Balancing act — working without a net

I am becoming more and more nervous about class balance in Legion. What spurred me to write about it today was this admittedly decent blue post from Celestalon, explaining the rather massive talent table changes announced in yesterday’s Legion beta build, summarized in this MMO-C post. Why am I nervous? Well, it strikes me that such massive changes to an extraordinarily complex and delicate balance system, this late in the development cycle, cannot be completed satisfactorily by the time Legion launches on August 30.

As Celestalon points out, these are not just tweaks of moving a few talents around, this is a major redesign of talents with a series of cascading implications for balancing not only each spec and class, but for balancing the system as a whole.

As you may have guessed, these changes are going to require a cascade of other related changes, mostly around tuning. Not every talent row is equally valuable (nor are they intended to be), and with new talents competing with each other, adjustments will have to be made. Talents that moved, and their competition, will have to be retuned to become viable choices. Additionally, the overall strength of some specs in common situations will be adjusted to account for talents now offering more or less in those situations. An example of this is Assassination Rogues, whose baseline AoE damage was low, but had an extremely strong AoE talent in Blood Sweat, creating one of those obvious ‘test’ problem cases. We opted to remove Blood Sweat, adding choice to that talent row, and will be improving their baseline AoE capabilities instead.


Tuning has not been Blizz’s strong suit for the last couple of expansions. In WoD, it was out of whack for several months, not only for some classes but also for  meta-class roles as well. In some instances — Survival Hunters (in a bad way), Discipline Priests (in a good way) — Blizz pretty much threw up their hands and admitted defeat, saying It was too hard to fix them at all this expansion, they would do it in Legion.

And now, less than 90 days before launch, there is a decision to redesign the basic operating numbers for nearly every class/spec, a decision that will require efficient and precise tuning in a system they have already demonstrated is often too complex to deal with.

Yeah, I am worried. And the main reason I am worried can be expressed in two words: Artifact Weapons. Artifact weapons, with their accompanying talent trees and with the design mechanisms for leveling them, throw a huge monkey wrench into what is already a nightmare system to tune properly, a system Blizz was never able to get a good handle on in WoD, even without Artifacts. 

What might this design decision mean for players? It means that now your class and spec selection for Legion is even more of a crap shoot than before. If you are lucky in your choice, it will be a spec that is a winner in what will be months of “retuning” and “tweaking” of the classes. If you are unlucky, you will “retuned” and “tweaked” into oblivion, like SV hunters were in WoD. And even if you seem to be a winner early in the expansion, that is no guarantee that halfway to filling your Artifact talent tree, Blizz won’t decide your spec needs tuning, and suddenly you become That Spec that is useless both for soloing and for raiding. Your choice will be to lump it or to start all over again with a new spec and a new Artifact weapon.

One of the things Blizz touted about their Legion testing plan was that the alpha phase was good because it allowed the real serious players to give detailed feedback on some basic design and playability features, the ones that are the foundation of the game. Things like overall feel for rotations, like obvious mismatches between talents and spells designed to have intertwining effects, like huge imbalances between specs. Now it appears that much of that feedback was futile, because Blizz is going back to the drawing board on all of it. Worse, as the inevitable huge problems arise in the coming weeks, there will be less and less time to correct them, and any attempt to do so can cause major cascading problems. Pushing in one place will cause something else to poke out on the other side — the system is so complex that this is its nature.

I do not necessarily disagree with what Celestalon had to say about the design of talent trees. But honestly, it is way too late in the development cycle for the Good Idea Fairy to be given control of the process. Celestalon may have laid out a very sound design philosophy for talent trees, but it was reasonable 6 or 7 months ago. Now it is just reckless.

Still, I am certain there is no going back at this point. Blizz is intent on trashing the delicate infrastructure of talents and class balance they have spent months developing and testing for Legion. Players will be caught in the path of the giant pendulum that devs will push ever harder as they attempt to bring the system into balance after Legion goes live.

So I make yet another plea for Blizz to rethink their current restrictive system for Artifact weapons: Make them spec-interchangeable, make it so that whatever level you have reached on your weapon in one spec automatically transfers to any other spec in your class. That way, if Blizz “tunes” your spec into the abyss, you will not pay a huge penalty for their inability to figure out a design philosophy until less than 90 days before going live.

I think it is time for me to start my weekend. You enjoy yours, too.

About Fiannor
I have a day job but escape by playing WoW. I love playing a hunter, and my Lake Wobegonian goal is to become "above average" at it.

2 Responses to Balancing act — working without a net

  1. Heh, we have a Good Idea Fairy in our group. Once we’ve agreed to do a speed run through Ulduar for a chance at the mount, we hear, “you know … now that we are here …. why not try for all of the achievements?”
    I had assumed that balancing was designed for end-game raiding and that they were peeling back layers (no tier, legendary necks, common Artifacts) to look at balance at early levels. If they are NOT doing that (as seen by the talent jumbling), wotthehell? Are they simply hoping that the paint will be dry as the right color on opening night?

    • Fiannor says:

      I think there is an element of “hoping the paint will be dry … on opening night” to this, as you say. I am certainly not a game designer, but I have been involved in the nuts and bolts of a lot of software development projects, and I think the effects of this will cascade much further than the devs have planned for. (Remember the surprise they got from the cascading effects of putting flying into Draenor — delayed for weeks beyond Watcher’s original estimate.) If so, it could have one of two very bad results: Delay the launch date (not likely) or go live with a completely unbalanced class/spec system that will undergo huge changes in the first few months (more likely).