ttRPG System Choice
Punchier than previous attempts, here are some high-level thoughts about the distinctiveness of ttRPG systems. I hope these will help me get my head around what I want to run next, but they might also be useful if you are thinking of designing a tabletop RPG.
The feel of the mechanic is an important consideration, but it’s also the place where you’re most likely to disagree with me, so I’ll leave it until last in each case.
Do you (and your players), want…
Classes (and other mutually exclusive character choices) excel at providing (or even enforcing) character roles, and providing some character distinctiveness. Everyone knows what they can rely on (e.g.) the barbarian for, and a barbarian is generally quite different to (e.g.) a rogue.
On the other hand, all the barbarians have certain things in common, and while different games might offer different customisations within classes, the world will be boiled down to a relatively small number of archetypes.
Smaller mutually exclusive choices (such as factions) provide these benefits and disadvantages in smaller amounts. Large lists of classes also influence these effects, as does allowing multiclass characters, although those situations are hard enough to balance that there will often be a more limited set of ‘most viable’ choices toward which players gravitate.
Classes also provide an element of structured character generation, as below.
To me, the fixed list of character archetypes makes a world feel a bit pulpy, larger than life, but prone to obstacles (or strange stretches of verisimilitude) when it comes to presenting fine-grained, complicated individuals. For example, in the WotC d20 Star Wars game (from the D&D 3 era) Darth Vader appeared as a Dark Jedi Guardian (fair), but still retained 2 levels of fringer from when he was a pod racer. It was nice that the system supported this, but it seems like a strange concession to have to make, to force his known history into the mechanics.
For that reason I tend to use class-heavy systems only for fantasy games. I’m fine with mutually exclusive choices with metaphysical roots (e.g. vampire clans) but you’ll often find me house-ruling out the mutual exclusivity justified only by training investment (such as Cyberpunk 2020 roles), since there are other ways to represent that. The popularity of games like Carbon 2185 suggests that plenty of people don’t feel that way.
On the other hand, classes excel where there’s an in-fiction reason for them to exist. My favourite example of this is the D&D3 Red Wizard of Thay, where a prestige class and its prerequisites sync perfectly with the requirements of an in-game social order. The d20SW may have Jedi Master going for it as an analogous case, depending on whether you accept the tension between a properly qualified master and a person who takes on an apprentice for plot reasons (because arguably the Jedi order does exactly that).
Structured Character Generation
Some games provide a structure for making characters with a small number of major decisions, rather than having every last point placed where the player wants. The systems provide direction, especially helpful to newer players, they make character creation less fiddly, and can help make sure that characters are mechanically viable and appropriate to the setting.
They can also lead players to more rounded characters, and those with interesting combinations of contributing elements. The system might even help players make interesting characters that subvert stereotypes, although without enough freedom it may instead lead players to subversions so frequently they become stereotypes.
As mentioned above, classes often provide a lot of generation structure. It’s relatively hard to make a fighter completely unable to fight, whereas other systems might make it quite easy to accidentally make a combatant with a crucial skill missing. There’s still a difference between a class-based system that lets you go free once you have a class and one that also invites you to choose a background, a subclass, etc.
Some players prefer to spend their own points. They may even have a concept in mind that simply can’t be made in the structured generation. Many systems with heavily structured character generation have a point buy fall-back, not least because it takes a fraction of the time to design.
Structured generation usually follows a character’s life to some extent, and is particularly satisfying when (for example) it covers their education and career to date. In some settings this is almost a given: a military game set within a unit should expect certain commonalities of training (and the reduced character distinctiveness is one of the challenges you take on when you start such a game).
Random Character Generation
I’ll keep this short, because I don’t like it, and if you do you probably already know it.
Random character features are good for challenging players to work with something they didn’t expect, but costs in terms of player choice. If players are OK with the really old-school ‘roll stats then pick a class they allow’, they probably know it.
Random character features are hard to balance, and sometimes there’s no attempt: some characters have bad stats (D&D), find their lifepath full of tragedy (Cyberpunk; classic Traveller famously allows characters to die during character generation), or have the game randomly start before they have achieved what they hoped (Twilight 2000). This can allow heavily unbalanced characters, which may be good for story in a mature party (better to design it in though, with deliberately unbalanced roles) but is bad for more competitive mechanical games.
Lists of Actions
Some rules systems have a relatively light touch: they tell you how actions are resolved in general, cover the basics (attack, damage) and leave a GM a lot of leeway in deciding on task difficulty and the specific outcomes. Others prescribe a list of likely actions and detail them completely.
This is a common feature of magic systems, generally with the actions gated behind character advancement: this is good for newer players because the possibility space expands as the character gains experience. The same approach can be used to gate even mundane abilities, in which case you must trade off the benefit of a list of available actions with the loss of player agency when you tell them their character can’t even attempt something.