's also a reason they show us the number of parses. You can't do any kind of real statistical analysis when the sample sizes are so skewed. Comparing 10k parses to 800 or 500 and then declaring them as "facts" is a bit misleading...I really wonder what outlaw and sub would look like at 10k parses.

The sample size does not matter at all. The assumption that Sub and Outlaw parses right now have more quality rogues in them is just plain stupid. Raiders will pick what performs best and Assassination does not perform best right now. 7.2.5 will hit reset on that but in the next couple of weeks Sub might pull ahead even more, since most rogues only now are catching up with artifact power for sub.

[DakonBlackblade]

The sample size does not matter at all. The assumption that Sub and Outlaw parses right now have more quality rogues in them is just plain stupid. Raiders will pick what performs best and Assassination does not perform best right now. 7.2.5 will hit reset on that but in the next couple of weeks Sub might pull ahead even more, since most rogues only now are catching up with artifact power for sub.

Of course the sample size makes a difference, where the hell do you think the brackets came from, warcraftlogs is not simply making numbers up. If there were only 10 Sub rogues parsees, all of them 1M+ DPS on Krosus, every bracket for sub would show 1M or more DPS on Krosus. Meanwhille the Sin brackets wich are being calculated based on 10K parsees, some of which are absolutly atrocious, would mostly be much lower, specialy on the low and mid end portion. You could have 10 Sin rogues who all did more damage than the 10 Sub rogues's parsees and the 90-99% bracket could still say Sub is higher than Sin. That could mean Sub is much better than Sin, or it could mean the 10 dudes playing Sub have the perfect gear/legendaries and are all very good at it, while the Sin parsees have dudes from all levels of itemization and performance because its the most popular spec.

And if you want to see a class/spec performance you look at the middle brackets, because those tend to ignore the luck factor, the only thing the 90% + brackets give an idea of is the maximum potential damange, Sin is lower on that department than other classes and rogue specs but to get to that bracket a lot of crap has to happen in the most perfect way possible for you so it is not realy a measure of anything, specialy not on progression kills where nothing tends to happen in the perfect way for anyone. Even this aspect is iffy because you could have 10-20 dudes that dealt more damage than everyone else on the Subs and Outlaws parsees, but because that when you bunch then up with the rest of the top 90% ppl the average is lower than that seen for the other specs which have averages based on much less ppl the bracket is also lower overall. You see many examples of fights where the top parsee is a Sin and te top 100 has the majority of Sins and yet the 90%+ brackets show Sub and Outlaw ahead. For instance on the top 20 parsees of Star Augur you have 8 sins x 1 Sub and yet the 99% bracket for Sub is ahead of Sin due to a much smaller sample size, heck the top 100 parsees are almost only Sin rogues and yet both Outlaw and Sub are suposed to perform much better on the high % brackets when its evident that Sins surpas both those specs very frequently.

To finish, Sin might end up not being the best rogue spec anymore, but who cares, as long as it is a good spec ppl will play it. Ppl played the hell out of Outlaw and Sub while Sin was the theoretical best and killed every boss there were to kill on top, mid and low end guilds. Same will go for Sin if it ends up worse than Sub on the long run.

[Deleted]

If there was only 10 Sub rogues parsees, all of them 1M+ DPS on Krosus, every bracket for sub would show 1M or more DPS on Krosus.

Nobody cares for every bracket. Did you even read the context of previous posts? Top log statistics will remain the same or become even better when the sample size increases. Who cares for low brackets. Also there are no 10 parses but everything is well above 500 which is more than enough to represent the sate of the spec. It's irrelevant if 10k of 500 parses show the potential of a spec in the form of top logs. So spare me with examples that have no meaning at all.

[DakonBlackblade]

Nobody cares for every bracket. Did you even read the context of previous posts? Top log statistics will remain the same or become even better when the sample size increases. Who cares for low brackets. Also there are no 10 parses but everything is well above 500 which is more than enough to represent the sate of the spec. It's irrelevant if 10k of 500 parses show the potential of a spec in the form of top logs. So spare me with examples that have no meaning at all.

You 1) don't undertsand statistics 2) dind't read what I wrote.

Brackets for Sin show a lower DPS average for the 90%+ range, and yet when you look at top parsees you see the top 200 almost entirely filled with Sins, often times the very top parsee is a sin and yet the 99% bracket is lower for Sins because it includes more ppl than just the very best because everyone and their mother plays Sin. A very crude and simple example: if 1000 ppl played sin and 10 of those dealt 1.2M DPS on boss X but the other 30 dealt less than 1.1M and the rest deal less than 1 M, while only 10 ppl play Sub and all of those dealt more than 1M DPS on that same boss the 90%+ bracket would show Sub ahead of Sin even tough top Sin rogues deal more damage than top Sub rogues. The bottomline here is that Sub simply has too low of a sample to draw any actual conclusion, which is what Im saying. If you tell me Warloks perform better than Sin rogues youre absolutly right, the sample sizes are btoh hughe and locks are ahead, now for Sub and Outlaw the sample is too small, it simply does not work to try to draw conclusions from these numbers.

If you want to know Sub had better 90-99% brakets for almost the entirety of EN, and we all know wich spec was the best at EN.

[Deleted]

I never follow the discussion but also I noticed after playing my rogue now since 5days on max level (AFT 37) that Poison Bomb should definitely a castable spell. If I want RNG I could play Outlaw.

[Deleted]

I understand statistics perfectly and I read your nonsense too. I played outlaw in EN because it was king at first. When they hotfixed the shit out of it many switched to Assass bc it was able to put out acceptable numbers even without legendaries. Most people could not just roll Sub because it was useless without Satyr's Walk. Only now we are heading in the direction that people can have the required legendaries for all specs to just switch to another spec.

[Deleted]

Not to sound like a jerk, but who told you that you understand statistics? You may be a smart guy but statistics is often counter-intuitive. Do you have formal knowledge about the subject?
Take a look here: https://www.dynatrace.com/blog/why-a...les-are-great/
And then maybe here: https://stats.stackexchange.com/ques...es/31330#31330

You still think mean across different percentiles over a sample size of over 5000 (60,000 for assa) is not representative?

[DakonBlackblade]

I understand statistics perfectly and I read your nonsense too. I played outlaw in EN because it was king at first. When they hotfixed the shit out of it many switched to Assass bc it was able to put out acceptable numbers even without legendaries. Most people could not just roll Sub because it was useless without Satyr's Walk. Only now we are heading in the direction that people can have the required legendaries for all specs to just switch to another spec.

OK, this has jackshit to do with what I said. What I said, and explained very well actualy, is that Sub and Outlaw rigth now does not have a big enought sample size to come to any conclusions in regards to its actual performance. The numbers could be skewed either way, the spec might turn out super awesome and when we have a large sample size it could show that its the best rogue spec by a mile, it still doesn't change the fact that right now the statistics for this spec (and outlaw) are not trustworthy, and that when you actualy break down the numbers and go trough the raw data you see Sin performing as well or arguably better than Sub. The reasons why ppl were not playing this or that spec, or the reasons why they might do so now is not relevant to this discussion, what matters here is that trying to say Sin is in some kind of danger based on untrustorthy statisics just doesn't fly. When we have a large enought sample size we will see.

Also none of this realy matters cause 7.2.5 will come and just take a massive dump on all the class balance (again).

[milkbonez]

Eh I do not agree with this. 75th percentile is either mediocre play or really bad luck with procs. 90th percentile is either great play without insane procs that would put you 95-99, or mediocre play with insane procs. So that is where I look. I feel Sin will get a boost since it probably performs better than most specs when played at a mediocre level.

It's...ridiculous how much people obsess over parses. Most of the orange parses are gonna be by guilds that are much more geared than other guilds logging and have the content super farmed, so their raid wide dps is veerrry high, which means shorter fights, the ability to now cheese abilities, and overall cheesing of logs. That's not to say the people who parse high are "lucky" or not good, quite the opposite. They were the first to clear the content and are committed to that, but it's only natural they're gonna parse super high on most occasions. No doubt RNG is a factor, especially when you see some of the procs on NBF...but 75th percentile, you know how to play your class well, that's pretty acceptable in my book....You see people asking why they parsed so low on kill, and their dot uptimes are all good, surge uptimes are good etc., but 1/3 of their dps died and the fight took an extra 30 seconds and their CD's didn't line up perfectly again. They are doing everything right, they are just not in those guilds that get super fast kill times that inevitably leads to a higher parse.

I feel like i put my $.02 in for this thread, i'm not gonna argue about fucking sample sizes and other obvious things. People can roll outlaw, hell it's easier now to switch specs imo with the increased leggo drop rate and how fast ak scales for the first few traits.

[Deleted]

Dunno why to stress yourself. I understand, that many here are theorycrafting fans and so on and so forth.
But: The current state of Assa is in a good (enough) spot for clearing NH Mythic. In fact: NH Myhtic is a joke now, compared to pre 7.2
Not only got the damage buffed across the board (including Assa) by a massive amount, they also nerfed Elisandre + Gul-Dan recently. As Assa with 910+ equiped (and BIS Legendaries) and proper setup you can go with 1mio DPS +- 100k (pretty stable numbers). Again: This is more than enough for clearing the current meta!

So bottom line: Wait till 7.2.5 - because this is where things are getting interesting and THEN you can start discussions.

Thats just MY opinion - I dont want to offend anybody =)

[Tyrarolls]

if you watch 95% parses with dmg to bosses we are pretty high overall.
however doing 50k more dps more ST but losing 400k overall/aoe dps is not the trade i like.
Also you often rely on stupid PB procs /bota at the right time to do even the ST which justifies to be very good at single target.

now i do not want to complain - Assa had a good time in NH progress (pre 7.2) and we were one of the most relevant and important specs in Night Hold.
We were needed for the pure ST on Spellblade before the 100 nerfs so we can kill her in time (yes if you killed her in the first ID you had some issues with the Enrage timer)

We were definitely needed at Krosus in the first ID and Tichondrius Singletarget (even if you had to soak) was also super important due the Enrage timer.
On Elisande our Dmg in P1 was needed (you could do 100+mio without Shoulders) and for Guldan we were soak bots but had good dmg in the 10% execute.

Noone can complain about Assasination in Nighthold.

This said - it is obvious that with 7.2 we fall behind sub and outlaw in terms of overall dmg and also sometimes ST - just for simple reasons:
Sub has the best new Traits and Outlaw has decent traits but always shines with max parses during nerfed farm content - also with more people getting Shoulder + Bracer finally you can even escalate more with 2-3 good rolls in the short fights.

I was always confident that Assa will stay relevant as Progress spec just cause of AP and with removing AP i doubt blizzard will realize what it takes as replacement to make Assa relevant. But let's see.

I got Assa currently on 49 and Outlaw nearly on 48 - even if we have to play Sub i m fine with that but in the end i dont even bother playing my DH if shit happens...

The only issue i have is that we dont have any information which spec we can prepare but well...we will see

Overall i really dislike Assa atm just because you just rely on PB procs nowadays...but well w/e