Are talent trees a failed design?

Blizzard’s plans to implement dual specs (ie. the ability to switch quickly from one talent spec to a  different one, complete with a change of gear and glyphs) is an admission that talent trees have failed them as a design.

It’s a band aid for a broken system that makes MMOs less friendly, more frustrating, and more inaccessible to casual players than they really should be.

Where did talent trees come from?

Talent trees on paper should never have grown into this strange monstrosity that they have become. They were intended to let players customise their characters and their abilities, to tailor the character to their play preferences. So as you went up levels, you had more points to spend on growing your character the way you wanted it to be. It was intended to mirror the way xp get spent in pen and paper games. When a character levels up in a tabletop game, they usually get some points to spend on stats, some to spend on abilites, and they may be able to buy new abilities also. In a tabletop game (D&D for example) you can easily build a cleric which focusses on buffs and damage. You can build a caster who chooses to specialise in support rather than damage.

Implementation of talent trees in games is associated with Diablo, so it isn’t surprising that Blizzard decided to go with the same successful model.

I hated talent trees in Diablo. You had to make some fairly fundamental choices about how your character played with no option for respeccing apart from starting again. You had almost no information to go on about how the different specs might play or what you might find more fun other than trying it (and starting again if you didn’t like it). The talent trees on offer gave you access to such different playing styles that each class was really a bundle of 3 (or more) completely different classes.

So on the one hand, there were lots of different things to try out, and lots of different aspects to each class. On the other, the only way to really try things out was to reroll a lot. It was also very easy to come up with a build that would be fine at the start but very weak later on in the game, and never realise until you had many hours of play under your belt.

Being able to tweak talents and try out different builds did give the game a lot of replayability. But ultimately people gravitated towards cookie cutter specs because sometimes you just want to beat the game, y’know? Without having to spend a lot of time starting again from the beginning.

The more hardcore players worked out the more effective talent specs. After all, they had the most time to spend on experimenting and starting again. They posted about them on forums. Less hardcore players read the forum posts and followed their advice.

Sound familiar?

Things I dislike about Diablo talent trees:

  • Being forced to make game changing choices without enough information to know how that choice will affect your game.
  • Being unable to change your mind easily when you find out that you wanted to try something different or have made a mistake.
  • Having the experimentation phase tuned such that only hardcore players really have access to it.

Note that these factors are frustrating for players like me, but may be good for replayability and building up a hardcore fanbase.

Three Characters in One, Bargain?

The peculiar thing about talent trees in WoW is that you sometimes feel that it’s like playing totally different character classes. The hybrids in particular suffer from using completely unrelated mechanics and gear from one spec to the next.

Paladins may be the worst example of class design ever seen in an MMO. They have three unrelated talent trees, each of which requires a totally different gear set, and playing style. There is no overlap at all. Balancing them in a sensible way has been a struggle for Blizzard right off the bat. It isn’t just the difference between switching from healing to dps to tanking, although that is also quite a fundamental change just from respeccing. It’s the fact that when prot/ret are balanced for mana regeneration, holy becomes way overpowered. It’s the fact that when prot/holy are balanced for survivability, ret becomes a nightmare.

Three classes in one sounds fine as a design goal, it gives players lots of options and lots of customisability. But in practice it’s proved almost impossible to balance. I think Blizzard generally does a good job these days, but they’re fighting against a talent tree system that is fundamentally broken.

Choose your talents to fit your playing style

If it was just a case of picking talents to fit your style, that would be fine. But it isn’t. The different talent trees support different areas of the game. There are PvP specs, group friendly specs, solo friendly specs.

This is glaringly stupid. People may prefer one or other part of the game but why on earth would you want to put barriers in the way of having players participate in all of it.

Why force healers to have to spec differently if they value soloing? Why force anyone to spec differently for PvP? Don’t they want people to have fun and be able to take part in every aspect of the game?

The cost of having talent trees that are specialised for different aspects of the game is that it directly flies in the face of one of the major strengths of an MMO. The fact that there are lots and lots of different things to do.

So it’s really not surprising that lots of people in WoW respec a lot. I used to respec my warrior about twice a week, and that was even without PvP. I welcome dual speccing with open arms as a band aid for a broken system.

Developments in Talent Trees

Other games have improved on the talent tree implementation. LOTROs traits, legendary weapons, and stances (I think most classes now have the ability to switch focus between group and solo style play) may not offer the wide range of customisation but let players tweak their characters but still recognisably remain the same core classes. WAR’s tactics let you easily switch from PvE to PvP focus and back again.

Guild Wars (which is on free trial at the moment by the way) has probably the best system ever devised for handling talent trees. You can respec for free any time you are in a town, and you can only pick 8 abilities to slot in your quickbars for any mission. A big part of the game is figuring out how to tailor your character for whatever it is that you wanted to do.

But what about actually just building and identifying with your character?

This is the flaw of easy respecs. Another aspect of pen and paper games is that you build up your character over time, and you can identify with it. Reworking all of its abilities every week would make it more difficult to do this.

This is one of the reasons I never really glommed onto Guild Wars, although I do think it’s a cool game. Sometimes you just want to grow into your character, not just switch it around every time you are in town.

It’s the reason that respeccing a lot on my warrior sometimes does my head in. I wrote before about identifying with a talent tree, but I know I’m not alone in this. People in WoW often do describe themselves by their talents, eg. I’m a moonkin, I’m an affliction ‘lock, etc. Easy respecs confuses that.

So from a gameplay point of view, I think that talent trees have failed. I think dual specs (and presumably triple specs sometime later) are a band aid for that, for making it easier for players to take part in all aspects of the game. I do wonder how future attempts at balancing talent trees will change to take this into account (eg. who cares how much damage holy paladins do, when they all have the option of just respeccing to ret?).

I think in future, talent trees will be tweaked more with PvP in mind than PvE, for this reason.