The problem with talents
There have always been different issues with the way talents work, and people prefer different iterations of the talent tree itself. The issue I’m talking about is primarily found in the iteration of talents we’ve had since MoP. These talents let you pick 1 of 3 options in each talent row, allowing you to tailor your character’s kit to the situation you are facing, through different choices in utility and throughput. Of course in practise, there aren’t typically very many choices to be made, you just pick the correct talent for the situation based on theorycrafting, math, logs or whatever. This is where the problems start.
Specialists vs generalists
The first issue that has been quite apparent since M+ got added is that the way talents and baseline kits interact is not uniform between classes. It seems like Blizzard intends for you to pick talents to specialize into a certain aspect of the game, whether that’s AoE, ST, burst, sustained damage or something else, however, that only works if the baseline kits are all somewhat equal. It also only works if talents are tuned well.
Here’s a few examples of specs where this has been a problem:
Havoc: The baseline Havoc kit can do basically everything already, and the way the talents are set up and have historically been tuned, means that Havoc can do everything at basically its full potential with one setup of talents(and other accompanying effects like legendaries, traits or essences). Other specs that are currently in a similar spot are Outlaw and Fire.
Feral: For the first tier of Legion, Feral was a strong sustained ST spec with basically no AoE. This was the case both with the baseline spec and the talent tree, which allowed for very little change of focus into other damage profiles.
Arcane: In patch 8.2.5, Arcane is in a situation where it has a baseline kit that can do everything to some extent, and a talent tree that allows it to specialize into different damage profiles. There’s the typical raid build, which focuses entirely on maximizing bursty ST, a bursty AoE build with Overpowered and a sustained AoE build with Arcane Orb, just to name a few. The first does very high ST, but sacrifices most of its AoE potential in order to do so. The second maintains the ability to do reasonable ST, but sacrifices some in order to do AoE also. The third build sacrifices most of its ST in order to do strong, short CD burst AoE/cleave.
The first example, Havoc, seems like a clear failure of the talent system. You have a baseline kit that can adapt to any situation already, which could be an okay starting point, but then the talent tree just doubles down on the idea of just doing everything at the same time, rather than forcing specialization.
The second example, Feral, seems like a failure of both the baseline kit, because of how specialized it already is, and the talent tree, because it does nothing to try to salvage the situation.
The final example, Arcane, is one of the most successful specs in achieving the initial goal of having a well rounded baseline kit and then having talents that force you to specialize.
Now, the way the game works is that Havoc is just the strictly superior spec, because while the Feral would only be doing ST, or the Arcane would have to choose where to allocate its power, the Havoc would be doing all of those things at the same time. A more apt comparison to Arcane might be Fire, which is essentially in the same boat as Havoc. Arcane is able to compete with Fire just fine on either ST or AoE, but not both with the same build, while Fire has to change nothing in order to do both. This is makes Fire objectively superior to Arcane for most content, especially M+.
This breaking of the rules Blizzard attempted to set up, where you can’t just be good at everything at the same time, is a big flaw that comes with how different the talents are between specs. In order to reach a balanced state, the developers would need to choose to either have every spec be a Havoc that can do everything at once, or every spec has to be an Arcane where you have to make decisions on where to allocate your power.
If talents are meant to allow you to adapt to your current situation, all specs would have to be like Arcane. If the goal instead is to allow every spec to be like Havoc and be able to be good in every situation, that’s fine, but why have talents at all in that case?
A second major issue with the way talents currently work is best highlighted by going back to WoD. The talent in question is Clarity of Power, in the Shadow talent tree.
Imagine for a moment that you’re picking a main, and that you want to play a DoT heavy caster spec. Your primary options are Shadow, Affliction and Balance. You end up picking the priest and everything is great, you’re playing the DoT spec you wanted to play. Then WoD, specifically HFC, happens. Now suddenly the best build for your DoT-focused spec is Clarity of Power, a talent that directly goes against being a DoT-focused spec by making you frequently cast Mind Spike, a spell that removes your DoTs from the target. This is not what you signed up for, you wanted a DoT spec.
This is basically the second issue. It’s difficult to have impactful talents in terms of playstyle without accidentally having something that goes counter to something people made their initial spec choice based on. It only really works if the different playstyles are tuned perfectly, so you can actually pick the one you prefer still, because otherwise people end up being forced into playing something that they didn’t want to, simply because it’s more powerful. This degree of tuning is effectively impossible.
The playstyle issue is primarily a problem with player expectations and feelings, rather than the direct balance issue like the rule breaking generalists.
My solution consists of 2 phases, each lasting at least 1 full expansion. It would be very disruptive at this stage of the game, a decade and a half into it, but they’ve made similarly large scale changes before, and in my opinion it would set them up for a more consistent baseline game for the future.
The first phase is to get rid of talent trees entirely, and focus on giving every spec a well rounded baseline kit and a strong playstyle identity. This baseline kit would most likely be based on the current kit, although ideally I would prefer seeing the “peak” iteration of the spec as the starting point, probably picked based on both the devs’ opinions as well as select players. The next step would be to consider the talent tree we just deleted, and see which parts of it would make sense to make baseline, both in terms of playstyle and to create a well rounded kit. After this, some iconic effects from other iterations could be added to flesh out the kit in terms of depth and well-roundedness. This could be something like a tier bonus, an old ability or passive, a legendary effect, a trinket proc that the spec synergized with particularly well and things of that sort.
The second phase is to add talent trees back, but have them be entirely focused on tailoring your mobility, survivability and utility to the situation you are facing. Yes, this would still likely end up with theorycrafted best setups for each situation, but it would avoid both the playstyle issue, because it wouldn’t be affecting the primary rotation or kit of your spec, as well as the generalist vs specialist issue, because now everybody would have a generalist kit that allows them to interact with and participate in any situation. These types of talents also make it easier to justify “off meta” choices, as in many situations which mobility option you pick is actually just preference, for example.
I’ve already spent some time thinking about potential phase 1 specs, with 3 concepts done.