Guild Wars 2 systems designer talks skill design, depth and balancing

Chris Thursten


ArenaNet have been vocal about the way they're trying to redesign the MMO. Guild Wars 2's design manifesto is a confident statement of intent that has faced intense scrutiny of the course of the game's beta programme - and rightly so. MMOs have been the site of some of gaming's promises, and biggest disappointments. The manifesto promises a combat system that will do away with the "rote and repetitive" nature of most MMOs, replacing it with something "immediate, active, and visceral" that provides players with "limitless choices."

The way ArenaNet have chosen to go about achieving this, however, is couched within the existing structures of the MMO. Guild Wars 2 has action bars, cooldowns, and status effects. Its innovations - mechanics like the dodge move, and the more-or-less absence of a mana pool - occur among familiar ideas. It's also a relatively straightforward game to get to grips with, at least at first, and as such it can be a while before its purported depth becomes apparent.

I spoke to systems designer Jonathan Sharp about how ArenaNet have built the skill system, where the depth comes from, and the options available to help serious players speed up their progress.

What's the appeal of giving the player so much freedom to slot skills in and out, as opposed to having a linear levelling tree or a talent system?

So there's two different things, right. We saw that in Guild Wars 1 people could basically go anywhere they wanted to with all the different skills. You had two different classes, you had a primary profession and a secondary. What we found was that for a lot of people that was just too much. It was a lot of overwhelming stuff. Some people absolutely loved it, though, so what we tried to do for Guild Wars 2 is we said "okay, with your utilities, with your heal, with your elite you can decide those things. But for your main skills, your first five, what if it's just tied to your weapon?" Which makes it much much easier for a new player - they have a sword thats the same as somebody else's sword. Once they get that sword down, they figure out how they want to use it, they can take the rest of their bar and customise their build along with their traits to figure out how they want to play a certain class. So we try to give the best of both worlds there.

There's a substantial difference in the way that people gain skills, from weapon skills to utilities and everything else. Do you have a deliberate pace in mind for the rate at which things become more complicated?

Yep, we have a lot of different skill charts and Excel sheets about the different reward tables that we have. Every time you jump into the game we want you to be doing something fun. We want you to be looking forward to new skill acquisition. Maybe it's a new trait you've just got, maybe it's a different item. We want you to always have something that you're looking forward to.

There's not necessarily a logical connection between what a weapon is and what it does in the game. Are you concerned with making that stuff logical, or is it more about variety?

It's really a balance. If I have a staff and I'm a Guardian, what do I expect to do? I've got this huge wave attack. It fits logically, but it's not like I would look at that and go “oh I should obviously do that”. Whereas a greatsword on a Warrior is much easier because he doesn't do anything that's really magical. It's when you start doing the magical stuff with weapons that you can just say “this will do whatever I want it to do”. Hammer's a great example for a Warrior, he just jumps around and beats everybody up. That same hammer on the Guardian, because he's more magical, he's putting down wards and symbols. Which is not really logical but at that point it becomes a videogame. You're making a videogame about what this magical guy can do.

Playing an Elementalist recently I did pick up the differences between my staff and my scepter, but at the same time there was a lot of trial and error. Like, “water heals people, okay, cool.” Is it something you're happy for people to discover themselves? Or are you working on ways to teach those mechanics more clearly?

That's part of why you asked your question earlier, right? About skill progression and unlocking the skills. That's exactly why we decided to do that system. If you have daggers for the Elementalist, you have the different attunements: fire is offence, air is control, water is healing and regeneration, removing conditions, earth is a lot of times defensive. That's consistent across all of the weapons, but when what you have with the Elementalist is the range component. so dagger is your melee, scepter is your mid range and then your big AoE is the staff. Then, because you're learning each of those individually and you're learning the skills one by one, we're saying “okay, learn this. Now learn this. Now learn this .” We don't want to give you option shock. We want to explain it to you slowly.

So do you want to give people the freedom to speed up and slow down based on how comfortable they are? Skill challenges seem like a way for people to speed up if they want to speed up.

Yes. What we've done is we've made systems that let you - maybe it's your second character and you're an expert - you can just... whoosh!. There's a lot of stuff that you can do to quickly start building your character if you know exactly what skills you want to get. You also know exactly what traits you're going to be building towards. Or if you're an expert player that just jumped into PvP maybe you know exactly how to build your character because you've just found a template online. Maybe you've decided that that's how you're going to jump the curve. The systems to allow for that to happen. But at its base we don't want to overwhelm the new players. So we've tried to make a system that's easy to get into but expert players can take and they can turn into this other beast.

Next: bringing guides in-game, building archetype default PvP builds, and creating a skill system that is balanced for both PvE and PvP.

If there had been a website somewhere during the beta where I could look up builds I probably would have done. Is that something you want to bring inside the game at all, or are you happy for people to alt-tab out?

That's one of the things that I can't talk too much about, but in Guild Wars 1 we had a lot of stuff to do exactly that and we have plans to do that.

Okay, so there'll be something to mitigate the terror of “oh god I've got 70 skill points to spend”?

What you'll also notice is that when you make a new [PvP] character, I've made the templates for all of those characters and they're all built for you. I know what weapons you're going to have, know exactly what heal you're going to have, your utilities, your elite and I've also picked all the traits for you along all your trait lines. And if you want to reset that you can but again, for the new player who has no idea what they're doing - you've got a pretty good build, go run it. And then if you want to break it down you can.

So the idea is that those default PvP builds are the jack of all trades for that class?

What I try to do with those is try to give what I think the archetype is being selected for. For instance, if I pick an Elementalist I probably want to be Gandalf, I want to be throwing some big spells, knocking up some walls. Now if I pick a warrior I'm probably thinking Conan the Barbarian, maybe Maximus from Gladiator, so I build the character that I think the typical player wants. For the Mesmer you want the tricksy guy, so I made the build that gives you the tricksy guy to begin with. We give them the base archetype I think they're going to want. If they don't want that they can just redo it themselves.

Where do you start balancing a system that's that open ended? Is there a baseline you aim for?

How do you go about balancing something like this? One of things you need to do is to set a standard. What's our standard for what melee damage should be? What's our standard for what ranged damage should be? So we take those things into account but then we also have the traits, and we have the items that augment your abilities. So you need to start with the baseline for those kinds of things. Then you need to make sure that all of your traits and all of your weapons, your utilities, your heal skills, your elites - all of that stuff fits within those parameters. So you do set down a baseline and then you let the class play within those baselines.

Then you try to take stuff like, say, a knockdown. If you can knock somebody down then well I'm not doing any damage but maybe I've stopped them from using a powerful skill. Maybe I've stopped them from attacking me. Maybe I've stopped them from healing. So you need to take all those intangible things and get them into your system and figure out how much is a knockdown worth. You need to figure out that out - if this is number four out of five skills, what's it worth to not just have another skill that just does more damage?

The difference between balancing for PvE and general PvP and balancing for competitive play is pretty massive. How are you going about making sure that stuff is nailed down?

So far we've been really fortunate - we've done very few splits between PvE and PvP, it's always the same balance numbers everywhere. We knew that from day one we needed to do the PvE and the PvP. So as we were making classes, as we were making skills, as we were making traits every time we'd propose something for consideration we'd say "how does this work in PvE, how does this work in PvP, how do we balance these two together?" Then we look at the game and we say "okay, what are the tools that the PvE creatures are using? What are the tools that PvP players have at their beck and call? Then we try to say "okay, what is a PvE player going to be doing moment to moment, what is a PvP player doing moment to moment?"

We have to keep all of that in our head at the same time as we're balancing things. We do have the ability, as we need to, to take any certain skill, split it, and say that in PvE it has these numbers and in PvP it has these different numbers. The reason we try to do that as little as possible is because once you get used to your class, once you get a feel for who you are in the world, we want you to feel like you're consistent. If you're in PvE you know what you can do, if you're in PvP you know what you can do. So that's why we try to keep that as minimal as possible so that once somebody learns their character in any given format, be it PvE, PvP, WvW, they can just take that same character somewhere else and they understand what their limitations are.

So if you do make a split, do you communicate it? Not communicating it is a way of creating that sense of consistency...

Sure, sure.

...or do you just straight up say "you've got a split now"?

So right now, we're not calling out splits. But what we might do for instance in dungeons and stuff like that, we might have something that is a specific buff that only dungeon creatures use. In which case we would just call it out there, and you don't see that buff anywhere else in the game. It's still immersive, but we know that that buff only exists there so we don't need to worry about throwing players off. It's not PvP. It's not in the events. That's the way that we can keep it immersive but call it out that it is different.

Thanks to Jonathan for his time

Check out our interview with Jonathan about ArenaNet's e-sports aspirations for more of his thoughts on competition and balance. Guild Wars 2 week concludes tomorrow with a guide to getting involved with the PC Gamer war effort.

About the Author
Chris Thursten

Chris is the editor of PC Gamer Pro. After many years spent turning beautiful trees into magazines, he now oversees our online coverage of competitive gaming and esports. To date he has written more than sixty articles about Dota 2 and does not know how this became his life.

Around the web