ID:154572
 
In a roleplaying-focused game, people use either skill-based systems, or class-based systems, and you can find almost an equal number of both. Both seem to be equally popular, but I can't seem to find any good discussions on which is better.

So discuss away! =)




Personally, I like skill-based, class-less systems, in that anyone can do anything they like with any character. By fuzzing the boundaries of classes, they can make multi-purpose characters, who will, however, have a definitive lack in at least one or more skills.

Since I've been more or less distracted from AntWorld's coding, I have been drawing out, in computer Notepad and on actual pen-and-paper notepad, plans for my sci-fi MUD, which is going to be a revamped version of the former squad-based realtime strategy TekForce game.

(Basically, a Cyberpunk/Shadowrun/Fallout/original game, which, instead of magic, has "psi", "chi", "phi", and "omega" psionic energies. I haven't made any code for it, though! I'm still true to my AntWorld-devoted word, mostly. =)

My main issue is, that when I'm designing my combat system and the world, it's obvious that some skills will be used way more often than others, and some skills will never be used.

Which is the precise problem with a skill-based system; people tend to abuse the skills to pick only the excellent combat skills, and tend to ignore the specialty skills. So you could have Billy the Berserker with maxed out skills in everything but non-combat skills like Disarm Traps and Pickpocket, since those skills are almost useless in a combat environment, and a large portion of the game is centred on those. And Billy would make a killing, literally and figuratively.

The major problem is to find ways of rewarding people for using the obscure skills, but at the same time not upsetting a balance; so if a bandit took Pickpocket, Lockpicking, and Support Weapons, that bandit would be practically impossible to defeat, and they could also get high rewards for using Pickpocket.

In other words, it's nearly impossible to make any lesser skill worthwhile without upsetting a balance, since a character will just focus on combat skills in addition to that lesser skill, as a kind of exploit.

Darn, huh? I have no idea what I should try...


A class-based system, on the other hand, is more restrictive as to the tasks that a player can follow, but at the same time the class-system prevents abusing the system, and rather raises the question as to "Which class is better if I want to win" rather than "Which skills should I take if I want to win".

(Of course, you can always learn something from reading through the source of Roguelikes, whose character classes are excellently balanced. But that's a digression.)

I like well-balanced class systems, too, since playing as either a wizard or a warrior has its challenges and benefits. In a Roguelike, a wizard has a tough start, but once they get going they're almost unstoppable, except by freak coincidences. Likewise, a warrior has an easy start but it begins to get harder and harder to stay alive.

The problem I have with class systems is that, usually, none of the classes mimic what I want to do precisely.


Anyway, questions, comments, ideas, input, output, throughput, bandwidth, bandwagon, suggestions?
With a big post right above it, it is possible that people missed this one, and it is kind of a fundamental concept that I need to get rid of before I finish AntWorld and Hunter and then start to work on it for real. =)
Well, as you said, the main problem with skill-based is getting all the skills balanced correctly, and the main problem with class-based is avoiding making the classes too restrictive. I can see three decent solutions to the problem:

1) MOUNDS AND MOUNDS OF CLASSES. This is probably the worst solution, but sadly, probably the most popular. I think it's pretty self-explanatory: just churn out classes by the dozen until you have almost every single possible character anyone could ever want to play as covered.

2) Combine the two systems. Have a class system that determines the general overview of a character, and then throw in skills to let players fine-tune. There's two ways of doing this: Having all skills accessible to everyone but making it very hard to do well in skills outside your class, or simply restrict skills based on what class you play as.

Now, as this is essentially the basis of several common MUD engines, I should hasten to point out that I don't mean merely do what most of these MUDs do. Most MUDs out there that use a combined level/skill system use the skill system almost entirely as a means of advancement. It's not "OK, you're a fighter, now choose what style of fighting you want to specialize in"; it's "OK, you're a fighter, now go out and level and practice progressively more powerful attacks which all function about the same." This latter view tends to lead to much less itneresting games.

3) Have an advanced skills system. i.e., rather than simply choosing a powerful skill and reaping the huge rewards that go with it, make those huge rewards contingent on having a large array of related, somewhat overlapping skills which all contribute towards a common goal. This leads to a system somewhat like #2, but a lot more flexible. If being a skilled burglar can get you massive amounts of cash, rather than just letting players being able to happily loot wealthy homes dry by getting the "Pick Locks" and "Sneak" skills or whatnot, make 'em practice 10 or so before they become master thieves. If they don't want to practice that many skills but still want some thieving ability, then they only practice 4 or 5 of that set of 10. Maybe they can get around alright but are very easy to catch off-guard. Or perhaps they're terrible lockpicks but can hear someone coming from a mile away and be out the door with their loot before the owner knows what hit 'em. You get the picture.

The big drawback to this one is complexity. I don't consider this a drawback at all, but unfortunately a lot of players do (curse their hides!) You can lighten the load a little bit by simply sticking to a few skills but making players gain them incrementally, so that while effective burglary still only requires 1 or 2 skills, you have to practice them a lot. This is a bit less flexible than the mounds and mounds of skills approach, but your whiny evil players will probably like you a lot better.
On 4/24/01 9:10 pm Spuzzum wrote:
In a roleplaying-focused game, people use either skill-based systems, or class-based systems, and you can find almost an equal number of both. Both seem to be equally popular, but I can't seem to find any good discussions on which is better.

So discuss away! =)

I prefer a skill-based, class/race-blind, free-form system.
The concept in daggerfall is excellent.
1. Have some guilds which take in members who applied.
2. Have lots of skills. And skill gains instead of levels.

What I would like to add to daggerfall:
1. No race! (What is the genetic basis for them anyway?)
2. No levels! (Just complicate matters.)
3. Plenty of skills!
4. Non-combat based skill advancements. (Logical as kill a rat would not increase my intellegence.)
5. No fixed story. (Players get to do what they wish.)
6. No reward system based on creature-killing. (Why do rats carry cash?)

If you use a class-based, levelling system you will end up with players who do nothing but kill creatures. That is not roleplaying.

I prefer class based to skill based because then players who create their characters usually end up with a swiss army knife of a character who never needs any help. In Maevea the majority of the game, I hope, will be based on players working together. A decent party for instance would need a fighter to do the bashing and the slashing, a priest to clean up and beef up the fighters, a thief to pick locks and scout, and a mage to give it some flavor, gotta run, class is over!
In response to DerDragon
In Maevea the majority of the game, I hope, will be based on players working together. A decent party for instance would need a fighter to do the bashing and the slashing, a priest to clean up and beef up the fighters, a thief to pick locks and scout, and a mage to give it some flavor, gotta run, class is over!

On the other hand, probably the biggest reason I left EQ was because around seventh level I got to a point where I couldn't realistically hope to make any kind of progress without joining up with others (or killing roughly a jillion dinky spiders and bear cubs). And it was no fun to be trapped in essentially two zones for fear of almost certain death if I proceeded beyond.
In response to Gughunter
On 4/26/01 11:49 am Gughunter wrote:
On the other hand, probably the biggest reason I left EQ was because around seventh level I got to a point where I couldn't realistically hope to make any kind of progress without joining up with others (or killing roughly a jillion dinky spiders and bear cubs). And it was no fun to be trapped in essentially two zones for fear of almost certain death if I proceeded beyond.


I think all the concerns expressed on both sides are relevant and that this is one of the toughest decisions for a MUD to make. It seems like in a perfect world things are skill-based, but that class-based systems are easier to "get right" than skill-based.

Anyway, as someone who has played EQ for 2 years and has a level 43 Druid who has almost exclusively soloed, I just wanted to add my own take on Guy's comment. How EQ addresses this issue is by providing classes who have the ability to solo, though they are supposed to take longer to solo than if they group (reality turns out otherwise, at least for Druids).

The druids and necromancers are completely soloable classes, shaman are more or less soloable, mages are somewhat soloable, and everyone else (wizards, paladins, warriors, clerics, rogues) are not soloable at all.

So this can be covered in a class-based system, though it is hard to do so perfectly.
On 4/24/01 9:10 pm Spuzzum wrote:
In a roleplaying-focused game, people use either skill-based systems, or class-based systems, and you can find almost an equal number of both. Both seem to be equally popular, but I can't seem to find any good discussions on which is better.

So discuss away! =)





the best is none at all.. a race and class should help in some areas and hinder in others but it should not prevent any skills uses. a class should be a start.. it should be so you deside what char is the closest to what you want.. as i said, both skills and classes should be in games but neather should be iron clad...
In response to DerDragon
On 4/26/01 11:06 am DerDragon wrote:
I prefer class based to skill based because then players who create their characters usually end up with a swiss army knife of a character who never needs any help. In Maevea the majority of the game, I hope, will be based on players working together. A decent party for instance would need a fighter to do the bashing and the slashing, a priest to clean up and beef up the fighters, a thief to pick locks and scout, and a mage to give it some flavor, gotta run, class is over!

This isn't so much of a complaint against skill systems as a complaint against BAD skill systems. Too many entirely skill-driven systems take the "jack of all trades, master of none" player archetype which seems so popular there and sort of neglect the second half, letting players pick all the most powerful skills and leaving behind a handful of useless skills players will never need. To recap what I said in my post below on this issue, there's two ways to go about making skill systems more balanced:

1) Make MOUNDS AND MOUNDS of skills, and whenever possible, rather than conferring a really useful ability to someone on the basis of learning a single super powerful skill, spread it out over a number of contributory skills that players must acquire all of to work at full effectiveness.

2) Make a more moderate number of skills, but put a lot more emphasis on degrees of learning. I notice a lot of games you-have-it-or-you-don't skill systems, so you're either clueless at how to pick locks or an expert. Even more games will make it so you have a learned level for skills, but most of these end up being played as you-have-it-or-you-don't because there's no real reason NOT to learn something at max. A big problem here is simply giving players too many points to allocate to skills, or spells, or whatnot; if you've got 10 spells per spell level and players get enough skill points per level to learn 9, then you better hope you have the most spectacularly well balanced spellbook in the game or else you'll end up having 1 or 2 useless spells per level, and so everyone ends up with the exact same spells, at the exact same proficiency. BOOOOOOOOOOOOOOOOORING. What if you had 10 spells per level, each with 10 skill levels, and made it so each skill level of a spell cost 1 skill point... and only gave players 20 skill points per level? Players would only be able to learn 20% of the spells in the game AT MAX, which means that, one, players only partly practice a lot more skills/spells, and two, there's going to be a LOT more difference between the spells your players know, letting players customize their characters a lot more.
In response to sunzoner
On 4/26/01 1:18 am sunzoner wrote:
On 4/24/01 9:10 pm Spuzzum wrote:
In a roleplaying-focused game, people use either skill-based systems, or class-based systems, and you can find almost an equal number of both. Both seem to be equally popular, but I can't seem to find any good discussions on which is better.

So discuss away! =)

I prefer a skill-based, class/race-blind, free-form system.
The concept in daggerfall is excellent.
1. Have some guilds which take in members who applied.
2. Have lots of skills. And skill gains instead of levels.

What I would like to add to daggerfall:
1. No race! (What is the genetic basis for them anyway?)
2. No levels! (Just complicate matters.)
3. Plenty of skills!
4. Non-combat based skill advancements. (Logical as kill a rat would not increase my intellegence.)
5. No fixed story. (Players get to do what they wish.)
6. No reward system based on creature-killing. (Why do rats carry cash?)

If you use a class-based, levelling system you will end up with players who do nothing but kill creatures. That is not roleplaying.

No, but it's fun.
Whoa, lots more replies than I expected! =)

Someone mentioned having 'neither'. Do you just mean that anyone could do anything, but your aptness for a particular task would be based off of your character's statistics?


Anyway, with some of the blanks filled in from the replies, here's what I have come up with for my TekForce sci-fi/cyberpunk/post-war MUD:

  • no classes or specific races; everyone is human
  • everyone has at least SOME capability to do something; in other words, if you don't have the skill, you can still try to do the action, and though you'll have a larger degree of failure, you can still succeed (in game terms, it uses an "Untrained Skill Roll", which takes the lowest two of three dice. Anyone played FASA's Mechwarrior RPG system?)
  • if you know a skill, the level to which you know that skill decreases the difficulty modifier of the task you're trying to perform; i.e. if you have Knifing 4, you have a -4 bonus to your target roll when attacking with a knife.
  • skills are loosely based on your stats; if you have huge strength and build, then you're really good with a knife, but you absolutely stink with a lockpick
  • you can join universities and academies, which give you basic skill packages (i.e. a bunch of skill points divided between certain skills) cheaper than normal. (Also taken from Mechwarrior). If you go and buy a package, you assign various levels to the various skills. So you can take one level 3 skill, two level 2 skills, and three level 1 skills from the list given to you at the academy (as an example of one of the packages). Of course, you can't learn a skill at an academy more than once; once you learned a skill, you have to increase it from your own experience. Of course, if you learned the skill on your own terms, if you go to the academy, you can pick that skill, and get a bonus of that many points to your current skill level.
  • you have a number of skill points... each skill takes a cumulative number of skill points to increase, and it also takes a cumulative amount of experience to receive skill points... i.e. you need 1 point to buy a skill in the beginning, than 2 points to raise that skill to level 2, then 3 points to raise it to level 3, and so on... thus, it took 6 points to raise to level 3
  • characters have a pool of 100 character points, which they can use to buy better stats or more skills; players may also purchase Advantages, which boost their abilities, or sell Drawbacks, which detract from their abilities. Of course, 1 Drawback gives you 15 points, but an advantage costs 30.


    Remember, I'm making this a MUD, not a personalised kind of MUSH creation. So some lack of realism is necessary.
In response to Spuzzum
Heh. Sounds rather GURPSy, which is always a good thing. Especially if it has as much background stuff as GURPS sourcebooks always have, cuz that's hella good reading.
In response to sunzoner
Another way of looking at this problem...
http://www.mpogd.com/editorials/huber1.asp
In response to sunzoner
On 4/27/01 1:24 am sunzoner wrote:
Another way of looking at this problem...
http://www.mpogd.com/editorials/huber1.asp

It's a good editorial and one which I agree with quite readily, but correct me if I'm wrong, it never once touches on the single most massive defect in the D&D-style HP system: IT MAKES NO SENSE WHATSOEVER. I've heard a few half-assed attempts to justify it realistically, but the result on a lot of games seems to be that a level 1 character goes out and kills a few rabbits, and this results in a significant boost in their ability to withstand punishment without dying--much as of course is the case in real life.

Riiiiiiight...

Also, I fail to see why the fact that a master assassin can kill untrained, naive apprentice wizards in 5 seconds flat every time is something to be avoided. Realistically, the main discouragement of PK shouldn't be that it's hard for master assassins to kill untrained, naive apprentice wizards, it should be that it's massively illegal, and, if caught by the city guard or whatever comparable law enforcement agency your game had, would result in the killer's imprisonment or execution. Note the emphasis on "in-game" punishment rather than "out-of-game" punishment such as banning or purging.
I'll talk about the games I know best, those being GS and Cerulea.

Both are basically the same in that your class determines how much it "costs" to train in skills that are available to everyone. A wizard can learn to swing a claidh, but it's so expensive he has to skip other vital skills. If you were to step onto the GS boards you'd hear lots of complaining about balance between the classes, but I think they're amazingly balanced considering the game was only planned with up to level 30 characters in mind, and now we have level 200 characters, and nothing "broke." The fact that most professions have unique spell lists that can be adjusted as necessary helps. To see how training costs per profession are laid out, click here.

Cerulea will be harder to balance because of the way I've straightjacketed myself. If you've seen the website, you may recall that the professions were lined up in a certain way, and the skills fell under them; how much a skill cost for you depended on which profession it was under, and how far away from your own that profession was. Rogues are pretty far from priests in the layout, so thieving would be hard for a priest to learn. If you've clicked on the above link, you'll see the GS costs are pretty much all over the place.

Both skill systems operate on a 'diminishing returns' philosophy where the better you get at something, the more it costs to train more in it (Cerulea) or the less benefit you get for training more (GS).

Cerulea isn't level-based. GS is, but for the most part very few people pick a certain profession with the notion it's the most powerful. The game is very rich, people tend to pick what they want to roleplay. Race-profession combos, though, are often chosen for power reasons.

One major controlling factor you can have in a level-based system is what you'll see on the GS training sheet there; number of times per level you can train in something. A rogue can train in hiding three times per level. A sorceror, only once.

I prefer these systems to one like, say, AD&D. The latter just isn't realistic. I want that halfling wizard to be able to learn how to swing a claidh. He's just going to have a hella time trying.

Z
In response to Leftley
On 4/27/01 2:12 pm Leftley wrote:
Also, I fail to see why the fact that a master assassin can kill untrained, naive apprentice wizards in 5 seconds flat every time is something to be avoided. Realistically, the main discouragement of PK shouldn't be that it's hard for master assassins to kill untrained, naive apprentice wizards, it should be that it's massively illegal, and, if caught by the city guard or whatever comparable law enforcement agency your game had, would result in the killer's imprisonment or execution. Note the emphasis on "in-game" punishment rather than "out-of-game" punishment such as banning or purging.

I agree with you on that count. But Pk'ing may not be what the players want, so a balance have to striked. Read the pk thread for more detail on this one...
That's easy. The better one is what you like more.
It's like anything else, nothing is necesarily better than the other, just what people want.
In response to Leftley
I'd agree also about the skill thing. Classes should have their own skill trees with varying levels of skill for those skills. I'm using basically every 2nd ed D&D spell you can think of plus some of my own for Maeva, and abilties for fighters and thieves and stuff will go beyond that. I'm just opposed to the idea that someone gets a character and just sits online all day working their way up to a demi-god that doesn't need help from anyone ever. I prefer players have a large list of skills at their disposal, but have all these skills tailored to their general class. That is, fighters will be doing things like jump attacks and throws, while wizards will have a library of spells. My friend is urging me to go 3rd ed with this, but I feel like its sort of bordering on classlessness, so I may not go there. Oh well, we'll see how everyone's games turn out, and the best of luck to all of us.
On 4/24/01 9:10 pm Spuzzum wrote:
In a roleplaying-focused game, people use either skill-based systems, or class-based systems, and you can find almost an equal number of both. Both seem to be equally popular, but I can't seem to find any good discussions on which is better.

Hrm, perhaps the problem is in the question. With class vs. skill based systems there is no 'better', only different and preference. Understand that class and skill based systems are designed for different purposes and, so, have differing strengths (and failings). Here is a brief rundown:

<h2>Class based systems:</h2>
Class based systems are the granddaddy of all rpging. Way back in the 1970's Gary Gigax developed a medieval-themed strategy wargame called Chainmail. "Classes" were just that- the different divisions of units. As time went on, however, Gygax continued to add more and more detail to the game until, at one point, it was decided to run an infiltration scenario where each player controlled only one elite character instead of an army. Thus the original rules that developed into D&D were born. Classes were a convenient way of organizing abilities and making rapid distinctions while levels provided clear indicators of advancement and ability. In essence, level/classed based systems are heavilly weighted towards quick comparisons/computations and game balance (as is appropriate to a game with such roots). These systems are best suited when statistics, ratings, or other such 'attributes' are the primary focus (such as in a combat intensive setting). Gathering of positive modifiers to statistics (treasure hunting) and advancement, usually by facing more and more powerful opponents, are the main goals. The disadvantage of such systems is gross power inflation and poor diversity. Rules must continually be added to do new things. In many was this system can be regarded as a 'positive' system, that is you can do only what you are told that you can do (though there is some leeway in this). As D&D's popularity grew, though, some gamers were growing increasingly dissatisfied with the number crunching, hack-n-slash, power gaming that such a system seems to encourage. Thus the rise of skill-based systems.

<h2>Skill based systems:</h2>
Before the pure classles skill based model arrived, there arose a number of hybrids such as Iron Crown's Middle Earth RPG. These games still had the statistics, classes and levels of D&D but brought in a wider array of skills and customization. Other lines such as GURPS and ICE's Stormbringer RPG got rid of classes entirely in favor of a skill based model. The trend was toward greater freedom for character development. Accrual of wealth and monster slaying diminished in imnportance as character development grew in popularity among those who dubbed themselves "true role players." This desire for weight on the psychological and social aspects of gaming (as opposed to emphasis on the numerical) has lead to some fairly sharp divisions within the gaming community. In fact some RPGs, such as AMBER RPG have become so abstract that dice are completely unnecessary! The emphasis is entirely on player interaction and decision making. This, then, can be seen as the greatest strength of skill-based models: freedom. Players may act and learn as they choose, free of arbitrary constraints. Of course there are many weaknesses as well. Play balance becomes much more difficult as one must attempt to compare a master chef with a rogue swordsman. The system is also heavilly reliant on external justice. Whereas a class-based player can expect regular rewards and advancement, a skill based model is much more dependant on administration (IE a good Game Master) and circumstance. Skill based systems are not free from exploitation, either. As Ultima Online and White Wolf have both clearly demonstrated, power gamers can be found in skill-based systems as well, exploiting the difficulty of proper play balance.

So in the final analysis it boils down to this:
What do you want your game to be focused on?
Social interaction and freedom/lots of choice? Consider skill based.
Dungeon hopping and slaying monsters while seeking ancient treasure? Look at class based.
Strict division of powers and abilities? Class based.
Characters who have a broad range of skills but less specialization? Skill based.

All in all, it is obvious why most CRPGs have chosen the class/level model: its straightforward, popular, and well suited to the combat focus of most games. But if you wish for a more social environment, it might pay to switch emphasis from numbers and abilities to player interaction. If players can gain skill by practice and watching others, they are far more likely to interact than if the only way to advance os to go kill something. However, recognize that, especially in the independantly oriented USA, many players may be turned off by forced dependance or cooperation. It's all a question of what you wish to create and who you wish to create it for.

<font face="exocet">-James</font>