Classes

I like the flavor that explicit classes bring to games' mythos/universe/story but I find myself always favoring the skill systems of games like Ultima Online with their flexibility and better-by-usage approach.

I think a nice marriage of these systems is something that begins as the UO system and has opportunities at a later stage of the game to opt-into a "class" system.
Something like completing a task to initiate yourself as a... Paladin? or something which grants specific abilities to your skill set that match the flavor of the job/class/archetype but maybe hinders other ones, based on flavor.

This way a base skill-setter isn't persuaded to ALWAYS opt-in to one since the restrictions maybe be seen as too big a negative.

I'm rambling a bit. Just wondered what others thought of class systems, ones they've liked, disliked, etc.
«1

Comments

  • Hi, I've decided AGAINST using classes for now. The reason is I want a game that can be quickly launched, where you create your character AS you play without spending an hour doing it while booting up a new game. I also think you can get more interesting permutations by not limiting to specific class types. That said, it may try to autodetect your closest class just so NPCs can identify you ("greetings, Ranger" - etc).
  • I hadn't thought about the game-start overhead, point taken.
  • having them point to your closest class would be good, but I think it could also go through a line of how popular you are.
    For example say your more towards a mage but often use a bow to attack.

    If your often doing quests in or around this town, some of the folk could have heard how this mage was also good with a bow.

    "Hello there mage, I have heard word that you are quite handy with a bow."

    Even if person doesnt give you a quest, it would be nice to see word spread around.
  • That kind of nuance might be pretty difficult to obtain, but along your same point having at least a binary noun for each archetype that carries a positive and negative connotation would be a simple addition I think --> "Mage" vs"Warlock", "Sorceress" vs "Witch", "Monk" vs "Brawler", "Knight" vs "Sellsword", some stuff like that, although binary good/evil system usually fall flat when they're leaned on heavily (Fable).
  • To be entirely honest, I've never really cared much for classes either. I like games like Skyrim/Morrowind/etc. where you decide what you are as the game progresses, and aren't quite so confined. In those types of games, if you want to min/max, that's great, and you'll be better for it, but you don't have any restrictions on what you're able to do, either. It basically removes any penalties/prohibitions from trying to branch out with your skill set.
  • I find that many games end up using classes as restrictions to how you make your character rather than a tool to help build them. I prefer classes to be descriptive; "you have these stats high and tend to fight with this weapon, so we will call you a X", rather than "You are an X, so you get these stats and fight with this weapon".
  • edited October 2014
    Talvieno said:

    To be entirely honest, I've never really cared much for classes either. I like games like Skyrim/Morrowind/etc. where you decide what you are as the game progresses, and aren't quite so confined. In those types of games, if you want to min/max, that's great, and you'll be better for it, but you don't have any restrictions on what you're able to do, either. It basically removes any penalties/prohibitions from trying to branch out with your skill set.

    What you describe is only Skyrim. The older TES games had all a class based system.

    Ultima Online had it right IMHO. Except maybe for the skill/stat cap. Lose that, and it would fit perfectly for a rogue-like.
  • As usual, design is still in the air, but these are my initial thoughts:

    It's been a perpetual battle for designers - class-based or skill-based. Each has unique advantages / disadvantages. In the base game design, I'm probably sticking with classless, but that doesn't mean I won't support classes. As mentioned, there is a lot of interest for people implementing their own games on the engine, and I can't tell them which is right/wrong.

    Skill-based does not have to mean muddled or uber-classing (making an all-powerful character who can do everything). In fact, the beauty of the system I am planning is that you only have slight choice as to how your character is going to develop.

    At first thought, I would have determined such a system to be crap. But I played an iOS game about a year ago that changed my thoughts on this (Desktop Dungeons also provided additional influence). The game was a roguelite called Hopilite for iOS:
    http://toucharcade.com/2014/01/08/hoplite-review/


    The game contains shrines on each level which allow you to pick one of three available buffs/skill increases. Its flexible enough that you don't feel you are being robbed of choice, but rigid enough that it forces you out of your comfort zone and intelligently pick among what is available to you, rather than resorting to the same old skill picks every time (guilty - I always pick a sniper class if I can in RPGs because I'm cowardly like that).

    Similarly, desktop dungeons lets you pick a class, but the available runes on the level are random, so you get to pick a few base skills, but you don't know what else you will end up with. To further this, each level has 3 or so random shrines that further vary your choices.

    Whether a roguelike or not, games will get stale if they give you too much freedom. The desire to play one more game at the chance that you will have those ideal skills available is a strong draw when your choices are always limited. This is the foundation of what makes CCGs fun - you are not in complete control of the circumstances / what is given to you, but you must find creative ways to work around what you are given. That said, it is fine line you must walk. There must be enough control in your game so that a player feels empowered by the choices they make, but no so much control that they can lay out the game an ideal manner.

  • In my own study of game design, I found the thing that would make character building stale was imbalance. for a beginner player it doesn't matter much, as you will explore the various options freely. However, as system mastery sets in, subpar choices start being obvious, and they drop away as viable options. Similarly, overpowered options become obvious, and start to dominate the landscape. With an imbalanced game, a masterful player is only working with a tiny subset of the choices offered, and it becomes stale. In contrast, if the game is well balanced, all options remain viable, and trying out novel combinations retains its appeal.
    Using some TTRPGS as examples, D&D 3.5 gave you many, many options to build your character with, but they were extremely unbalanced against each other. By the time you filter out everything of an inappropriate power level (yes, too powerful can be a problem too, esp. in the group context of D&D), your actual options are very sparse. Most things you would try to do hit the wrong power level, the group you are playing with has difficulty coordinating their power with each other, and the game suffers for it.
    In contrast, I now play Legend, which has far fewer options, being only a single rulebook compared to the many, many expansions D&D has gained over the years, but it has a far better balance. The game feels more flexible, and more balanced, so even after years of playing it and gaining system mastery, nearly everything in it still seems like a valid choice, and I am still coming up with fresh new combinations on a regular basis.
    Balance is key. The disadvantage of a more flexible system is that balance becomes harder, the more ways the player can tweak things the more likely it is they do something you didn't anticipate and cause issues. However, when you can pull off a system that is both sufficiently flexible and sufficiently balanced, the results are amazing.
  • Just wanna say I play Legend as well and love that rules system.
  • edited October 2014
    I need to get @Flatfingers over here. lol This sounds just like the kind of discussion he'd enjoy...
  • edited October 2014
    Watch this. Wait for it... Wait for it...

    disclaimer: he may appear elsewhere first
  • Hello! Hello! :D

    I don't want to start flinging out detailed ideas specific to VQ until I feel like I understand Gavan's preferred design direction first. That said, I can maybe offer a couple of quick general notes.

    1. Morrowind and Oblivion actually had a kind of crossbreed ability system. When defining your character, you could specify five major and five minor skills from a reasonably large set of skills. If you wanted, you could speed up this process by selecting a "class" that had these major/minor skill selections pre-defined. But gamers who enjoy character creation could define (and, at least in Oblivion, name) their own collection of major/minor skills. Gaining a level in one of these major/minor skills then contributed to increasing your overall character level. (The notion of "character level," BTW, deserves its own thread, mostly to discuss the effects that mechanical structure has on roleplaying.)

    I've long thought the TES solution was an excellent middle way between a pure unbounded skills system and a hard class system in which skills obtained through progression are pre-determined.

    Question: should the freedom (for players) to choose whatever skills they want include the freedom to make "bad" choices? What makes a choice of some skill "bad?"

    2. This is just my personal design philosophy, and should not be considered a criticism of the approach to progression management that Gavan describes above.

    That said... I do have what I've called a "player-centric" game design philosophy. When I create a game design, I think in terms of systems that work together to achieve the overall design vision for that game, and that maximize the variety of allowed player choices for all challenges.

    In other words, I tend not to think in terms of control -- "I must limit you here for your own good" -- but in terms of giving good feedback prior to decisions so that players control themselves by making the choices that are the most fun for them. That way, even if the result of an individual choice is not optimal, the player still feels satisfied because the process itself was one that they naturally enjoy. For example, if I prefer exploration over achievement, then if the game lets me open locked containers in an interesting way I'm going to be happy doing that even if not every container has some awesome goodies in it -- the game was designed to allow me to enjoy the kind of process that's fun for me. (Naturally it also should provide choices in that same area of gameplay that appeal to achievement-oriented players.)

    Ultimately of course the developer has to set some limits. Even vast teams of programmers can't create all possible options for all possible gameplay activities. With good planning, though, I think a minimal team can get close enough to that player-centric nirvana through careful design of systems. When the core systems of a game are all created from the very start to try to maximize the variety of ways that different kinds of players can address gameplay challenges, then even if there are limits on the specific mechanical effects of a player action (as there must be), those challenges don't feel controlled because each player can solve them in the way they find most enjoyable.

    Another example, which occurs to me from Dishonored (and thinking of it in this context is probably not an accident, as I'll mention in a moment), is choosing how to get past a guard. Three obvious choices: 1) sneak up and stab him, 2) sneak up and knock him out, or 3) Blink past him entirely. Even though there are mechanical limits on how to accomplish those actions (which is a big part of the gameplay in Dishonored), players still have considerable freedom to decide which of those mechanics to apply. There is control over player actions, but it's at a fairly low mechanical level. At the next level up of tactical choice, the player in Dishonored has considerable freedom to choose what to do... and that respect for multiple ways of solving challenges feels incredibly empowering.

    As I said, I don't think it's an accident that Dishonored offers this feeling. It was made by Raphael Colantonio/Arkane, whose stated goal was to make games that feel like the RPGs from the much-missed Looking Glass: the creators of Ultima Underworld, System Shock, and Thief. Those games shared (I might even say popularized) the fundamental design concept of "multiple ways to solve challenges." That's a philosophy I embraced after doing some work of my own on synthesizing several models of general personality and styles of play -- these work together in that different playstyles enjoy different ways of solving gameplay challenges. When you (as a game designer) can see what makes these styles of play distinctive, and have a good toolkit for building specific play activities that appeal particularly to each of these styles, it becomes possible to create the challenges of your game so that they're systemically solvable in ways that are fun for different kinds of gamers. And ultimately that whole game ends up feeling (for the player) less like you've been controlled by the developer to have the kind of play experience they thought you should have, and more like you've personally created the kind of experience that's fun for you.

    I'm not suggesting that won't happen in Voxel Quest! This is just me thinking out loud at a fairly high (i.e., vague) level about ways that game designers can communicate through game mechanics with the people who come to play their game. In the end it's about what kind of game you're trying to make. Some games benefit from a tightly managed experience -- Valve's games tend to be that sort -- while others are fun because their loose mechanics are well suited to a highly dynamic world (of which Minecraft is a fair example). My only advice would be that the level of control imposed on players through class/skill mechanics probably feels most right when it's consistent with the feeling of openness or control that the world of the game itself promises.

    Which of these design paths (or some other path) makes more sense for Voxel Quest is a question only Gavan can answer.

    But we here can, I hope, discuss what we think are the possible effects of any of those choices of design paths, I hope!

    Also, yes, this is what I consider "quick general notes." :D
  • edited October 2014
    image

    Amusingly enough, I can tell you're holding back. :P I very strongly agree with everything here. I had a feeling it'd help to get you in here - you have a valuable knack for explaining things I couldn't even begin to put into words.
  • In regards to "should players be free to make bad choices", the general answer is yes. It is the flip side of players being able to make good choices. However, no choice should be added just to be bad. In the context of character creation, this means none of the character building options should be subpar. Combinations may be; avoiding the possibility of bad combinations is hard, even if you think its desirable, and attempting to do so tends to limit just as many interesting combinations. It is important to avoid combinations of abilities which are too powerful, but combinations which are sub par are too be expected, and it falls on the player to avoid them. That said, which combinations are poor should be obvious. If I put all my points into spellcasting then run around stabbing things with a sword, that mistake is on me. However, if I try combinations that should work well, like training my spell power and mana capacity, and then find that being a mage is nonviable , the system needs rebalanced.
    When designing a character building system, the aspects I suggest paying attention to are the flexibility and bance of the system. The more flexible the system is, the greater variety of archetypes are possible, and the more details within the archetype you can vary. Balance comes down to how many of those possibilities are viable, and how easy it is to make a overpowered combination. Ideally you have a flexible and balanced system.
  • edited October 2014
    I can't remember exactly where I read it, but I agree strongly on "No choice should be added just to be bad." In fact, I'd like to take that further: No choice should be inherently bad. There should be good situations for each choice, and the rewards for making the choice under the correct circumstances should equal the rarity of using it - within reason. If any weapon or statistic is something that nobody would ever put anything into - if someone says "No, never put any points into this skill" or "this is a trash weapon, don't use it", then you, as a game designer, are doing something wrong.

    That's not to say all weapons and skills should be equally useful. Someone focusing on spellcasting, for instance, might not find putting points into Strength as useful as a melee-oriented character would. There should be different ways to approach any given situation, just as Flatfingers said. This also means that someone who's been playing as a mage for the first half of their game should suddenly find themselves useless upon coming across their first enemy that is heavily resistant to magic. By this, I mean that the game ought to make you balance out what you do so that you don't find yourself backed into this sort of a corner. How you do that can be achieved through different ways - perhaps making occasional enemies that are easier to kill through melee, or perhaps making it so that spellcasts need to be carefully managed to keep from running out of charges, or perhaps even making it so that the character needs a few points in strength if they want to carry larger spell tomes. The method is irrelevant, so long as the effect it has balances out the rest of the game and doesn't feel like a nuisance to the player.

    Another good example might be weaponry. Suppose you have a crushing/piercing/slashing mechanic. Crushing weapons might not deal damage for most of the game, but say you occasionally come across some particularly powerful enemies that are weak against that type of attack damage. In that instance, having a crushing weapon handy would be exceedingly useful, and might make the difference in whether or not you survive.

    On a slightly different subject matter... Something that could be exceedingly helpful to discuss here is the matter of choices and problems. Both of these are decisions, and at the most basic level of game design, we're taught that all games are made up of a series of decisions. And this is true - but what most people don't understand is that there's multiple types of decisions that a player can make.

    A problem is when the player is presented with a number of options, and must choose the best one. Things like deciding whether a +5% resistance to fire charm or a +5% resistance to poison would be better is considered a problem. Another problem might be, "I'm a mage, so do I level strength or intelligence?" or "Do I spend my money on health potions or do I buy a new piece of armor?" In some cases, the solution might not be immediately obvious, but the outcome is ultimately quantifiable. These are unavoidable - but not necessarily bad - parts of any game. In fact, it's safe to say that the great majority of decisions the player makes during any given game is made up of problems, and some people even prefer games that only feature this type of decision.

    A choice is when the payer is presented with a number of options, but there is nothing then, nor later, to indicate which is the best. These are extremely important, as it leads the player to customize their experience, and they're forced to think about their options instead of flying through on "this has better armor class". It leads to a more immersive, meaningful experience - but only if the choice has a very clear impact on the game. Here's two good examples: "Should I make camp outside of town and risk being attacked, or should I try to fight my way there to get a better night's sleep?" or "Do I talk to the enemy " Even if you load your old save and pick both routes, the outcome is different each time, and as long as one way or the other isn't a "better decision", it's considered a choice. Another choice might be, "Do I make this character Good, or Evil?" Unfortunately, this doesn't apply to some games, where being evil (or good) has quantifiable drawbacks - in these games, "good or evil" becomes more of a problem than a choice. Choices are often purely story elements, but it's usually best (and hardest to create) when they have some effect on gameplay.

    But back to my good/evil example. The reason it's important to distinguish between the two is because it's all too easy to make something that was supposed to be a choice (say, choosing to save or kill the little sisters in Bioshock) into a problem. Perhaps you don't know on your first playthrough, but you only get the good ending if you don't kill any of the little sisters, and you get the most ADAM overall if you don't kill any, too.

    ~~~

    After typing all that, I decided to look up what I was thinking about and found an article on it, courtesy of Gamasutra. http://www.gamasutra.com/php-bin/news_index.php?story=22456

    There's another article I'm thinking of expressly related to "no weapon should be a bad choice", but I think Flatfingers may have the appropriate article for that... I know I read one somewhere about that topic. I think it was somehow related to the development of Dark Souls/Demon Souls, but I'm not entirely sure.
  • I agree with tavlieno. I really like the problem vs choice dynamic he layed out. That clarified a few things in my mind. It therefore becomes important to know what should be a choice and what should be a problem.
    I would say that the archetype of character you build should be a choice, with obvious differences between them and no right or wrong answer, but how you build that archetype should be a problem. So, a fighter is not a worse choice than a mage, but there are better and worse ways to build a fighter.
  • edited October 2014
    I agree. Choosing what to focus on - stealth, combat, magic, personality, crafting, etc. should be a choice with no clear "best" answer - but the way you get to the top in each should be the same - assuming the world is set up the same way each time. In a dynamic world, you're also faced with the problem of, "is it worth exploring farther to try to find this better item I know exists", thus upping the replay value.

    Here's that second article I was talking about, and yes, it was indeed courtesy of Flatfingers. He posted it on Limit Theory's Technical board a while back, and while I may not personally have commented, I most certainly read it, and it appears to have stuck. :P
    http://kotaku.com/how-to-balance-an-rpg-1625516832

    To quote his original post:

    Josh Sawyer -- the veteran game designer of Icewind Dale II, Fallout: New Vegas, Alpha Protocol, and the upcoming Pillars of Eternity -- has written a bang-up article for Kotaku on the technical process of balancing computer-based role-playing games (CRPGs).

    I think it's worth posting here for those who enjoy this kind of behind-the-scenes game development knowledge because it offers some valuable lessons for Limit Theory. LT is not a CRPG, exactly. But RPGs are usually highly systemic games -- there are a lot of moving parts that need to express the right dynamic behaviors. That's a characteristic they share with Limit Theory.

    So I thought I'd share his article here, as it is full of things to consider when balancing the many components of a systemic game to insure that every choice a player might make remains a fun one.

    For example:

    How do we tune? The Sid Meier way: doubling and halving. For many people, the instinct when tuning is to try small incremental adjustments. This is usually not an efficient process. It's almost always much faster to halve or double the values being adjusted and only make smaller tweaks after you've already passed the target. ... Once you've confirmed you've gone too far, you can roll back so you've gone just far enough.

    His article is rich with hard-won, practical advice of this kind for how to efficiently construct a game that is fun all the way up from the moment-to-moment mechanics to the feelings of achievement and aesthetic satisfaction when you're done playing.

    This isn't to suggest that Josh isn't thinking about many of these things already. But I'll bet even he would find some useful ideas here.

    In the meantime, if anyone here is interested in how you make a complex computer game enjoyable, you might like this article as well.
    http://forums.ltheory.com/viewtopic.php?f=6&t=3432&p=61104&hilit=+balanc*#p61104

    (Flat tends to be knowledgeable about this stuff, which is why I'd hoped to get him over here. :P He and Gavan had a huge, in-depth discussion over on the LT forums, and it impressed me how open Gavan was to new ideas.)
  • Talvieno said:

    I need to get @Flatfingers over here. lol This sounds just like the kind of discussion he'd enjoy...

    SUMMON THE KRAKEN

  • edited October 2014
    @Flatfingers‌ More excellent points - don't hold back though; I want to know what YOU would like to see in a game and I weigh this against community feedback.

    Also, this addresses some of @Talvieno‌ ' s comments as well

    It is another question for the ages: how much freedom should you give the player?

    Raph Koster (designer behind Ultima Online and Author of A Theory of Fun) kind of addressed this, but not directly, in his book (iirc) - not that he is the Oracle of Game Design, but I think he has some valid opinions. He had this theory - if the player has little/no control then the game is not fun (it is, in mathematical terms, just "noise"). Many people think games like this are fun regardless, like Bunco, which has no skill at all (100 percent luck).

    On the opposite end, you have games which offer 100 percent choice and no luck (i.e. Chess). I tend towards this type of game, but not completely. The downside to chess is that it always starts out the same (also an upside, depending on your preferences). There are several types of fun (adrenaline, vicarious/story, beauty/aesthetic, twitch / coordination, strategy, etc). The strongest, IMO, is about mastering pattern recognition, which is typically classified as "strategy." Once you have realized the pattern, a game becomes boring.

    Imagine if, in chess, every piece could move like a queen does. The game would be much more boring because it does not limit your control of pieces and diminishes the strategy involved. This is why Chess is generally considered a much more intelligent game than Checkers.

    Let's take another example: Hearthstone. Overall, a good game, but too much based on luck for my preference. The order in which you draw your cards can make or break a game, and since you have no control over this it is frustrating at times (but it does make you play the odds - i.e. create a deck that is statistically likely to work and whatever card you draw will tend towards a certain balance).

    And another example: Puzzle Quest. It is not entirely random, but it can be very difficult to calculate chain reactions in a match 3 game beyond the first two iterations - i.e. destroying some pieces, then having another match fall into place. At some point, it just becomes noise and getting a 5-x chain of matches is almost always pure luck.

    Which comes back to why I favor determinism. I think that it is ok to be presented with a random scenario (imagine a chess board that is randomly arranged and you have to make the best move that turn). But I think it is always best if there is a way to tackle the current challenge deterministically, and better still if you are never presented with a scenario that is impossible to solve (i.e. you will lose regardless of what choice you make).

    Against total determinism: Finding the jetpack in Spelunky is a game-changer (and it is often hard to find). It often means the difference between life and death. If you were always guaranteed to find the jetpack by level 3, the game would become more boring because it loses the thrill of actually locating the jetpack. In the same way, I think it would be interesting to introduce a few rare skills or artifacts that are much harder to find, so that when you do find them it is more thrilling.

    Total freedom could be given to making a class or skillset, and choice can be limited elsewhere, this is just my initial design idea and we will see how well it works. But here is the case against too much player choice. Lets look at two simple scenarios:

    1) You can pick 3 of 10 randomly picked skills out of a total pool of 100 available skills on each level up / shrine / whatever.
    2) You can pick 3 out of 100 skills of a total pool of 100 available skills.

    In the first method, there are 10*9*8 possible choices, but out of 100*99*98*97*96*95*94*93*92*91 possible permutations from the larger pool (if I remember my statistics 101 correctly).

    In the second method, there are 100*99*98 total ways you could pick skills on each level up.

    In the simplest mathematical terms, limiting player choice intelligently creates far more permutations of choices that they are presented with. Pattern learning becomes more difficult or impossible, and theoretically the game becomes more fun

    Given such a system, it becomes impossible to take advantage of imbalanced skills, because there is no guarantee that you will get to pick one. Thus the need for nerfing is slightly reduced (I always find it annoying when my favorite skill gets nerfed in a game).


    Of course, as mentioned, you have to walk the line between putting the player in control and not making it feel too random.

    In the end, it comes down to player taste I think, which is why I will provide at least support for classes, if not direct implementation. I also like the idea of mixing class and skill-based systems.

    The draw of CCG mechanics is difficult to explain to anyone who has not played (or learned how to enjoy) a CCG. I used to hate CCGs, but my friend forced me to play with him enough that I got into it. The beauty of CCG mechanics is that you have an infinitely-extensible set of rules, and a virtually infinite set of permutations for each game scenario, meaning you are rarely presented with the same challenge twice. Which makes mastering a specific pattern impossible - instead, you must master the general mechanics of the game. If you want a nice introduction to CCG mechanics I recommend Hearthstone - it is well polished and free to play (you don't have to pay to enjoy it). You can play on iOS or PC/Mac. It takes a few hours to build up your first class, but once you get into the game it is worth the learning curve and time investment, even if the game has minor flaws.






  • I'd love to see a combination of skill based abilities and boons/buffs.
    Skill based:
    You cut down a lot of trees, so you get better at cutting down trees. If you do nothing but cutting down trees, you become an expert. When you stop cutting down trees, your skill level stays put for a while, then trickles down slowly. If you never cut down a tree again however, you'll still stay somewhat proficient at it.
    A high skill makes you more efficient at it.

    Boons/buffs:
    There could be altars to the gods (not unlike the birthsigns in TES games) that give you boons to certain attributes. Attributes affect skills. A high strength would make your woodcutting faster, or lets you equip a larger axe.

    Now, this is all in theory. I realize that in practice it's extremely hard to let all the different attributes have a (sensible) effect on all the different skills.


    Something else that appeals to me, inspired by Flatfinger's blog: http://flatfingers-theory.blogspot.be/2008/06/living-world-massively-single-player.html
    It's to play in a persistent world, where you can inhabit an NPC and take over their life. When this character dies, you could take over another NPC.
    Each NPC would have a fixed set of skills and their place in the world, which could be a farmer, knight, guard, priest, healer, ...
    When inhabiting a character the player could visit shrines to boost certain attributes, and of course train skills but respective to the original "job" of the inhabited NPC.


    In all of these modes however, I'm imagining the game not as turn based (at least not in the literal sense). More like Baldur's Gate, with the ability to pause the game during a fight to pick spells or drink a potion.
  • "Which comes back to why I favor determinism. I think that it is ok to be presented with a random scenario (imagine a chess board that is randomly arranged and you have to make the best move that turn). But I think it is always best if there is a way to tackle the current challenge deterministically, and better still if you are never presented with a scenario that is impossible to solve (i.e. you will lose regardless of what choice you make)"
    I agree, random scenarios with deterministic solutions works best, and that is what voxelquest seems to be heading towards. It is still important to balance the random scenarios. The example of finding the jetpack in spelunky was brought up. I found elements like that to detract from the game; my success on a run is going to vary wildly based on the specific scenario I am given, independently of how well I handle it. A good scenario offers good opportunities to act on, while a bad scenario lacks those same opportunities.
    On a related note, risk of rain's biggest problem was random scenarios. There were many goals in the game which could only be completed if the scenario was generated right. You have to be lucky enough to generate this map, and that map must generate with this feature, before you can even attempt to accomplish the task. You have failed before you even began because of pure chance.

    I think random selection of attributes is a bad way to balance a game. The main effect that has is to obstruct a player's abilities to make the choices they want to make. One effect of this is that they cannot steer down an optimal path each time. This has the additional implication that you can be forced down a suboptimal path, perhaps to the point of getting screwed over because the options you need were never present(this also happened a lot in risk of rain; one run may be a breeze, and the next a slaughterfest, just because of which items were available). Preventing players from going down optimal paths is only nessecary if the game is imbalanced in the first place. Risk of rain becomes much more interesting when you get the artifact that lets you choose your items. It also kills the longetivity of the game because the balance is so awful, and so taking optimal paths leads to you being extremely overpowered.
    Perhaps ironically, the better balanced your system the better randomness works. Unbalanced systems hiding behind randomness have arbitrary difficulty curves, as one time you may get a powerful selection and another time a weak one. With a more balanced system, the randomness can force you into different areas of the possible build space, but without large swings in power from it, but if the system is balanced you should be able to let the player explore the build space at their leisure.

  • To go back to the original question of classes -- which I hate to do, as so many comments here have pushed so many of my game design buttons :D -- the idea of identifying a "class" from the player's individual selection of skills is an idea I've long wanted to see implemented in a game.

    A few years ago I wrote up a needlessly detailed design doc for a MMORPG. Basically this came after I'd played a few, studied some more, and found myself critical of most. I tend to believe that the price one should pay for the privilege of criticizing someone else's work is to try doing it yourself. So I decided to stuff my quirky personal beliefs about what would make a maximally fun MMORPG into a design, and see if the thing wound up having its own identity that might hold together as a game.

    One of the pieces of that design was the skills system. This was to be a space game, and it was going to follow my notions that gameplay for a game of any size needs to be optimized for fun at the tactical, operational, and strategic levels. The other key design goal was that all four of (what I believe are) the fundamental styles of play should be equally represented by mechanics and content.

    So the skills I created tended to be things that spanned the range from immediate bonk-bonk-on-the-head fun to things you could do for a full play session to abilities that you'd want and need to apply over several play sessions. And to decide which skills your character got, I specified that the first part of the game (which could be skipped) would be a simulator not unlike how System Shock 2 starts. You're presented with various scenarios, and each choice you make for how to resolve each scenario grants you the skill that's associated with that choice. Talking to someone would get you a diplomatic skill; shooting someone would get you a tactical combat skill; and so on.

    Once you were done, the game would tote up all the skills you picked, then calculate which of several "professions" were defined to have most of those skills. You could then choose to enter that profession, which would have appropriate content generated for you to use those skills. Or you'd be free to keep those skills and not choose any profession at all, but simply be a galactic adventurer. And as I noted above, yet another option would be to skip the training simulation and just pick a profession from the full list -- you'd then get the skills considered optimal for that profession.

    I liked this idea enough that I actually created a little Java program for it. You can pick a profession and it then displays the skills associated with it, or you can pick individual skills and the program shows you the scores of the profession for which the selected skills are optimal -- that would be the profession suggested for you in the game after picking skills through the simulation section.

    I don't think I've yet seen any game that does this, but it seems like an obvious enough idea that I won't be surprised if someone does use it, or something like it (maybe without the framing device of the training simulation).

    In the meantime, I think I'm pretty happy to hear Gavan's preference for a skills-based RPG style rather than a hard class-based system. But that's me; I also suspect most typical gamers -- especially ones who've grown up with MMORPGs over the last ten years -- will probably consider hard classes the One True Way. If today's typical gamer is impatient with ambiguity, anything other than a class-based, "holy trinity"-style character definition system may be criticized for being "too complicated."

    Some way to generate a class (even if only a text string) from individually chosen skills may help some, but I have a feeling that there'll be a number of gamers who will expect this generated class to be meaningful mechanically as well.

    Or maybe this game will mostly attract people for whom a predefined class is regarded as a lost opportunity for making interesting choices about the character they'll be playing. :)

  • Once you were done, the game would tote up all the skills you picked, then calculate which of several "professions" were defined to have most of those skills. You could then choose to enter that profession, which would have appropriate content generated for you to use those skills. Or you'd be free to keep those skills and not choose any profession at all, but simply be a galactic adventurer. And as I noted above, yet another option would be to skip the training simulation and just pick a profession from the full list -- you'd then get the skills considered optimal for that profession.

    I don't think I've yet seen any game that does this, but it seems like an obvious enough idea that I won't be surprised if someone does use it, or something like it (maybe without the framing device of the training simulation).

    Oblivion and skyrim use a system not unlike this. The starter dungeon serves as a tutorial and a means for the game to look at what kind of playstyle you prefer. Afterwards it gives you a few (general) choices but lets you override them.
  • They did do something like that, you're right. The GOAT test in Fallout 3 was another framing system for choosing skills.

    The system I was imagining made that more overt: "Due to your quick action in this scenario, you've earned the Field Medic badge. Well done, soldier." But the way that BethSoft handles skill allocation in character creation is darned close -- thanks for the reminder.
  • A few years ago I wrote up a needlessly detailed design doc for a MMORPG. Basically this came after I'd played a few, studied some more, and found myself critical of most. I tend to believe that the price one should pay for the privilege of criticizing someone else's work is to try doing it yourself. So I decided to stuff my quirky personal beliefs about what would make a maximally fun MMORPG into a design, and see if the thing wound up having its own identity that might hold together as a game.

    One of the pieces of that design was the skills system. This was to be a space game, and it was going to follow my notions that gameplay for a game of any size needs to be optimized for fun at the tactical, operational, and strategic levels. The other key design goal was that all four of (what I believe are) the fundamental styles of play should be equally represented by mechanics and content.

    If you wouldn't mind sharing, I'd love to see this! Your comments have piqued my interest in gameplay as a whole.
  • I hope a hybrid caster/martial will be viable. There are many ways that could manifest, and imo the more of them that are feasible the better.
    The first type is the spellblade. They mix their martial strikes with magic to get magical effects on top of their melee.
    The second type is the self-buffer, who can invest their magic in improving their martial abilities.
    The third is the flexible. This will likely work naturally, though its viability may not be there. they have some blend of magic and martial skill so they can use either. they would fall into three categories based on their blend:
    -off-mage : a warrior who has a bit of magic for versatility.
    -blended : can use either equally well; depends on alternating between the two to allow stamina and mana to recover
    -off-warrior: a mage who has some martial skill to help preserve mana or to use if they run dry.


    The primary disadvantage to all of these is the necessity of learning both magic and martial, which means they are worse at both than a specialist, but the cross-synergy should let them make up for it. Mixing these types would also be possible, but mana and action limits will limit how well that works.
  • *applauds* All for this, Mystify. :D
  • Mystify said:

    The primary disadvantage to all of these is the necessity of learning both magic and martial, which means they are worse at both than a specialist, but the cross-synergy should let them make up for it. Mixing these types would also be possible, but mana and action limits will limit how well that works.

    In real life, there's a limitation to what you have time to learn and practice, but you're not limited to one thing or another. I don't see why someone can't specialize in one thing but develop skills in another area. Of course, if you want to prohibit a character from becoming all-powerful, you have to set some limitations. I don't think that we should set skill limits, though (numbers of martial vs magic skills, for example).

    Gavan has already put into place mechanics for mediating a character's progression. He mentioned that skills will be chosen from shrines: given three options, the player can choose one. This randomizes character progressiom. But as long as each NPC also has the chance to develop skills similarly, I see no problem with a "limitless" system.

    This creates a new dynamic: if all of the characters in an area have access to the same skill choices, many of their choices will result in "similarly-classed" characters. Of course they won't all choose the same skills, because their personalities will drive them towards certain skills. This could be a good thing, but you don't want to end up in a steady-state (in my opinion).

    One way to mix this up would be to include a Dwarf-Fortress-like mechanic of traders. A "living world" would of course have a functioning economy, and perhaps characters could choose to buy skills/education from a traveling market. If NPCs are fairly mobile, they won't be able to take advantage of traders every time bc they have tasks in other areas. So characters would develop different numbers and types of skills over time.

    If a character recognizes that he or she needs a certain crafting skill or ability, it might be interesting for each city to develop specialties of training. A school in City A could become well-known for teaching students about how to manipulate the environment, for example. Characters (both NPCs and the player) could opt to take intensive courses that will teach them different abilities. To avoid all-powerful characters who have taken courses all over, make the trainings long enough to discourage "skill-hoarding" but short enough to be a worthwhile investment. Travel distance could also become a limiting factor, especially if the land between cities is dangerous or if your character (or possibly just NPCs) have lifetimes that make being away for extended periods damaging to relationships or politics.
  • I still hate the idea of randomized character progression. That destroys my ability to customize my character as I want, and severely cuts down my interest. A game shouldn't need to interfere with your ability to choose for fear you will be able to choose too well.

    However you gain the skills, someone who gains skills in both magic and martial will be weaker at each of them than someone who spent all their skills in one. Oftentimes, in a game like this, you can simply use one of the two, and get it at its full power every round. A character that devoted part of his power into a different skill gets to use their main skill at partial power every round, with the option of using a weak alternative. The jack of all trades and master of none is often nonviable, since they are always weaker than the specialist, unless their versatility allows them to hit the weak points of every enemy or some such thing to make their average effectiveness higher.
    First, let me draw a distinction. There is cross discipline training and intradiscipline training. a magic user and a fighter are cross discipline. The distinction is what you spend your time doing. if the training is boosting things you are already doing, it is intradiscipline training. if it boosts things that you don't normally do, it is cross discipline. My above comments were in regards to cross discipline training. With intradiscipline training, variety is often beneficial. For instance, say I do 1 damage and I have 1 health. I have 2 points I can use to increase either. if I put both in damage, I have 3 damage and 1 health. If something does 1 damage a round, I will do 3 damage before dying. if I put both in health I deal 1 damage and have 3 health, and will 3 damage before dying. However if I put 1 in each, I have 2 damage and 2 health, and do 4 damage before dying.
    With intradisciplinary training, everything you train is actively used, and hence you gain the full benefit of each, and the synergy between the different parts generally means you want to train up evenly. The costs of upgrading each part compared to their effectiveness can skew where the cost effective upgrades are, but in general you want to be well rounded.
    With cross disciplinary training, you are sacrificing a skill point that would help you out in general in order to get one that has occasional use. This is much harder to make worth it. If we take a cross-section of every situation the character will find itself in, we could make a plot of how effective they are. improving intradisciplinary skills will improve the plot everywhere, albeit unevenly. If the build is ineffective in certain circumstances, that part will increase less, and form a dip. If that dip is lower than the necessary ability to succeed in that situation, the character will fail (this is a fuzzy concept, its mainly useful for theoretical considerations, not for actually calculating things).
    You effectively have such a cross section for each discipline you have (again, this is broadly speaking. if the two disciplines can work with each other, like casting a buff, you would be considering the fusion). for your overall effectiveness, you take the maximal discipline for each situation. So, if training cross-discipline can raise the circumstantial ability to be higher than your main discipline, it can improve your overall ability curve. however, this comes at the cost of not raising your main discipline. So, in order for it to be worthwhile, you have to do it enough that it will raise the overall fitness curve to the point of patching your weak point, without doing it so much that the rest of your curve falters.

    since failing your primary curve is generally a bad thing, you either end up in the situation where you /have/ to go cross discipline to shore up your weaknesses, which serves to counterbalance being a specialist, or going cross discipline is a bad idea as you just flatten your overall curve. Neither is a good design; you don't want to force people to go cross discipline, as that is making basic archetypes into improper choices, nor do you want to make cross-discipline choices invalid. The counter to this is to create synergies. One simple one is being able to pull from multiple resource pools, thereby allowing you to use each resource faster to do bigger things at any given moment, to make up for your things being innately weaker than the specialist. Another way is to make the weaknesses of various enemies such that the cross-discipline character can dart between styles to fit the enemies. They may not take out an enemy as fast as the guy who specializes in their weakness, but maybe they can still do it fast enough to be competent. Or perhaps there are combos built into the system; after I hit this guy with a cold spell, my hammer will do more damage. Other ways are to create more explicit fusions, like the spellsword I mentioned above, you have attacks that are dependent on both abilities to be effective.

    Other methods of controlling such things involve variable costs to increase skills. maybe training my magic to half of my skill doesn't mean my martial skill is also halved; if each point of advancement is more expensive, then the specialist can work because they get, say, 3/4 of each ability, which raises their total skill enough to make up for the inefficiencies with weakening their primary skill. Another is to have the other discipline's stats influence things that you are doing anyways, which gives you intradisciplinary benefits alongside the cross discipline benefits. Another way is to have them on non-competing resources. This is hard to pull off, simply due to opportunity cost, or in the other direction there being no reason not to get both.

    In regards to everyone in the world leveling up, this creates its own sets of concerns to balance. Namely, the repesctive leveling rates become important. if you can level faster (or optimize so your effective power level is increasing faster), then you will create a level gap to squish people from. If you level slower, you get the reverse scenario. Hopefully monsters, by and large, are not also leveling, or you end up on a treadmill, where you have to level fast enough to keep up, but you don't really make any progress. You don't want to take your time exploring a town and talking to people only to find that the other adventurers have been killing monsters for the last 10 hours and are 50 levels above you. You also want dynamics to bring in new npcs. Actually, we would want that anyways, so offset the death rate in the world. This can allow for new blood to go out and level up, and then the longer a npc is out doing dangerous things the more likely it is he will eventually make a mistake, which will help keep the power levels in check. This can keep the overall world of a fairly fixed difficulty, while being dynamic. You are just another character running up the power slope. If you live long enough, you may get higher than most.
Sign In or Register to comment.