Unholy DK
Guardian Druid
Restoration Druid
Augmentation Evoker
Devastation Evoker
Marksmanship Hunter
Arcane Mage
Frost Mage
Mistweaver Monk
Windwalker Monk
Holy Paladin
Protection Paladin
Holy Priest
Assassination Rogue
Elemental Shaman
Enhancement Shaman
Affliction Warlock
Protection Warrior
Our Class Writers share their expert opinions based on rigorous testing and theorycrafting in hopes of contextualizing the changes coming in future patches of the War Within. Check out all of our released articles in our Editorial Section.
Wowhead Editorials
Subtlety Rogue in the War Within So Far
Subtlety Rogue was part of both Race to World First Events and, as such, had a great start in both seasons. However, its popularity outside of the World First Race was low. There are multiple reasons for this. We will go over them. Let’s start out with a recap of the Seasons.
Season 1
Nerub-ar Palace:
The Raid tier started out fairly competitive between Rogue specialisations. Each one brought its own set of benefits, which could decide on the best spec for the encounter. The list below highlights encounters, one particular spec would be noticeably stronger.
- Rasha’nan – The encounter started on 60% health, benefiting Assassiantion’s execute niche from the Zoldyck Recipe talent.
- Broodtwister Ovi’nax – The encounter was very add heavy, it allowed for funnel damage, which Assassiantion is outstanding at.
- The Silken Court – The encounter had multiple short phases of vulnerability, the timing lined up well with cooldowns on Subtlety, making it the stronger option for the fight.
- Queen Ansurek – One of the difficult to handle mechanics of the fight was Silken Tomb. Subtlety could help with the mechanic immensely with the more dynamic cooldown management of Invigorating Shadowdust, making it the prefered choice for the encounter.
Going quickly over the list, the most important fights gave Subtlety a slight edge, especially on the mythic difficulty. One specific fight Subtlety had problems on was Nexus-Princess Ky’vez. The fight had a lot of forced downtime and movement, making it difficult to efficiently use your cooldowns. This is also visible on Warcraft Logs, where Subtlety is positioned last.
Mythic+:
Subtlety saw little to no play in Mythic+. This was a result of the changes to Affixes as well as a tuning problem. Competing damage dealers like Assassination were allowed to put out absurdly high AoE damage on top of providing additional benefits like additional crowd control or funnel damage. Survivability was less of a problem, interestingly enough, because both Subtlety and Outlaw bring slightly higher amounts of it.
The mid-Season Update:
The mid season update removed Invigorating Shadowdust, this made sense to reduce rotational complexity. The talent offered incredible amounts of freedom in the rotation but was so difficult to understand that external tools like Spreadsheets needed to be created to lay out the specific use on each encounter, often multiple sequences depending on encounter strategy or kill time. The replacement talent was Death Perception. The talent on its own is extremely boring, but the mid-season update also came with changes to the general rogue tree.
The hightlight of the general tree change was Supercharger, replacing Echoing Reprimand. The talent ended up extremely strong and was a general buff to Rogue. For Subtlety, it increased damage during cooldowns, further increasing burst damage, which in general synergizes well with the design of the specialization. There was also a roughly 20% buff to Black Powder to improve competitiveness in AoE. This is quite a big buff still positioned Subtlety behind competing damage dealers, but made the gap smaller.
Season 2
Season 2 started with one week of pre-season, there was an initial tuning pass done with raid release based on seemingly collected data during this period. Subtlety got hit with a small nerf to Secret Technique.
Liberation of Undermine:
Similar to Season 1, Rogue specs ended up competitive during the initial release of the Tier. Outlaw was even ahead in output statistically, but did not see a lot of play during events like the World First Race due having a worse damage profile.
The list below highlights encounters, one particular spec would be noticeably stronger on.
The Raid design is a bit unfortunate. Chrome King Gallywix is undertuned, which makes spec choice matter less. The hardest boss is for this reason Mug’Zee a fight a competing rogue spec is noticably stronger.
Mythic+
The AoE buffs in Season 1 did help, Subtlety has at least a slightly higher play rate compared to the previous season and can compete slightly better overall.
The mid-Season Update / Current Position:
Rogue, in general, was positioned rather middle of the pack this Season and as a consequence fell slightly behind over time. Outlaw did look like the strongest performer overall before the mid-season tuning changes. The tuning changes buffed only Assassination by roughly 5%, moving it slightly ahead of Outlaw. This leaves Subtlety behind by roughly the same amount Assassination was buffed. Updates to non-rogue classes also moved many of the worse performers up, leaving Subtlety as the statistically weakest damage dealer in the current raid environment, going from Warcraft Logs (excluding: Augmentation Evoker).
The additional increase in power level from gear did make dealing with raid mechanics, especially on earlier fights in the raid easier. Removing reasons to bring Subtlety on fights like Rik Reverb. For what its worth, the gameplay loop is fun but the output is not high enough for most players to justify playing it over competing dps options.
Current Issues for Subtlety Rogue
There are a lot of issues to discuss, i try to highlight the important ones.
Bugs
I need to bring add a bit of historical context to explain the situation. Subtlety not only has bugs, the time to fix even the more annoying bugs is long.
For a bit of context, Subtlety ended up as the worst damage spec in the statistic on Warcraft Logs at the start of Season 4 Dragonflight due to a bug in the update to Warcraft’s combat logging. The majority of damage done by Secret Technique did not have a reference to the player, resulting in overall ~15-20% less damage in logs. The issue was overall harmless because individuals could go and take out the calculator and sum it up manually. The problem took one entire month to fix, making subtlety not only look weak but also adding the extra effort to explain the problem to your fellow raid team. A different, very annoying bug causing delayed combo point refunds was fixed just very recently after celebrating its one-year anniversary. This context is important not only to mention the existence of a long list of bugs, but also the time it takes to get issues fixed.
A short list to highlight some bugs:
Hero talents have bugs related to them too, so more bugs in the Hero Talent section.
Talents
There is a lack of options to specialize on Subtlety. We essentially play the same talent everywhere with minor adjustments. This leaves us very vulnerable to tuning because we can’t add niches or meaningfully change our damage profile. I also don’t mind the current approach if talents that have been unused since the introduction would be replaced or reworked.
A very good example of such a talent is Goremaw’s Bite. The talent was seemingly added to fill the tree but was not part of any talent setup since the introduction. It seems the main reason it was added was nostalgia, but it leaves one of the capstone talents unused. This is also not an exception, the capstone area has at least three such talents.
🟠 – Has very low impact or is extremely situational.
🔴 – Will never be picked.
Complexity, by Design
A lot of Time in design was spent on adding complexity. This started before Dragonflight with things like splitting Eviscerate into multiple instances and adding the dependency of Find Weakness. Design decisions like this are all over Subtlety. Secret Technique is also a good example. The Spell is interesting in design as a three-part attack, but is a AoE attack which requires you to be in melee range, has the second and third attack coded as pet attacks.
To give a more recent example of similar decisions, The First Dance was reworked in Season 1 of The War within to now simply extend the duration of your first Shadow Dance when entiering combat. However, the development team decided this to be too simple and also added a second condition to trigger the effect. You need to be out of combat for 6 seconds.
To finally get to the point. Design decisions like the mentioned do increase the amount of time individuals need to learn the rotation but more importantly, add a lot of dependencies the development team needs to keep track off. Adding a simple buff changes from simply affecting one spell to often three or more secondary spells. It leads to a high chance of introducing bugs, at least half of the bugs of Subtlety are related to secondary effects not working the way they should.
It can also introduce problems when spells behave differently than you would assume as a player. A good example is Secret Technique on the Cauldron of Carnage fight. Using the spell while the two bosses charge, each only hits the one boss you target due to how hitboxes are designed. To make things worse, the second and third attack of the spell is only attributed to the rogue correctly in combat logs when the person logging the fight is in the same group due to the distance between the two groups. And this can be even worse. A example from Mythic+ is the first boss – Kyrioss from the dungeon The Rookery where all clone attacks of Secret Technique dissapear while the boss channels Lightning Torrent.
This problem extends to Hero Talents, which will be discussed in the Hero Talents section.
Raid buffs
Atrophic Poison is a great Raid buff but depends on raid encounters. It still remains good on most fights, but we see an increase in the amount of environmental damage effects in the raid. Any effect that’s not cast by targets you can apply the poison on lowers the effectiveness of the poison and with this, the reason to bring a Rogue to your Raid group. This makes Rogue in general more dependent on tuning than other classes.
Tuning
Rogue is unique in its position as the only pure melee dps class. Specializations are also designed similarly enough to make tuning an important factor in terms of competitiveness and popularity. The problem we face since 3 seasons now is a more one-sided tuning strategy which often leaves Subtlety Rogues left behind.
For a reference tier Subtlety was clearly ahead in tuning; we can look at Vault of the Incarnates:
This chart clearly shows that Assassination was tuned lower than the other two, with Subtlety being the probably best spec of the tier. If we look at the same data from The War Within we get the following:
These pictures are only snapshots as they show, especially for raid a specific point in time and can miss details. However, it is easy to make out balance by simply looking at this data. Going with the assumption that Subtlety is in need of some tuning, we can look at the upcoming tuning changes (as of writing the article). I visualized the buffs to specs in the graphic below:
We see a strange picture, where Subtlety is statistically seen one of the weakest damage specializations but does not get buffs. Now i don’t want to claim that statistics are a golden bullet for tuning, but this situation is very difficult for me to explain. There is no fight in the raid Subtlety does overperform, neither is there evidence for it to have a unique strength to justify the lower tuning.
And the tier did not start off this bad. I don’t understand why Subtlety is not kept competitive to allow more choice between Rogue specializations. And it is a big problem, Subtlety already has a higher learning curve than competing classes and specs, so people who invested time learning wasted their time, the majority switched to Assassiantion, which is now for three consecutiveseasons the best spec in Raiding and Mythic+ or worse switched away from Rogue. It would be a welcome change to have more transparency in tuning to know why situations like this happen, because it is a very frustrating experience for players.
Subtlety Rogue Hero Talents: Trickster vs. Deathstalker
Give a brief synopsis of your spec’s interactions with Hero Talents. Has there been one dominant choice or has it changed each patch? Can a player pick whichever Hero Talent and be successful or is there a major gap in throughput? Does your spec routinely change Hero Talents, or has it been a mostly set-and-forget type of situation?
Both Hero talents ended up close enough to be played during The War Within. Deathstalker ended up as the primary choice for single-target heavy encounters during Season 1. Trickster ended up as the primary choice for fights with damage amplifying phases, adds, or multiple targets.
Season 2 changed this, the Tier Set Bonus benefits Trickster more, and as such, removes reasons to play Deathstalker. This isn’t a bad thing because Trickster is the better-designed option. Both Hero talents also come with their own set of problems and bugs, which will be discussed in the follow-up sections. The overall situation is very set and forget, which is comfortable.
Trickster
Trickster is the better-designed Hero Talent of the two. It comes with Nimble Flurry providing cleave damage and emphasizes damage during cooldowns. The hero talent overall leads to a fun and engaging gameplay loop but is not without issues. I try to quickly go over them.
Design Problems:
Trickster has a lot of spell dependencies, and it makes it difficult to understand. To go into details, Unseen Blade is the capstone talent with a 20-second internal cooldown, which applies a debuff on the enemy called faced. The fazed debuff lasts 10 seconds. Triggering Unseen Blade also adds a stack of Flawless Form with a duration of 12 seconds. your next Eviscerate turns into Coup de Grace after triggering Unseen Blade four times. Reaching the fourth stack is less predictable due to talents that add free casts or resets to the internal cooldown.
The combination of the above means your damage can often vary depending on random chance, but also lead to inconvenient optimization tricks like using Backstab during multi target or not using Eviscerate for a short period of time right before getting your major cooldowns back. It also adds multiple spells for what could have been one spell. There is no reason we need Fazed and Find Weakness with additional buffs like Flawless Form and stacks of Unseen Blade. We have multiple buffs and debuffs for what could be only one or two.
There is a slightly diffrent problem with Nimble Flurry. The talent is not only target capped but the target cap makes you assume that using Black Powder is worse. The reality is complicated. The use of Black Powder depends on Shadowed Finishers. Making the amount of active Find Weakness debuffs on targets the deciding factor. Using Black Powder can be stronger on as little as 5 targets for this reason if every target has the debuff. Players will typically not track the amount of active debuffs and simplify the finishing rules to 6 or 7 and more targets.
Bugs:
This is not a complete list but highlights some specific bugs.
There are more bugs outside of the mentioned, but most of them are damage amplifiers not applying correctly to spells.
Deathstalker
Deathstalker is design wise less appealing. Deathstalker’s Mark limits your ability to target swap efficiently and Corrupt the Blood makes dot management more difficult and punishing.
Design Problems:
Deathstalker has no access to additional cooldown reduction on Secret Technique. This leads to a very tight cooldown window during Shadow Blades, this means small interrruptions or delays caused by e.g. raid mechanics can quickly lead to bigger losses in damage from not being able to use two instanes of Secret Technique. Additional the slighlty lower combo point income compared to Trickster lead to less cooldown reduction from Deepening Shadows. Combine this with the requirement to stay in melee for Hunt Them Down and the importantce to not drop Rupture to keep Corrupt the Blood stacks high and you end up with a less engaging rotation.
The absence of Nimble Flurry also means spending a lot more time maintaining Rupture compared to competing Hero talents.
Bugs:
This is not a complete list, but highlights some specific bugs.
The Deathstalker’s Mark bug sounds small in impact, but a fix would probably be enough to make Deathstalker competitive or even slightly stronger than Trickster on single target fights. There are more bugs outside of the mentioned, but most of them are damage amplifiers not applying correctly to spells.
Season 3 Subtlety Rogue Wishlist
The gameplay loop is fun, but there are a lot of rough edges that would be great to see fixed. The list below highlights my wishlist for the future.
- Tuning – Output will decide whether or not you are brought to the raid. It would be a welcome change if Subtlety would compete in damage better, being mediocre at best compared to other damage dealers is not a great position to be in.
- Update to Dead talents – Perforated Veins, Lingering Shadow, Goremaw’s Bite, Silent Storm, Gloomblade and many more talents will never see play. It would be nice to see these talents replaced with new options to allow for higher talent diversity.
- Update to outdated design – Many of the core Abilities have problems which would be beneficial to solve:
– Secret Technique is just a very complicated spell with a lot of bugs. It would be nice to see the melee target requirement removed and have the spell work more like Black Powder. Changing the Shadow Clone attacks to be cast by the player instead of Pet attacks would help eliminate all the problems around logging or talents not working on Clone attacks.
– Cold Blood is due to the Tier set being very annoying to use. A change to something like “your next Finishing move after using Symbols of Death critical hits” would lower the button count while still allowing room for optimization.
– Flagellation’s main benefit is the stacking mastery buff. The spell has a damage effect that is ignorable. Having the requirement of having a target in melee and the stacking process on the target is cumbersome. Updating it to be less target-dependent would allow for better dynamics.
– Deepening Shadows is currently set too low, the cooldown reduction, especially on Deathstalker is very punishing. It would be beneficial to slightly increase the amount of cooldown reduction.
– The First Dance should get the six-second out-of-combat requirement removed. It only leads to frustration and makes it hard to understand the power level of the talent option.
– Shuriken Tornado does not fit well into the rotation anymore, combo point generation is already high during cooldowns and the damage output outside is low enough to make little diffrence. - Iteration on Hero Talents:
– Deathstalker leaves a lot to be desired. I personally would love to see a rework instead of simply buffing talents long enough to be competitive.
– Trickster has some rough edges, I would love to start using AoE spells instead of standard finishers closer to the target cap of the spell. Some of the talents could also be more streamlined. - Bugfixes – Many of the bugs lower damage output or make things more difficult to understand.
Final Thoughts
Subtlety remains one of the least played specs in the game and seems to get lower amounts of attention and weaker tuning as a consequence. This is unfortunate because it has a fun rotation of all Rogue specializations and a very cool theme. A buff would be welcome to even out the playing field. The Specialisation would love some attention to talents too, many of them remain unused since their introduction, or in general, don’t hold up with updates. The learning curve is also high, and mistakes are costly. A higher risk with a bad payoff isn’t exactly a great position to be in.
Concerns aside, Subtlety has an interesting rotation and fun gameplay loop. Even if the output lags a bit behind currently, it’s worth giving it a try if you want to try something different.
author: fuuRogue Theorycrafter of ~8 years, moderator in the Rogue Discord. Guide writer and Subtlety enthusiast. Happy to answer questions about Rogue. Youtube (soon) | Discord | Twitter | Simulations | Bug Tracker |
About The Author
