Thoughts on Diablo III: Skill Trees

We generally don’t spend much time reviewing “main stream” games on SoMoFos, but Diablo III is a sufficiently epic event to warrant some comment. It’s been over a decade since the last installment in the Diablo franchise, and anticipation reached a fever pitch last week. I’ll admit I had to step out on the sweet thing I had going with Magic: The Gathering to take a peek. It’s okay, M:TG is the understanding sort.

There’s a lot worth talking about with respect to Diablo III, but I’m just going to focus on certain game design decisions that I find pretty interesting. Diablo III made a number of simplifications since Diablo II, which I view as a great example of the increasing trend toward accessibility in games. You might recall me talking a bit about this last week. Let’s break open the skill system and see what we see.

It’s safe to say that I have a deep and abiding affection for inordinately sophisticated systems in games. What can I say? They give me the warm fuzzies. They’re like a fluffy puppy on Christmas morning, tumbling about dispensing joy unto the onlookers. My attachment to such systems is strange and perverse affection, and I am rightfully shamed by it. A man has his wants.

With that in mind, my experience with Diablo III was somewhat disappointing. Apparently the fight against the dark evils of the world needed simplifying, because most of the systems I so treasured in Diablo II have been removed. I assumed others felt a similar sense of loss, but they were too busy with their 4th play through on their 3rd character to be bothered. Alas, it looks like I am a somewhat specialized niche audience. I, for one, am appalled that “PwnLaw” was not considered a sufficiently large market for a blockbuster game.

Without further adieu, a breakdown of the first of the simplified systems and my thoughts. For whatever they’re worth (not much).

Pruning of the Skill Tree

Well, it’s more like they ripped that bitch out. For the uninitiated, the skill tree is a system of ability advancement wherein prior skills operate as prerequisites for subsequent skills. By selecting a particular skill, the player opens up a “branch” of downstream abilities. Often, selecting a particular branch prevents a player from exploring other options, which means tree is never in “full bloom” in the sense that the player may not make use of all abilities on any particular character. Here’s a visual representation of a typical skill tree:

Skill trees server a number of purposes in a game. The limited nature of the trees prompts differentiation between characters and enhances replayability. For example, a person may select a “cleric” class, but the constraints on skill choice may force specialization in a specific branch, resulting in numerous subclasses – attack cleric, paladin, healer, etc. This means a game with a 4 main classes can play like it has 16-20, which can create very interesting dynamics in multiplayer settings.

Another important implication is the enhanced replay value. Since there are now effectively 16 classes, a gamer needs to run through 16 toons before they exhaust the content. A secondary consequence is the impact on loot selection, with certain item builds being preferable for each subclass rather than the class as a whole.

Of course, there’s a serious drawback to the skill tree system. Often the player is making decisions with limited information (typically a basic skill description and a few graphics) and little grasp of the consequences of their branching decisions. This means there’s a very real possibility that a player will invest weeks, if not months, into a toon only to discover that their chosen branch is not what they expected.

Diablo III sidesteps this problem by removing the skill tree and granting the player access to all skills available to that class. This has the advantage of granting significant flexibility and removing a lot of the balancing issues that comes out of a skill tree setup. A player who invests time into a character will be given the opportunity to experience the full range of content associated with that character. This might mean they operate as a healer at some points and a paladin at others.

The constraint on skills comes in the limitation that a character may only have 6 skills active at any given time. I can see arguments a brewin’ that this forces a player to “specialize” for a given fight while given them the opportunity to respec between battles to optimize for circumstances. I think the more common practice is that the player will just use whichever 6 skills they think are strongest, and ignore everything else. I also imagine that science will show over time which 6 skills are most effective and there will be a fair amount of aggregation around these skills within each class.

Outcome? 4 classes means 4 classes. We lose differentiation and the complication that arises out of subclass/item selection, but we gain the blissful silence of those inclined to QQ for making bad decisions on skills. We also gain the knowledge that a person who signs in as a particular character is guaranteed to have access to certain abilities, which makes grouping easier to the extent group composition is important (such as on the harder difficulties).

Overall, I’m a bit sad to see skill trees go. I really enjoyed experimenting with different characters and purpose driven building. I liked that sorcerer X might be entirely different from sorcerer Y. The skill enhancements available on items further encouraged specialization, which led to some really interesting play styles (particularly when it came to the PvP elements). I know it sounds a bit odd, but I also felt I got value out of being forced to play the game multiple times to experience the content. I felt like it was a fair compromise for the feeling of differentiation.

On Wednesday – Diablo III: Items

JM: I can’t comment, I haven’t played the game yet. Funzio had a free Diablo III play day on launch, but I worked. And I don’t want to get started at home, since I may get a new computer for home. So those are the excuses. But yes, much love to skill trees. Sometimes that’s all that’s left to a lone mage in the world.

 

2 thoughts on “Thoughts on Diablo III: Skill Trees

  1. I had the same first impression when I was in the beta – the lack of trees sucked. But then I did the math, and the combinations for D3 characters are far more complex than they seem. The real difference between D2 & D3 is how quickly you get to make choices about your character, which is pretty important in a roguelike (Diablo is a casual gamers’ roguelike) because character development is an important part of the genre.

    Using the example of the D2 barbarian, there are 5.8845552e29 ways to distribute skill points for a maxed out character (without adding the additional complexity of pre-requisite skills and stat points). You will never experience every build – there isn’t enough time in your life. Not only that, but most of these builds are non-viable.

    While the D3 barbarian has significantly less choices, it still has enough options for character differentiation due to the skill runes transforming skills into entirely different skills than the base skill. So, crunching the numbers for skill/rune/passive combinations for the D3 barbarian (without the additional complexity of elective mode), we have 4.42248925e20 combinations That’s still too many combinations to experience in one lifetime.

    • Really excellent point Adam. I would respond by saying that the structure is unlikely to result in that type of outcome. Already I see a lot of aggregation around specific skills rather than exploration of the full range of possibilities. More importantly, a person’s choices never force them down a specific path, which means that folks are probably most likely to just seek a single optimal path rather than attempt subtypes (particularly since the item system doesn’t have the same sort of skill tie-ins that we saw in Diablo II).

      I do think it’s a fair point that the range of possibilities is pretty impressive though. There are a lot of things I find attractive about the skills themselves, particularly in the Witch Doctor class, which I thought introduced a pretty fun iteration on the necromancer pet class.

Leave a Reply

Your email address will not be published. Required fields are marked *

*


3 + = twelve

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>