Artificial Restrictions: Classes
It’s no secret that I’m the opposite of a fan of artificial restrictions. When you could logically do something but are restricted completely from doing so, it suspends suspension of disbelief. While I’m not deluded to the point that I believe it’s possible to achieve complete immersion, I do wholeheartedly believe we should take all possible measures in trying to preserve it.
On to what I think is the most difficult artificial restriction to argue against: a class system. Most games have one, and class systems are generally good in that the developer can inject classes with flavor. Class systems also making balancing a game much more manageable. I’m getting ahead of myself. Suffice it to say that there are many positive aspects of having a class system in a massively multiplayer game.
Why, then, am I going to argue that classes suck? A few reasons: One, they infringe upon the freedom I value so highly. Two, because I would be a hypocrite if I were to argue against all forms of artificial restrictions except class systems. And three, so I can convince myself that there is a better alternative.
To begin with, let’s look at why the opposite of a class system–that is, an open skill system–sucks. Naturally, since balance is a plus in a class system, imbalance is a drawback in its opposite. Since injecting flavor is relatively easy in a class system, it is relatively hard to do in a skill system (it’s harder to determine what type of character you are dealing with algorithmically).
So, the greatest benefit to an open skill system is enhanced freedom. I might argue precisely the converse. An open skill system means players will create templates (which are fundamentally the same as classes, only player-created). The problem with player-created templates is that players have the freedom to create them without the all-balancing eye of developers. This means that certain templates will reign supreme. The flavor of the month.
Assimilate or perish. If you don’t go with one of the uber templates or create one of your own, you will suck. You now have less freedom in an open skill system than a class system, which is exactly the reverse of what the developers were going for (unless you like to suck, in which case you have the freedom to do so).
Even in a perfectly mechanically balanced skill system, there will still be flavors of the month. Player perception will dictate what is seen as the superior template, and it will be so (indeed, player perception is to blame for most “imbalances” in any game, including those with class systems). So in a perfectly balanced system (which I’ll probably eventually argue that a mechanically balanced game would be altogether boring, I won’t do that here), you’ll still have templates that are seen as superior.
How do you conquer this roadblock? Logical counterbalancing. An eye for an eye. Everything at a cost. For every action there is an equal and opposite reaction. Whatever other quotes I came up with in my last entry on artificial restrictions.
Essentially, this means that every attribute and every set of skills has an “opposite” (in quotes because they don’t have to be diametrically opposed). Gaining in one means sacrificing in other nonreciprocal areas.
Note how I just used the words “skill” and “set.” This is important. I’m not in favor of a completely open skill system because logically counterbalancing in such a system would both be difficult to represent visually and would still be a royal pain to balance.
What I am in favor of is an archetypically structured system. Knowing one skill (or combination of skills) may lead to greater understanding of that skill set, allowing you to become better in your area of expertise. There is nothing barring one character from learning what you want it to. You don’t actually “choose” an archetype, except by how you decide to train that character.
For example, training with a hammer will improve my skills with blunt weapons, battle tactics, parrying, etc. (assuming that training is against an opponent). The main attribute tied to blunt weapons is strength, with stamina/endurance as secondary, and probably some dexterity sprinkled in there. Improving my skills with the hammer will allow me to train bigger, badder abilities.
Since I’m focusing on improving my physical prowess, my mental processes (intelligence, wisdom, etc.) begin to atrophy because they aren’t being actively maintained. Any character has a certain limit to how many attribute or skill points it can have, so skills/attributes in opposition to what you are training would start to deteriorate.
While the attribute and skill cap is itself somewhat artificial, it could be integrated fairly seamlessly and may even serve as a cool feature. What if someone could train hardcore in opposing archetypal roles and go beyond the cap temporarily? For as long as you are actively honing the roles, the skills and attributes you are training would not deteriorate (of course, the farther you get above the camp, the harder it is to maintain and faster it will deteriorate afterwards–think “hot water loses its heat faster than cold.”)
I don’t think I need to go into much detail about how to inject flavor into characters with this system. NPC reactions and the like can be based on the role tendencies of a character. Ultima Online already does it to some degree; your title is based on your highest skill. In this system, a cloth-wearing avatar with many elemental powers would be a sorcerer. A plate-wearing, bastard sword-wielding, horse-riding avatar would be a knight. You get the idea.
At some point I may describe what I’m talking about visually, but if you’ve made it this far you’re probably one of few and I don’t want to drone on longer. The gist of it is that certain skills naturally complement each other based on the attributes they are tied to, and certain attributes are mechanically opposed to each other (strength vs. intelligence/wisdom, dexterity vs. strength/stamina, etc.). Opposing skills are determined based on what attributes they are tied to, not on a complex matrix of potentially dozens of skills and how they related to dozens of others.
Designing for such a system would be a careful process of checks and balances. But the result could be something truly amazing. Freedom with real repercussions for each decision made. A system that lets players decide what their character is rather than developers. A system that is fun.

A concern with open skills systems is that the new user isn’t sure what role to fill. I loved GURPS (still do) and the very open skill structure can be min/maxed considerably (I think GURPS is a great training ground for future FOTM builders…).
But… It also offers templates to guide new players. similarly, here… you could offer “template overlays” that allow a player to pick an “archetype” and it would filter out the skill sets that are unrelated with that archetype’s theme, serving as a crutch for the people comfortable with the archetype-based games.
Heck, I wouldn’t mind a hefty and intentional break from the archetypes that have been reinforced over the last 30 years… One of Charlemagne’s “Paladins” was Malagigi, the enchanter (not to mention the Turpin, the Archbishop). The idea of such a “knight” in title and ability but not necessarily in armor or weapon is rather appealing to me…
I’ve played more than one game that had a skill system instead of classes and games that had classes with a pretty open skill system. They all left a bad taste in my mouth. Talking about skill systems makes it sound like a great idea, but in practise I have never had a good experience with them because people could min/max too reliably.
In terms of balance considerations surrounding an open-skill system, the three obvious solutions to me are as follows:
1) Dilemmas – like you mentioned, increased strength leads to decreased intelligence and so forth. Dilemmas should exist at all points of character development and in all arenas. Arguably the best portion of character development in World of Warcarft is their talent-spec system, which forces mutually exclusive tough choices that change gameplay
2) Contrived counters. In the world of uber-specs, there needs to be a designed counter to every possible ability. Perhaps it shouldn’t be a “guaranteed win” when water mage fights fire mage, but fire mage should have a rough time of it, even though he pwns leaf mage all over the place. Most games attempt to do this early in development, then lose sight of the structure as they add on.
3) Something I’ve never seen in a game – a game in which the power of your spec or ability is augmented or nerfed algorithmically according to how many other people are running it. Itcan make logical sense in some spheres (the planet only produces so much fire mana) and is rather nonsensical in others (my massive sword of the Titans no longer does much damage because everyone else is using one).
Another possible implementation would be to scale the “cost” of developing the ability with respect to the amount the ability is currently used in the game. If the developers failed and Fire spec is dominant for any given level, then X amount of “character development” gameplay time would yield either a L45 fire spec or a L60 Ice spec. Assuming no absolute caps on development, this encourages variety and almost eliminates the possibility of “gimped” specs.
This really runs about like a market model, which requires some oversight, but ultimately runs itself.
This last idea would very much upend the play experience of gamers used to man-max speccing. It would be challenging to implement correctly. But it would require innovation, tradeoffs, and variety from players all the way through the game’s lifetime.
Interesting ideas, Irond_Will. The main issue I see with the last one, however, is that players can be assholes. Like me. I would figure out what template my enemies were using and create characters on the same server (along with all my guildmates) that mirror their templates, thus reducing the power of my enemies.
[...] I am still one of the few MMO developers who is convinced that skill-based games are better. Read up on most of my opinions in an article I wrote a few months back called Artificial Restrictions: Classes. What it comes down to is that I don’t like being restricted to defining a class for my character. I prefer for the actions of my character to dictate what my character is best at, and don’t feel any developer is capable of determining that for me (as a player). [...]
[...] I was away, of course, but i am sure that plenty of people have expected me to jump in on the never-ending debate on class-based systems. So I thought I should, but with brevity. [...]
[...] The topic of “Class vs Skill Systems” is a very very very very very very very very popular topic right now. [...]
, who is also a fan of classes for the most part. I am still one of the few MMO developers who is convinced that skill-based games are better. Read up on most of my opinions in an article I wrote a few months back called Artificial Restrictions: Classes. What it comes down to is that I don’t like being restricted to defining a class for my character. I prefer for the actions of my character to dictate what my character is best at, and don’t feel any developer is capable of determining that for me
[...] How did I miss this? There’s been a flare up of the Class vs. Skill debate recently. You can track it across multiple blogs: Nerfbat, Zen of Design, Broken Toys, Moorgard, Nerfbat, Probably Not, Sierra Kilo, Raph Koster, World IV. [...]
The whole balance arguement just needs to be thrown out. Everquest was never more balanced than UO. WoW is no more balanced than SWG was. Both systems suck at balance and frankly I don’t think either one contribues equally towards balance.
Balance is always and will always be in the eyes of the player. A rogue can do 10 damage per second and a warrior can do 30 all in a single blow every 3 seconds… the warrior is percieved as overpowered, but on paper its 30-30 damage over 3 seconds. The developers may call it balanced, but in a dynamic setting it never pans out.
Anyways this is a dead horse that really has been well discussed by the smarter people in the industry. Still don’t see any of the answers getting put into any games, but we all know they are really (and I mean REALLY) working on it.
What you descibe (something I like and have often thought about) is more or less a class system. However, it’s fuzzy. It allows a player the illusion of natural growth, but, except in extreme circumstances, has still essentially given rise to a class, or a multi-class character.
The major benefit is you don’t have to reroll if you think warriors suck. You just don’t hit shit no more. Eventually you will become something else.
It’s an enhancement of the UO system to some extent but with many of the benefits (I hope) of the class system. You’re designing classes, but allowing the player a mechanism of changing classes. You’ve just hidden the classes and the mechanism somewhat.
I think the answer or solution is to realise the flaws each system has from a fundamental point of view regarding our expectations. Class systems are great because we know when we meet another class we know what to expect, but if you choose a class and build your character up, he/she will be exactly the same as mine with different equipment/augmentations. If it is skill based character then you don?t really feel like you are anything and you can become a ridiculously strong character that is a fast as lightning, and can jump and can do all sorts of things?.well here is where I think lies the caveat. If you are taking a particular skill and improving it to the extreme, then shouldn?t other skills be negatively affected? If you are ridiculously strong, shouldn?t you be ridiculously slow? and have a low dexterity? ( i know many of you will say, but I want to combine power and speed!!!) well fair enough, but something needs to give, otherwise that is where the imbalances come from; you can?t have your cake and eat it too! LOL Seriously, if there is some sort of skill restricitions put in place that follow the rules of the ?fantasy? world in which the game is set, like opposing skills that have a negative proportionally-inverse effects on each other or a maximum skill-total you are allowed to have, then the classes will be ?born? out of these limitations. You therefore have a skill based, class generating RPG engine. Some class systems limit you skills points but that is not what I mean, and some skill systems generate certain types of characters and that?s not what I meant either – I mean balance the skill engine so you can not dominate through supremacy – you have to choose your niche and there are certain type of characters that will be your rivals and others will be a push-over – which also introduces the need for allegiances with complimentary characters in MMOG?s if you want to survive?.. Just think about it!!! To me it is staring developers and everyone in the face – don?t allow supremacy to beachieved in either game system – only specialization through skill limitations/dynamics that allow classes to be ?born?.
Think about it.
In the project I have been developing for the past two years, Genesis (www.playgenesis.com), I chose to go with classes versus skills, although admittedly it was a tough debate for me. It sounds like what you are proposing is more “maintaining the illusion” that the game is free-form. Because even though players can select whatever skills they want, they are still ultimately working towards building one archetype. I too despise artificial constraints, but I believe they have their virtues. Artifical constraints are what makes games games. Imagine what chess would be like if you could train your pawn into becoming a Knight? If something becomes to freeform, it can lose the art of strategy. If there are an infinite number of paths, it becomes much more difficult for players to develop strategies and recognize patterns, and as Raph Koster puts it, these are the things that determine fun. Classes are simple, and they allow for role-playing…who wants to be involved in a story that involves hundreds of mutt-classes running around with no defined roles? One way to alleviate the artificial constraints a bit is to have classes, but allow players to custom tailor their classes a bit (that’s how I decided to handle it with Genesis
).
The buzz in the MMO blogosphere is yet another resurrection of the Class system vs. Skill system debate. A number of prominent online gaming bloggers have chimed in with their opinions on the subject, including: Scott Jennings, Raph Koster, Ryan Shwayder, Steve Danuser, Damion Schubert, and a host of others you can find linked on those blogs. The conclusion? Most of the devs favor class systems because of their simplicity and ease of communicating character roles, while a few devs and many players
Irond_Will’s #3 was tried in the original Asheron’s Call magic system. If I recall correctly, post-mortem from the developers was that they ended up removing it because players couldn’t stand it.
“The buzz in the MMO blogosphere is yet another resurrection of the Class system vs. Skill system debate. A number of prominent online gaming bloggers have chimed in with their opinions on the subject, including: Scott Jennings, Raph Koster, Ryan Shwayder, Steve Danuser, Damion Schubert, and a host of others you can find linked on those blogs. The conclusion? Most of the devs favor class systems because of their simplicity and ease of communicating character roles, while a few devs and many players
“The buzz in the MMO blogosphere is yet another resurrection of the Class system vs. Skill system debate. A number of prominent online gaming bloggers have chimed in with their opinions on the subject, including: Scott Jennings, Raph Koster, Ryan Shwayder, Steve Danuser, Damion Schubert, and a host of others you can find linked on those blogs. The conclusion? Most of the devs favor class systems because of their simplicity and ease of communicating character roles, while a few devs and many players
[...] Posts I agree with: Nerfbat, Moorgard, Nerbat (again), Sierra Kilo, World IV. [...]
Good discussion and I felt it time to chime in.
I’m also a fan of skill systems as I hate artficial restrictions, especially as more of a storyteller/roleplayer myself. This is not to say there should be “NO” restrictions, the restrictions should simply be a lot more “natural”.
What makes a thief “incapable” of putting on a suit of plate armor? Nothing really. But the natural constraint of an armor that is noisy and slows him down would make it bad for his chosen role. A good skill system includes enough natural expected detriments for each choice to make choosing them useful. Class systems make that choice for you, and if the choice design is bad… you’re stuck with it.
Plus with the supposed idea that classes are easier to balance, I don’t think I’ve seen it actually in action. The worst cases of game imbalance that I’ve seen have all come from Class based games. If this imbalance comes from game adjustments, then you may be forced to create an entirely new character to feel useful and competent at your intended job.
I do believe that one major detriment of skill systems is not knowing how to work towards your role when you start. These games often need more starter “archetypes” for the new players to choose and work through.
Both systems target a different niche of player, and the disadvantages of both can generally be overcome with more creative and thoughtful design.
Random note: I’m going to use the term “artificial restrictions” a lot less often. Instead, I’ll use “logical” and “illogical” restrictions. Since all restrictions on a computer are somehow artificial, those two make a bit more sense.
Weekend Design Challenge: Classes vs. Skills…
A while ago Damion Schubert opened a can of worms about Classes vs. Skill-based systems (http://www.zenofdesign.com/?p=704). This caused a lot of discussion on many fronts. A good summary can be found at Slashdot (http://games.slashdot.org/article.pl?...
Not sure if this solution has been mentioned yet. I played the old school text MUD (multi user dungeons) and there was one format called “ACK MUD” that had a true multi class system. Basically, you pick the order of your classes, giving you access to all of the abilities. Depending on which order you put the class in determined how effective you were at each class ability. Example: You picked … Fighter, Priest, Mage, Rogue.
Fighter abilities 90%
Priest abilities 70%
Mage abilities 50%
Rogue abilities 30%
This helped solve the following problems.
1. Defined primary role in group play.
2. Simple character creation for newbies.
3. Freedom to solo play without dependancy for certain abilities.
4. More flexibility compared to single class archetypes.
5. Somewhat balanced in that everyone has access to the same abilities just different in their effectiveness.
Reply for the two main problems you mentioned above.
1. Skill based less horrible balance.
2. Class based more player control.
1A. Eventually, power players are always going to get the similar templates of skills for PVP, so why bother limiting everyone else? If everyone has access to the same abilities this creates an anti monopoly. Different class combinations can be made to counter other class combinations. In real life, you use the right tool for the job, no body cares about the ultimate screwdriver! Maybe some do but that is not the point.
2A. Freedom of choice without limitations. This is not a perfect solution but it’s a happy medium. You’re not really forced into a particular class role because you can still perform all the other functions. The power to choose your class order, gives you more creative freedom.
The one main complaint was uniqueness. Since everyone can do the same thing, no body felt unique. Aside from that drawback, this system was very popular with MUD games. I don’t know why they haven’t tried it with MMORPG. As for uniqueness, there are tons of ways to make characters unique. Just to name a few: character appearance, class geared equipment, race selections, race vs prime class appearance.
Elf Mage Prime = Skinny Elf
Elf Fighter Prime = Buff Elf
If anyone wants to discuss further, you can email me at: dreygore@hotmail.com
[...] players stats they could manipulate that would determine their specialization (see Nerfbat’s writeup for ideas). Players could create their own concept and not be bound by Artificial Developer [...]
[...] Perhaps at some point I might write a blog with ideas. In particular, I think NerfBat’s article gives a good picture of why Skill Based systems are hands down stronger for roleplay servers due to [...]