Hi everyone. This is an issue that's not relevant to me yet, but is rapidly becoming so. I want to design a classless MUD, but I'm afraid of the perils that come with it.
Classes are useful tools. They define sharp boundaries that make balancing easier (Not easy, mind, but easier). If you have 5 classes with 40 skills unique to those classes, then you have 5 things to balance, and 200 things you can tweak to achieve it, where as if you just have 200 skills and allow the player to pick any 40, you have an absurd number of possible builds to deal with.
The problem you have won't be "Are all possible combinations of skills equally viable", since that is a pipe dream. The only feasible way to achieve that is have all skills just be clones of each other, dressed up to look differently, and that's boring. The problem is that skills can be divided into tiers, like this:
- Useful: This is a skill that makes some aspect of the game easier once you possess it. It can be a travel skill, or an appraising skill - Powerful: These are the top notch combat skills. - Fluff: These are theoretic useful skills, but aren't really useful, or are overshadowed by better options - Weak: These are combat skills that are overshadowed by the better combat options - Fun: These skills serve little to no purpose in or out of combat, but have some mechanism that make them interesting. Note that all the other skills could be "fun", but for the purposes of this post we'll stick with the specific definition.
In a class based system, you can provide a mix of the above skills to each class, giving them variety and options. Even the weak and fluff skills can be tweaked so that instead of "useless", they're just situational.
In a classless system, people will just pick skills from the first two, and sprinkle in skills from the last one, while ignoring anything fluffy or weak. This would be fine, except if your skills are evenly distributed, and you have 200 skills, fully 80 of them will not be used by players at all. And they aren't usually evenly distributed, because it's difficult to come up with 40 unique and equally powerful skills, or 40 unique and useful skills, so the distribution actually looks more like 40 "good" skills, 20 "fun skills" and fully 140 weak options. That means that instead of unique diversity, most players will have a core of 20-30 must have skills, and diversify themselves with the leftovers. Things start to look a bit samey, if everyone has a horse because it's the superior travel option to teleportation (or vice versa).
One of the ways to reduce this is to have skill paths, where one must accept X less powerful skills to get at the good skill. This encourages a degree of specialization, and encourages players to be creative with the extra skills, since they only get so many of them. It also forces more meaningful choices. A good example with this is IRE games. They are class based, but only 3 skill lists out of many available to any given player are class specific. The none class specific skills are very useful, so deciding how much to allocate to those vs your class skills is a meaningful choice. If the MUDs were to go class free, it would be even better, since the "must have" skills fall at the end of the lists for class skills (This is done for economic reasons as much as anything, I'm sure, but it's still a good example). The problem is that if you have too much forced specialization, you start to have things that look like classes again. If to be a competent healer you need to focus on healing at the exclusion of everything else, you might as well just have a healer class. Generally, having skill lists (or trees) and putting the good ones at the back is a good way to ensure that anyone attempting to take multiple trees halfway gets screwed over by having many options but none that are good, and if you don't support those players, then you're creating an implicit class system, the thing you're trying to avoid in the first place.
So where do you draw the line? The options I see are:
1) Try to have as many powerful/useful/unique skills as possible. If there are many equally good options, players will differentiate on personal preference. This is terribly hard to execute 2) Have a class "lite" system where sticking to a paradigm is rewarded by power in the endgame, but is not necessary. You can mitigate some of the inflexibility by sprinkling relative powers into the lists all the way through.
Are there any other options?
Also, suppose you're picking option 2. Are there any guidelines into making each tree attractive, so that you don't get everyone following the magic path and not have any melee users? I suppose this is the same question about having unique but equally viable and interesting classes, extended to a more free-form system.
*I should mention that the numbers provided don't represent anything about the MUD I'm planning. I haven't begun planning, let alone coding skills
The problem is that if you have too much forced specialization, you start to have things that look like classes again. If to be a competent healer you need to focus on healing at the exclusion of everything else, you might as well just have a healer class.
Well, I see it as you having to work hard to get good at something. If you want to be a doctor, you won't be able to heal anyone if you can't figure out how to apply bandaids. That's just how the world works. What if you want to get to a point in a medical field, but only just enough to get a certain level of competence? You really want to be an engineer so you max out the engineering skill tree, while still having some lower skills that a medic could do.
This is the approach we use in HOPE. In addition, some skills in one tree will actually require skills from another tree. You can't be a pilot and just know how to steer. You need a background in multiple fields. I am sure that many of you will say this is too forced and structure, but I don't really think it's feasible, or even a good idea, to incorporate a completely unstructured system.
I am not saying this is your answer, I'm just offering another viewpoint from what we are trying to develop. We sure think it will work.
Thanks, that's exactly the sort of answer I'm after (not THE answer, but I'm looking for different perspectives on the problem).
You bring up a good point, that if someone wants to be good at something, realistically they will need to specialize at some point, just because there's only so many hours in a day. I guess the main benefit of a class-lite vs a class system is that you can kind of grow into your class, rather then have to pick it at the start. In your early days, you can wield a sword, cast spells, pick locks, heal, and be stealthy, and do poorly at all these things, but only because you are low leveled, and since you will tend to level up the mechanics you like, you pick your class almost by voting. Leveling up lockpicking is like saying to the game, "I like this lockpicking thing, so give me more of it".
Your second point of having requirements from different trees applied to certain skills is interesting. It solves the generalist problem, to an extent. For example, you could have a mage tree and a swordsman tree, and if you have certain skills from each tree you skills in the swordmage tree. You could even have a "generalist" tree that comes from having a bit of everything. The only problem I can see is that you're going to need to support the many combinations explicitly. At some point you need to sit down and say "Let's make swordmage skills", and after you do that, are you going to address every possible combination? Sword-theif? Theif-Mage? Sword-Theif-Mage?
The other option is to not really have explicit boundaries in your trees. Just make design all your skills, and link them intuitively. The skills can be "grouped", but more out of rationality an convenience then any real limits. For example, you could create a "use sword skill", a "cast ice spell" skill, and have having both of those unlock an "Ice sword" skill, while leaving the "improve sword skill" and "improve ice spell" options open to unlock later trees. It's almost the same approach, but I think it assumes much less work, since it reduces unnatural combinations. That's the approach I think you're going for, just based on info you gave, and I think it's a sound approach.
So this isn't exactly what you want to hear, but I'm gonna come at it from a different angle. I like to play games that aren't carefully balanced to have different classes match raw output numbers. What I like is games where the content is balanced to play to the strengths and weaknesses of different builds. For example, classes that can do things that other classes simply can't. Modern MMOs often break this down into 3 "things". Healing, tanking, or damaging. Some MMOs break down the roles of those three even further based on traits like specializing in multiple or single target in their healing/tanking/damage dealing. wrt "pure healers ending up best" or whatever, that doesn't have to be the case. There are ways to design a game to be balanced so that pure healers are more powerful at healing than a non-pure healer is, at the expensive of extreme frailty, no damage output, etc. The trick is making diminishing returns at the apex of the design. IMO the sweet spot for the build should be completely balanced from the perspective of being a fairly stand alone type player.
An example of a mechanic where a healer could be a worse healer based just on the numbers, but perhaps more useful in some cases is a build where damage dealt is converted to healing to the lowest party member. Not only is it perhaps dealing decent (if on the low end) damage but the "smart" aspect of the heal makes it perhaps more responsive to the healing needs of the party than manually targeting may be.
While I would definitely agree that pure mechanical balance isn't necessarily something to strive for, I consider balance of fun to be important. To me, this means relatively balancing the strengths and weaknesses like you said, but it also means that the core mechanics of each class (in a class based system) or skill (in a skill based system) need to be fun. It's a bit less important in a skill based system for everything to be fun, since people can just use the mechanics that are fun. In a class based system, especially if there isn't much role overlap, making each class equally fun is more important.
Of course, as a game designer, I should be striving to make every element of my game fun. It just seems like skill systems feel more forgiving in that respect.
Ok. Ive attempted this post a couple times, but its get very lengthy and im not even half way done. So lets try this again. Its gonna be short and to the point. With that in mind some of this may come off as rude. Its not ment that way, just blunt cause i hate typing. This all takes in mind your goal is a classless system.
1) You skill/spell break down. If a skill or spell falls into the fluff, fun, or weak spell category. Toss it, your doing something wrong. You should have any skill or spell that is there for the sake of being there. Either re-work it into a useful, spell people will want, or just get rid of it. Weak spells shouldnt be weak, but instead grow in power over time just like the caster. Filler skills and spells are bad for the reason you said, no one wants them, so they arent gonna use them. A classless system just makes it easier to avoid these waste of space things because you dont have to take em and train em. If you had them in a class based mud, sure people would have them, but im betting their use would still be minimal.
2) Classless systems balance themselves if done properly. The more skills your working on learning and raising and being proficient in, the more time you have to invest to stay on the same level as someone who is only training a small segment of skills. So person A working on a 'master' build given 10 hours may be level 2, where as person B working on a single class specialized build might be level 6 in that same 10 hours. That is the trade-off you should work to have with your skill system. This way given the same length of time the 'master' build person is never as powerful as the 'single build' person. Now if the master build person puts in more time so as to have all their skills on a comparable level as to the single build person, welll then yeah they are WAY more powerful, but more time was invested.
3) The above never makes anyone happy, cause there are always the 'power gamers' that go with master builds and play crack-headed amounts of hours. Thus making them EXTREMELY powerful. This is no flaw of the system, it is working as intended, this person just had WAY more time to put in then everyone else. Sadly the answer is to the complainers to get over it an invest more time if they dont like it. They of course wont like this cause they invest what time they want/can and are unrightfully upset this other person has more time and has thus accumulated more power.
Time investment is the only thing that can properly balance a classless system where your letting everyone do everything. Think about it logially as well. Why do we have specialized jobs in life? Cause thats the time we can invest. Whats stopping me from being a doctor and a car mechanic? I dont car to invest the time to do both… but if i did, i could surly be both and be good at both.
Ok. Ive attempted this post a couple times, but its get very lengthy and im not even half way done. So lets try this again. Its gonna be short and to the point. With that in mind some of this may come off as rude. Its not ment that way, just blunt cause i hate typing. This all takes in mind your goal is a classless system.
No problem :). I know how you feel. The original post was almost scrapped when I realized I actually wanted to ask a different question. I figured what I had written would also get me good responses, which it has. And I'll try to read your post in the best possible light (my standard, anyways, since I find message board posts that don't go out of their way to be polite usually sound rude anyways, regardless of intent).
1) You skill/spell break down. If a skill or spell falls into the fluff, fun, or weak spell category. Toss it, your doing something wrong. You should have any skill or spell that is there for the sake of being there. Either re-work it into a useful, spell people will want, or just get rid of it. Weak spells shouldnt be weak, but instead grow in power over time just like the caster. Filler skills and spells are bad for the reason you said, no one wants them, so they arent gonna use them. A classless system just makes it easier to avoid these waste of space things because you dont have to take em and train em. If you had them in a class based mud, sure people would have them, but im betting their use would still be minimal.
While I agree with you in principle, the reality, I think, is that it will be impossible to make all spells/skills equally desirable. Eventually, people will start making a mental list of which skills rank higher then the others. This is especially true if you want to provide multiple ways to do the same thing (e.g. travel). I think the best that you can hope for is that a given player's list will differ from another's.
That said, very likely there will be lots of tweaking on my part to get all the skills as close to equal as possible. At least so that my players aren't consistently picking one set over the other.
2) Classless systems balance themselves if done properly. The more skills your working on learning and raising and being proficient in, the more time you have to invest to stay on the same level as someone who is only training a small segment of skills. So person A working on a 'master' build given 10 hours may be level 2, where as person B working on a single class specialized build might be level 6 in that same 10 hours. That is the trade-off you should work to have with your skill system. This way given the same length of time the 'master' build person is never as powerful as the 'single build' person. Now if the master build person puts in more time so as to have all their skills on a comparable level as to the single build person, welll then yeah they are WAY more powerful, but more time was invested.
I guess I should specify that I'm looking to set a total skill cap below having a max value in every skill. The player will be able to change focuses over time, but should never be walking around with everything maxed out, regardless of time put in.
It's this auto-balancing effect that I want to harness, but also what scares me. People will obviously gravitate towards the powerful options, but I need advice on detecting and stopping the "master build". It's ok if players select powerful skills as long as they aren't all picking the same ones.
3) The above never makes anyone happy, cause there are always the 'power gamers' that go with master builds and play crack-headed amounts of hours. Thus making them EXTREMELY powerful. This is no flaw of the system, it is working as intended, this person just had WAY more time to put in then everyone else. Sadly the answer is to the complainers to get over it an invest more time if they dont like it. They of course wont like this cause they invest what time they want/can and are unrightfully upset this other person has more time and has thus accumulated more power.
This is the primary motivation for the skill cap I mentioned. In my end-game, I hope to have specialists for groups of skills, as well as generalists who do fairly well at everything. I also want to promote retraining when players get bored of their current skill set. I don't think it should be free, but it also shouldn't take as long as the initial grind (unless the player is trying to change every point they've set to a different one). For this to work, each option needs to be fairly unique (no sense of "you've tried one, you've tried them all"), and generalists need to be a little more powerful then they would be ordinarily since versatility usually doesn't beat specialty (This can be done by making the highest level skills significantly more costly, so the generalist can get reasonably high in multiple trees for the same cost).
1) Try to have as many powerful/useful/unique skills as possible. If there are many equally good options, players will differentiate on personal preference. This is terribly hard to execute 2) Have a class "lite" system where sticking to a paradigm is rewarded by power in the endgame, but is not necessary. You can mitigate some of the inflexibility by sprinkling relative powers into the lists all the way through.
Are there any other options?
3) You arrange the abilities in a skill web, with multiple groups of mutually-exclusive abilities that are intentionally a bit stronger than the others. Players will generally pick one ability from each of these groups to build their character concept, and then customise further with auxiliary abilities.
This is much easier to handle than the first option, as you only need to balance abilities against other abilities in the same group. These "core" abilities also provide a good starting point for newbies, who can focus on a small number of essential choices at the beginning rather than wading through hundreds of equally viable choices. But because the core abilities are divided into multiple groups, and can be extended in unique ways through auxiliary abilities, you still end up with a lot of variety among players.
The real answer here is, you can never stop the cookie cutter builds. They are driven by power gamers. There is ALWAYS a best build to go with, the power gamers will figure it out, and others will follow suite to keep up with them.
Beyond that, like i said its just trying to make everything so dang useful they want it all, but obviously cant pick it all. That will help drive diversity.
Skill trees as others have mention can help in this diversity, as it forces people to make decisions based on what deeper skills from the tree they want. This is still vulnerable to best build syndrome though.
Ill think on this more and possibly have some other input later today.
The real answer here is, you can never stop the cookie cutter builds. They are driven by power gamers. There is ALWAYS a best build to go with, the power gamers will figure it out, and others will follow suite to keep up with them.
For any sort of complex system there will always be some builds that are better than others, but if there's one "best build" then you should probably rethink your design.
I'm not convinced that a true classless system is worth it for anything other than a heavy RP focused game where viability isn't a concern. Once you have competitive gameplay then players will tend to gravitate towards a handful of optimal builds. This may not be that noticeable when dealing with the relatively small player numbers most muds get, but the experience from big MMOs bears this out.
I think the first question you really need to be clear about is why you want to have a classless system to begin with? Once you've identified what exactly you want to achieve I think it's easier to come up with a hybrid system.
EVE Online is a good example of a hybrid system where you can learn skills from any class but can only play one class at a time. This gives players a lot of freedom in which skills they learn and also makes it easier to switch classes without penalty (two traditional weaknesses of many class systems) but avoids the balance problems of a free skill choice by forcing players into a class role when they select a ship to pilot.
I'm not convinced that a true classless system is worth it for anything other than a heavy RP focused game where viability isn't a concern. Once you have competitive gameplay then players will tend to gravitate towards a handful of optimal builds.
By breaking the stronger options into separate groups of mutually exclusive skills, you can increase diversity, because players will use different combinations of abilities.
For example, supposing each player can select one of the five weapon specialty skills (swords, axes, clubs, polarms or flails), one of the five fighting style skills (berserk rage, formal duelling, assassin training, martial background or streetfighting) and one of the five protective skills (shield expert, armour proficiency, evasion, stamina or iron will). If you can balance each skill with the other four skills in the same group, while making sure that each skill has a strong impact on gameplay, and avoid crossover between the groups, you'll now have now 125 viable warrior builds.
On top of that you can add auxiliary skills, which are slightly weaker, but which allow the character to be extensively customised in smaller ways. Warriors will still be fundamentally built around one of the 125 "optimal" builds, but can be made to look, feel and play as if there's a lot more variety. Players won't feel as if they're being pressured into a specific role.
The real answer here is, you can never stop the cookie cutter builds. They are driven by power gamers. There is ALWAYS a best build to go with, the power gamers will figure it out, and others will follow suite to keep up with them.
Beyond that, like i said its just trying to make everything so dang useful they want it all, but obviously cant pick it all. That will help drive diversity.
Skill trees as others have mention can help in this diversity, as it forces people to make decisions based on what deeper skills from the tree they want. This is still vulnerable to best build syndrome though.
Ill think on this more and possibly have some other input later today.
How about a system where the baseline of skills segments (to throw a number out there, levels 50 out of 100) are reasonably easy to obtain, and it's expected that every player obtain level 50 in most skills (not required, but easy enough that they will). The first 50 levels are "free", in the sense that they're capped only by the time it would take to maximize them all. However, after level 50 the player must distributing points which are capped. So, for instance, the player could have 4 skills maxed out (for 200 points), or 20 skills above average (also for 200 points), but not 5 skills maxed out (which would cost 250 points and be over cap). From a design perspective, there should be powerful effects placed around the 50-60 range, more powerful effects placed around the 70-80 range, and some really, really nice powers around the 80-100 range, particularly a capstone at 100. Skills at similar levels should be similar in power, though the real concern is that the capstones are equally valid and useful.
For instance, you have the lock picking skill. Each level in the skill up to 50 gets you from simple locks all the way to moderately difficult locks. From there, further development will let you see and unlock magical locks of increasing difficulty, until you reach the capstone which allows you to unlock treasures that players who didn't master lockpicking are not capable of possibly even seeing. Silly, kind of boring example, but it illustrates where I'm going with this. Most people in the world will be able to unlock doors, but the master lockpicker will be able to unlock inter-dimensional portals, that sort of thing. The key will be designing each skill so that they're worth going through the trouble for.
I think the first question you really need to be clear about is why you want to have a classless system to begin with? Once you've identified what exactly you want to achieve I think it's easier to come up with a hybrid system.
There's a couple of reasons I'm leaning in this direction: 1) My source material doesn't lend itself well to classes. The characters in the world tend towards being capable of doing ok at everything, but tending to do better at a couple of specific things, mostly out of preference for practicing those things, and it wouldn't make sense for most characters to be terrible at most things, but incredible at one thing (This is certainly possible to achieve, but it shouldn't be normal. There are less mages and warriors then people who swing swords and cast the occasional spell, and people who cast spells but aren't bad with a blade in a pinch) 2) I dislike (not strongly) "pure" class systems where all decisions are made at character creation. I know these are rare, with most people (designers) favoring some kind of specialization within classes, but I state this as the extreme of my view. I can appreciate the benefits of strict classes, but I wouldn't want to play or design a system using them. 3) I'm enamored with the idea of people developing into archetypes simply because they enjoy the mechanic behind them. The "newbie" player is a blank slate, who becomes a warrior not by joining the warrior guild or selecting warrior at the start, but because he picks up a sword and starts hitting things with it. If he sees his friend the mage, thinks "That looks cool, I want to be a mage", he can slowly migrate his skill to being magelike just by acting like a mage. (For sure, he'll be a really terrible mage at first, but it will be possible for him to make the change). Suppose then he stops partway, and progresses as a warrior who slings the occasional fireball, that should be ok too.
EVE Online is a good example of a hybrid system where you can learn skills from any class but can only play one class at a time. This gives players a lot of freedom in which skills they learn and also makes it easier to switch classes without penalty (two traditional weaknesses of many class systems) but avoids the balance problems of a free skill choice by forcing players into a class role when they select a ship to pilot.
If it were possible in my design for players to simply swap "bodies" every time they wanted to try something new, I would definitely consider it. (Actually, in the strictest sense it will be possible, but it's likely to be a specialist ability rather then something players can just do from the outset).
For example, supposing each player can select one of the five weapon specialty skills (swords, axes, clubs, polarms or flails), one of the five fighting style skills (berserk rage, formal duelling, assassin training, martial background or streetfighting) and one of the five protective skills (shield expert, armour proficiency, evasion, stamina or iron will). If you can balance each skill with the other four skills in the same group, while making sure that each skill has a strong impact on gameplay, and avoid crossover between the groups, you'll now have now 125 viable warrior builds.
On top of that you can add auxiliary skills, which are slightly weaker, but which allow the character to be extensively customised in smaller ways. Warriors will still be fundamentally built around one of the 125 "optimal" builds, but can be made to look, feel and play as if there's a lot more variety. Players won't feel as if they're being pressured into a specific role.
That's an interesting approach to it. If a class can be seen as a vertical design (where you pick a set of predefined skills), your idea is horizontal (You have X slots to fill, build your class from these options).
I'm concerned that it appears to presume classes already (For example, a mage likely won't get access to a weapon specialty skill). You could probably design your slots such that they're generic enough to support multiple different styles. For example, you could have a "damage" slot and a "buff" slot, where the options for the former are "sword mastery" and "fireball", and the options for the second are "critical hits" and "fire damage". So that gives you the option to have a sword which does critical damage occasionally, a flaming sword, a fireball that does critical damage , or a fireball that does more fire damage then normal consistently. You can then build in exclusivities, so that you can't have a frozen fireball (Though the damage could be "energy ball" pretty easily, with the type determined by the buff). This won't work for all things, where do things stealing and lockpicking fit in, but it's an interesting idea to build on. It probably won't work for what I have in mind, though.
I'm concerned that it appears to presume classes already (For example, a mage likely won't get access to a weapon specialty skill).
I was trying to keep the example simple, to show how a single traditional character concept (in this case "warrior") could be represented through skill groups, therefore avoiding the "optimal build" problem.
However you could certainly extend it to magic as well, either by adding some "spell" weapon skills (like your fireball example), by making the skill names more generic, or by creating a separate group of core magic skills and balancing those against the combat skills.
In my mud, for example, I handle magic through talents. The Magically Gifted talent gives you the ability to cast spells, but it's classified as a "heritage" talent, and you can only have one heritage talent.
Each heritage also unlocks a few "lesser calling" talents, but you're limited to just one of those as well. Magically Gifted unlocks the four elemental Adept talents for example, but you can only pick one. Or you could take the Dark Lineage heritage talent and the House ShadowSworn lesser calling, making you a dhampir with access to shadow magic. Or Feral Spirit and Spirit Wolfkin for spirit magic, or Tainted Soul and Witch Spawn for infernal magic, etc.
But spending your lesser calling slot on a magical talent means you can't take any of the other lesser callings, and they each offer powerful bonuses.
Quote
This won't work for all things, where do things stealing and lockpicking fit in
They'd be auxiliary skills.
The idea of this approach is to help balance different builds in combat situations, assuming that (as with most muds) combat is a major part of the game. Of course it's possible that combat may play only a minor role, or even no role at all, in which case you should change the core skills accordingly. If the main focus of the mud is crafting, for example, then the core skills should instead ensure that each character has a solid set of crafting abilities.
If you can balance each skill with the other four skills in the same group, while making sure that each skill has a strong impact on gameplay, and avoid crossover between the groups, you'll now have now 125 viable warrior builds.
But avoiding crossover seems like a lost cause. There'll always be synergies between skills in different groups, and I'm not sure this is something you want to discourage anyway.
There's always a lot of talk about balancing a system you're designing, but to me this seems backwards, as there are so many factors (not even mentioning the particular playerskills of your playerbase) that a more fruitful approach would be to get it approximately right in the design and then profile the system to make changes. Granted, constant nerfing and revision of skills in a live game is hardly the ideal solution, but it may be the only practical one. Depending on the game, you might be able to make this feel more natural by making sure resource pools and eq play a large role in character effectiveness. This will have the dual effect of making the game more dynamic as well.
If you can balance each skill with the other four skills in the same group, while making sure that each skill has a strong impact on gameplay, and avoid crossover between the groups, you'll now have now 125 viable warrior builds.
But avoiding crossover seems like a lost cause. There'll always be synergies between skills in different groups, and I'm not sure this is something you want to discourage anyway.
There's always a lot of talk about balancing a system you're designing, but to me this seems backwards, as there are so many factors (not even mentioning the particular playerskills of your playerbase) that a more fruitful approach would be to get it approximately right in the design and then profile the system to make changes. Granted, constant nerfing and revision of skills in a live game is hardly the ideal solution, but it may be the only practical one. Depending on the game, you might be able to make this feel more natural by making sure resource pools and eq play a large role in character effectiveness. This will have the dual effect of making the game more dynamic as well.
In order to further analyze this question. I feel i need to have a better understanding of the skills you are working with, and how the increase.
For instance. Do you have a ROM style approach where you "learn" a skill, then work on mastering it up to 100%? or is it a rank based system where your skills gain ranks with use, and there is no correlation between your rank and a percent of mastery. Rank being more just like a rating of how good you are, higher number the better.
Can you give us an example of some of the skills you have. This might help to better determine a proper way of attempting to avoid a cookie cutter build.
Well, like I said in the original post, it's not quite relevant to me yet. That is to say, I haven't completed planning the system. I'll try to give an example of what I currently have in mind, but it'll be subject to change on a whim.
Firstly, the character will have 3 types of experience pools. General exp, skill group exp, and skill exp. Skill exp is earned by doing a skill. Over time, some of the skill exp will turn into skill group exp, and the player will find that they've gained the ability to perform related skills better. Also, some experience from the group exp will turn into general exp, which will tend to have a similar effect to leveling in other games (all your skills get better). The 3 pools have maximums of around 30, 50 and 20 respectively (Just to throw easy numbers out there). The game will be focused mainly on mages, so general experience can be considered general magic power, if that makes more intuitive sense.
Every skill in the game will have a group, and a minimum and maximum effective value. For instance, a really basic skill, the effective range will be from 1 to 10, while a moderately difficult skill will have an effective range of 10-25. This means, roughly, that you have a 10% chance of succeeding at the easy skill if you try it without any experience at all, and a 0% chance of succeeding if you try the same with the medium skill.
The success formula is min<= (genexp+skillgroupexp+skillexp+Rand(0, max-min)-(max-min)+1) (< max )
The last part is parenthesized, because you only get experience for results between the two numbers, but the skill will succeed any time you get a result higher then the minimum. The general experience and skill group experience become important here, since by the formula, you can only start learning a skill if genexp+skillgroupexp > min. Also, the result of the formula, up to max, minus min, will be passed to the skill in question as a power parameter, so the first skill can have a power range of 1 -10 (or 0 to 9 if you prefer), and the second can have a power range of 1-15.
You might be wondering how I can say I don't have the system planned, if I have a formula for how it will work? The answer is that I haven't really begun planning skills, or putting them on my number line. I'll try to come up with a brief sample, to give you an idea of how it should look. Keep in mind that the numbers are arbirary
(min, max)Transformation: //a skill for changing one object into another. Just a note, this is a feature of the MUD. Most objects in the game will be non-trivially mutatable (0, 10) Transform solid //Change one aspect of the target (0, 20) Transform solid two parts (10, 20) Transform simple fluid (10, 20) Transform three parts (10, 30) Transform fluid two parts (10, 30) Transform solid 4 parts (20, 30) Transform solid 5 parts (20, 40) Transform fluid -> solid (20, 40) Transform solid -> fluid (20, 40) Transform fluid 3 parts (…) it's actually hard coming up with things that aren't just more of this (100, 110) Transform anything -> anything (Note that even at max power we have a 1/10 chance of success to mitigate the game destroying potential of this a little :)
A simpler example would be herblore: (0, 100) Identify - reveal information about the plant (10, 100) Appraise - reveal quality of the the plant (e.g. ripeness) (10, 100) Diagnose plant - reveal maladies present in the plant (10, 100) Harvest - breaks a plant into it's useful parts
Note that the skills are all X-100. This is because the actual min-max will depend on the plant. In effect, there would be as many harvest skills as there are plants in the game. This is not a terrible thing, since people going down this path will be able to deal with the easiest plants through general and skill group exp alone, and use actual skill exp for the plants that have a difficulty greater then 80.
The problem that I'm having is that the transformation skill may be more interesting to players then the herb gathering skill, so I'm looking for strategies at keeping them relatively equal in fun/utility. The obvious problem being that it's subjective :). It also might be tricky to analyse, since none of the skills I listed are trivially identifiable as "fun". I really want to say that the issue is that they're off the top of my head, but I really want to code/playtest/tweak them if they aren't fun. For example, plant harvesting might need a minigame where inherent skill is a factor, but ultimate success results from performing the right actions. Or it might be good as just a % success skill.
I was trying to keep the example simple, to show how a single traditional character concept (in this case "warrior") could be represented through skill groups, therefore avoiding the "optimal build" problem.
The skill groups themselves certainly don't solve that problem. They limit the number of overall combinations of course, and they may provide a starting point, but you can't balance every combination simply by balancing individual skills within each group. The exception would be if the skills in each group were just synonyms, or there was no synergy at all between groups, but that wouldn't make for a very interesting system at all. Even in the simple example you gave it's obvious that polearm + shield expert is less optimal than other combinations.
But avoiding crossover seems like a lost cause. There'll always be synergies between skills in different groups, and I'm not sure this is something you want to discourage anyway.
For something simple like my example it should certainly be possible to design three groups of core skills that don't have any crossover with each other. I agree that crossovers can be a good thing, but they can be covered by auxiliary skills.
The skill groups themselves certainly don't solve that problem. They limit the number of overall combinations of course, and they may provide a starting point, but you can't balance every combination simply by balancing individual skills within each group.
The point isn't to balance every possible combination, but to provide a solid foundation upon which the rest of the character is built. Balancing a skill against four other skills is certainly achievable, and if those five skills are stronger than the other skills they'll provide the core around which character concepts are created. Offer three separate skill groups and you've got 125 basic builds, which can then be extended through auxiliary skills.
The exception would be if the skills in each group were just synonyms, or there was no synergy at all between groups, but that wouldn't make for a very interesting system at all.
You only avoid synergy between the stronger core skills, not the auxiliary skills.
Classes are useful tools. They define sharp boundaries that make balancing easier (Not easy, mind, but easier). If you have 5 classes with 40 skills unique to those classes, then you have 5 things to balance, and 200 things you can tweak to achieve it, where as if you just have 200 skills and allow the player to pick any 40, you have an absurd number of possible builds to deal with.
The problem you have won't be "Are all possible combinations of skills equally viable", since that is a pipe dream. The only feasible way to achieve that is have all skills just be clones of each other, dressed up to look differently, and that's boring. The problem is that skills can be divided into tiers, like this:
- Useful: This is a skill that makes some aspect of the game easier once you possess it. It can be a travel skill, or an appraising skill
- Powerful: These are the top notch combat skills.
- Fluff: These are theoretic useful skills, but aren't really useful, or are overshadowed by better options
- Weak: These are combat skills that are overshadowed by the better combat options
- Fun: These skills serve little to no purpose in or out of combat, but have some mechanism that make them interesting. Note that all the other skills could be "fun", but for the purposes of this post we'll stick with the specific definition.
In a class based system, you can provide a mix of the above skills to each class, giving them variety and options. Even the weak and fluff skills can be tweaked so that instead of "useless", they're just situational.
In a classless system, people will just pick skills from the first two, and sprinkle in skills from the last one, while ignoring anything fluffy or weak. This would be fine, except if your skills are evenly distributed, and you have 200 skills, fully 80 of them will not be used by players at all. And they aren't usually evenly distributed, because it's difficult to come up with 40 unique and equally powerful skills, or 40 unique and useful skills, so the distribution actually looks more like 40 "good" skills, 20 "fun skills" and fully 140 weak options. That means that instead of unique diversity, most players will have a core of 20-30 must have skills, and diversify themselves with the leftovers. Things start to look a bit samey, if everyone has a horse because it's the superior travel option to teleportation (or vice versa).
One of the ways to reduce this is to have skill paths, where one must accept X less powerful skills to get at the good skill. This encourages a degree of specialization, and encourages players to be creative with the extra skills, since they only get so many of them. It also forces more meaningful choices. A good example with this is IRE games. They are class based, but only 3 skill lists out of many available to any given player are class specific. The none class specific skills are very useful, so deciding how much to allocate to those vs your class skills is a meaningful choice. If the MUDs were to go class free, it would be even better, since the "must have" skills fall at the end of the lists for class skills (This is done for economic reasons as much as anything, I'm sure, but it's still a good example). The problem is that if you have too much forced specialization, you start to have things that look like classes again. If to be a competent healer you need to focus on healing at the exclusion of everything else, you might as well just have a healer class. Generally, having skill lists (or trees) and putting the good ones at the back is a good way to ensure that anyone attempting to take multiple trees halfway gets screwed over by having many options but none that are good, and if you don't support those players, then you're creating an implicit class system, the thing you're trying to avoid in the first place.
So where do you draw the line? The options I see are:
1) Try to have as many powerful/useful/unique skills as possible. If there are many equally good options, players will differentiate on personal preference. This is terribly hard to execute
2) Have a class "lite" system where sticking to a paradigm is rewarded by power in the endgame, but is not necessary. You can mitigate some of the inflexibility by sprinkling relative powers into the lists all the way through.
Are there any other options?
Also, suppose you're picking option 2. Are there any guidelines into making each tree attractive, so that you don't get everyone following the magic path and not have any melee users? I suppose this is the same question about having unique but equally viable and interesting classes, extended to a more free-form system.
*I should mention that the numbers provided don't represent anything about the MUD I'm planning. I haven't begun planning, let alone coding skills