At this point I have several pages worth of feedback, but I’ll try to keep this relatively consolidated to each tree. The TL:DR version is that while the individual talents within the Class, Arms, and Fury trees are largely great additions, the layouts and pathing in particular are very linear, resulting in the majority of builds using 90% of the same talents. I really like what I can make, but I feel like I keep making the same thing every time. This may be intentional, as I don’t know how many “hard” choices there are supposed to be before the final tier, but it feels like I end up taking all the throughput talents and have nothing left for any of the defensive/utility/tertiary options. Consolidating some of those “automatic choices” (like the many “improved” rank talents) and adding additional pathing could open up new alternatives.
In general, each tree could use more cross-pathing between branches, particularly around rows 4-6 and 8-9), in order to alleviate cookie cutter build dependencies, but the biggest offenders here are the “+2% secondary stat” nodes in Row 8 of all three trees. These feel particularly uninspired, especially when the Protection tree has vastly superior versions (Cruel Strikes vs Focused Vigor). I desperately hope they just are placeholder, considering the way they wall off the signature and capstone talents below.
Another issue is that the Arms and Fury trees feature several “Improved” talents which are virtually required for the spells to function correctly, the most egregious of which include Improved Execute (both), Improved Whirlwind (Fury), Improved Rampage, and Improved Enrage (this one may be borderline insofar as “required”, but so high in value that you’ll never not take it). This severely limits early build diversity and actively works against nearby talents which don’t provide such specialization-defining effects. Highly recommend baking those listed above into their respective baseline abilities.
- Battle Stance and Berserker Stance both do the same thing - increase damage. Players will simply choose the stronger of the two (Battle is currently ~4x stronger for Fury and even more for Arms) without ever switching to the other. Stances may be nostalgic, but these don’t really serve a purpose or oppose one another the way Defensive Stance’s combined benefit and drawback does.
- Fast Footwork and Siphoning Strikes seem like nice to have talents that will rarely be used. You will never not talent Rally, and if you do have leftover points, they will go into Inspiring Presence, Spell Reflect, Storm Bolt, or Piercing Howl before Cacophonous Roar. In short, the number of “required” utility nodes alongside prerequisite talents for progressing deeper in the tree make it very hard to take some of these otherwise compelling choices.
- Rallying Cry is missing the rank 2 effect, similar to many other spells which did not have them turned into follow-on “improved” talents.
- Cacophonous Roar needs to be much stronger in order to justify its placement as dead end talent. Dan_War suggested replacing it with some form of Menace (if not the knockback, at least the cower in place), which would be much more useful from a PvE perspective.
- Intervene is a great mobility tool that I hate using, as the attack-interception has a very good chance of getting you killed in split second situations where you don’t have time to choose who to use it on. I would very much like a version of this without that effect, perhaps with an optional follow-on talent re-adding that effect alongside safeguard for increased value.
- Double Time, Honed Reflexes, Massacre, Die by the Sword, Furious Blows, Overwhelming Rage, and Frothing Berserker are fantastic talents that every Warrior will want… and every Warrior will have, because there is literally no other way to path down the left or right hand side of the trees without taking them. There’s zero diversity here.
- Rend, Thunderclap, Blood & Thunder, and Crushing Force are questionable inclusions for Fury (technically usable, but bloating the rotation and taking rage away from Rampage), but the only way to progress down the middle path toward Dauntless Duelist. More on this in the Fury-specific feedback, but better pathing options would solve the issue without necessitating their removal. Another option would be simply adding Rampage damage to Crushing Force.
- Thunderclap costs less rage and deals more damage than Rend, and while my intent isn’t to focus on tuning in this feedback, the addition of Blood & Thunder creates some awkward gameplay, whereby Rend is only cast one time and then continually refreshed with Thunder Clap. Although Thunder Clap has a cooldown, its less than half the duration of the bleed, making Rend a mostly one-time per-combat keybind, which just doesn’t feel good.
- Two-Handed Specialization is pointless for Protection/SMF Fury and Dual Wield Specialization is unusable by Arms/Prot, yet they railroad those specializations into the left or right hand sides of the tree for no particular reason. While this creates some semblance of forced diversity, it’s not really a choice, and there’s further concern that Titan’s Grip actually benefits from both simultaneously. Perhaps that’s an intentional balancing mechanism, but it feels like these two talents might be better served as a split-talent choice more centrally located within the tree… or replaced entirely, as currently they feel too strong to pass up - build defining, without affecting gameplay.
- Spear of Bastion is a good addition, but now that it no longer deals Arcane damage, please consider toning down the visual, as it can be nearly impossible to see ground target effects underneath the graphic.
- There’s also a very good argument to be made here for reducing Spear to a 45s cooldown in order to better align with other Warrior cooldowns (Recklessness, Warbreaker, Avatar, and the slew of new 45s CDs), or otherwise baking in Mikanikos’ CDR portion of Effusive Anima Accelerator, which was near universally used with the ability throughout Shadowlands - this could be a good follow-on talent in place of Piercing Verdict (both increasing damage, but reducing the cooldown instead of generating Rage).
- Signet of Tormented Kings is extremely divisive due to having very strong potential but significant room for failure when you get the “wrong” proc (triggering Bladestorm on single target, or not getting Bladestorm in AoE). Most players don’t enjoy that RNG, but consider it too strong not to use, so if this is carried forward out of Shadowlands, I feel like a good change would be to make it only proc spells that specialization can use (e.g. Avatar can proc Recklessness and Recklessness can proc Avatar for Fury, but not Bladestorm. Likewise, Avatar would proc Bladestorm for Arms and Bladestorm would proc Avatar, but not Recklessness). This would remove the frustrating RNG element and make it more of a tactical tool, while in turn allowing for more controllable tuning.
- Shockwave has very high opportunity cost, to the point where DPS will feel very hard pressed to take it over integral tools like Spear/EM and Avatar/Signet in M+, though that might actually be a good thing to keep AoE stuns in check.
- Thunderous Roar is virtually identical in function to Odyn’s Fury, which seems strange. Should probably continue to move further away from the unrealized dragon-theme as well.
Overall, the class tree presents a number of downright wonderful tools, though while the top portion is very flexible in terms of cross pathing for customization, the middle and bottom portions have far less, especially when certain specializations are driven toward their associated weapon-passives.
I do wish there were a little more Charge/Intervene/Heroic Leap support though. Things like double/triple leap would be a very nice inclusion.
Some players have suggested Anger Management take the place of Dauntless Duelist, though I’m not sure this would work out too well, since it would become a mandatory pickup and only affects spells in each spec tree. Another popular request is moving the 2h/DW Spec talents there.
Arms and Fury specific analysis to follow in separate posts.