Approaches to RPG Design

[attevil]

Every month, as I’ve been redesigning Cyber Run, a futuristic table top RPG, I’ve also been writing an article based on subject I was redesigning. This month I was going to write about skill systems, but after overhauling our attribute system, I thought it might be more helpful to other game designers to talk about different approaches to designing a game system.

Top-down vs. Bottom-up Design

When I get a game idea, it begins from some detail of another game system that I want to change. After I get enough ideas, I start making a game. This approach is the Bottom-up design—trying to build something with a group of piecemeal ideas. One problem with this approach is linking each idea together in a logical way can sometimes be difficult. Bottom-up design does have its place, but mostly toward the end of the development cycle when I want to use the results of play testing.

Top-down design is required for designing a game’s core mechanics. A game needs to have a foundation that the other game mechanics revolve around. Creating a core mechanic for a game is difficult; it not only determines the type of game I am creating, but also how it will be played. I might have an idea about how to make an existing game system better, but not an idea about how to make an entirely new system.

What helps me most in creating a new system is determining what I want to communicate to the players. In the case of Cyber Run it is morality, which as a result determined a list of character actions, which allowed me to build the entire core mechanic.

D&D has a morality system called alignment. I and probably a thousand other people have spent hours debating over the interpretation of an alignment. I disliked the alignment system, but it was central to the game, expressed in spells, abilities and class restrictions. White Wolf games have a humanity meter, which attempts to keep player characters from becoming psychopaths.

For Cyber Run, I didn’t want to judge a character’s actions as being good or evil. Instead I wanted the system to handle consequences to actions in a realistic way.

Realism vs. Numbers

D&D took real medieval society roles to create character classes, and assigned a character attribute for each class to create their core mechanic. The hit point, saving throw, and attack systems revolved around this core mechanic.

Some games don’t need or want to reflect reality in anyway, but in RPGs I found that I become more absorbed in the games that do reflect some aspect of reality. I sometimes get too focused on this reflection in the details of a system and lose sight of the big picture. I find that it is better to limit reflecting reality in a game to the core mechanic, and then just concentrate on how the other game systems can work with the core.

D&D has a terrible, realistic representation of shields in their Armor Class system. In realistic shield-wielding combat, the shield is the primary defense against attacks, and in D&D it can be a simple +1 to AC compared to full plate armor that gives a +8. This mechanic can be addressed with house rules and the expanded rules that more recent editions provide, but in first edition D&D, reflecting realistic shield combat was not as important as reflecting realistic medieval societal roles.

Basic Rules vs. Detailed Rules

I like the idea of a game with few and simple rules, such as Guardians of the Order games or something like Ghost Echo, but the simple games I play are only fun for a inadequate amount of time due to their limited content and complexity. Heavily detailed games can be fun, eventually, but the initial cost of time for reading and learning one is a barrier that prevents many gamers from ever playing the game.

In addition, simpler game systems rely heavily on the ability of the game master. An experienced GM is free from the shackles of strict rules to craft a story, but an inexperienced GM can become frustrated, often ignoring some of the rules of the game systems, to simplify the entire game as a result, providing the players with a less desirable game experience.

Detailed game rules allow inexperienced GMs to perform better without making an experienced GM perform worse. Finding the balance between the two approaches can be tricky; I make the most basic rules possible for a play test and slowly add more rules after a few gaming sessions. What I found most important in creating a game is to not get burnt out on the design. Only if I maintain my passion for the project can it get finished. I hope this helps anyone struggling with their own game design.

Follow the Development of Cyber Run

http://cyberrun.grumpogames.com/

- See more at: http://grumpogames.com/blog/approaches-to-rpg-design/

 

[GMMichael]

I thought it might be more helpful to other game designers to talk about different approaches to designing a game system.

Agreed. Why should only players and GMs get all the forum-use?

This approach is the Bottom-up design—trying to build something with a group of piecemeal ideas...
Top-down design is required for designing a game’s core mechanics. A game needs to have a foundation that the other game mechanics revolve around. Creating a core mechanic for a game is difficult; it not only determines the type of game I am creating, but also how it will be played...

In my world (my fantasy version of reality, not campaign world) bottom-up sounds like designing the "core mechanics" first, and stacking the peripheral rules on top of them. Top-down sounds more like taking an entire game and chiseling at the unsightly bits. But your definitions work too.

D&D has a morality system called alignment. I and probably a thousand other people have spent hours debating over the interpretation of an alignment...For Cyber Run, I didn’t want to judge a character’s actions as being good or evil. Instead I wanted the system to handle consequences to actions in a realistic way.

One reason that alignment was troublesome was that it was subjective. Everyone has a slightly different idea of Good and Bad. It did give tremendous flavor to the game. It may have been implemented at the core, but it always felt like an optional rule to me anyway.

I'd like to see Cyber Run's character actions that are tied to morality. I bet it's awesome, but it sounds like it would have the same problems alignment did.

Realism vs. Numbers
Some games don’t need or want to reflect reality in anyway, but in RPGs I found that I become more absorbed in the games that do reflect some aspect of reality. I sometimes get too focused on this reflection in the details of a system and lose sight of the big picture. I find that it is better to limit reflecting reality in a game to the core mechanic, and then just concentrate on how the other game systems can work with the core.

D&D has a terrible, realistic representation of shields in their Armor Class system. In realistic shield-wielding combat, the shield is the primary defense against attacks, and in D&D it can be a simple +1 to AC compared to full plate armor that gives a +8.

Interesting contrast. When I hear "realism" in games, I think "simulation," and the next thing I think: "numbers."

There's one good way to reflect reality in games: get out your physics books, and do a ton of math. The further you get from that, the less realistic the game is, and perhaps coincidentally, the more fun the game is. (Yes, there's probably a limit to this.)

Shields were a fun one for me to handle in designing Modos RPG. I did the same thing as attevil: using another rule as my starting point. I believe it was Final Fantasy Zero...or just the video game. My observation was that shields serve a different function than armor, so they should be simulated with a different rule or rule implementation.

Heavily detailed games can be fun, eventually, but the initial cost of time for reading and learning one is a barrier that prevents many gamers from ever playing the game.

In addition, simpler game systems rely heavily on the ability of the game master. An experienced GM is free from the shackles of strict rules to craft a story, but an inexperienced GM can become frustrated, often ignoring some of the rules of the game systems, to simplify the entire game as a result, providing the players with a less desirable game experience.

Let's throw the GM under the bus. An experienced GM (depending on personality type) can run a good light- or heavy-rules game; an inexperienced GM can't run either type of game. This issue might be served by talking about rules that restrict. A poor GM can run a great game of Monopoly, because the game is highly restrictive. Early video games were highly restrictive, but could still be awesome if they had enough artistic flavor. A GM is responsible for conveying, amongst other things, that flavor - and if he fails to do so, the game suffers. Light or heavy.

 

[attevil]

One reason that alignment was troublesome was that it was subjective. Everyone has a slightly different idea of Good and Bad. It did give tremendous flavor to the game. It may have been implemented at the core, but it always felt like an optional rule to me anyway.

I'd like to see Cyber Run's character actions that are tied to morality. I bet it's awesome, but it sounds like it would have the same problems alignment did.

I agree, even with Christian values as a norm, which the alignment system seems to be trying to reflect, there is a lot open to interpretation. To me this is an indication that a game system is bad. I think D&D should have actually written up a list of things that are good, evil, chaotic, neutral and lawful, but that is a daunting task, which is another indication its a bad game mechanic.

Cyber Run uses NPC opinion and reputation to give consequences to "bad" actions, if someone uses the steal, cheat or intimidate action, then they have a chance of the NPC opinion of their character lowering as well as their reputation lowering because the NPC makes a post on social media of what the character did. If the character used one of these actions for a "greater good" then the GM can decide not to change the NPC opinion or the character reputation level. This has a potential of players arguing with the GM, but instead of saying there is an Objective Good and Evil like the alignment system, this is based on the subjective view of a single NPC.

Shields being a different mechanic makes a lot of sense!

A poor GM can run a great game of Monopoly, because the game is highly restrictive. Early video games were highly restrictive, but could still be awesome if they had enough artistic flavor.

Read more: http://www.enworld.org/forum/showthread.php?436815-Approaches-to-RPG-Design#ixzz3WC3mAfGH

I think you described it better than me, you are right it is more about being rule restrictive that determines if an inexperienced GM can pull off a decent game.

As always, thanks for your comments!

 

[Neonchameleon]

First before continuing I'm going ot say thank you for posting this - and I'm going to appear to be disagreeing with it in my reply more than I actually do. The areas of conflict are the ones I'm focussing on because they are interesting.

Top-down vs. Bottom-up Design

When I get a game idea, it begins from some detail of another game system that I want to change. After I get enough ideas, I start making a game. This approach is the Bottom-up design—trying to build something with a group of piecemeal ideas. One problem with this approach is linking each idea together in a logical way can sometimes be difficult. Bottom-up design does have its place, but mostly toward the end of the development cycle when I want to use the results of play testing.

Top-down design is required for designing a game’s core mechanics.

I think that there is a very important distinction round here, but missed quite what it is. I'd rather differentiate between Hacking and Design whith hacking mostly equating to your bottom up design and Design being your top down design. Hacking is changing the rules of a pre-existing game, and it is entirely possible to reach the "Same old axe, three new handles, two new blades" situation through hacking.

A game needs to have a foundation that the other game mechanics revolve around. Creating a core mechanic for a game is difficult; it not only determines the type of game I am creating, but also how it will be played. I might have an idea about how to make an existing game system better, but not an idea about how to make an entirely new system.

And here I am going to praise two specific game engines that have core mechanics that are immensely flexible for design work and seriously blur the boundaries between design and hacking. They are the Powered By The Apocalypse engine (Apocalypse World, Monsterhearts, Dungeon World (not a great implementation), Black Stars Rise, Monster of the Week, and over 100 more including my Houses and Wands and Panem et Circenses) and Cortex Plus (Smallville, Leverage, Marvel Heroic Roleplaying, Firefly).

Cortex Plus is the easier to explain - and is a very swingy system. You roll your dice pool and take the two highest numbers as your outcome - and any 1 is a botch. Your opponent does likewise. The core mechanic is apparently simmilar between all the games. But each dice represents a different layer of what the game considers important and these vary game to game. Firefly's very trad - the two core layers are Stat and Skill. In Smallville on the other hand the two core layers represent your values and your relationship with the most important person directly or indirectly involved.

Powered By The Apocalypse games are trickier to explain because the core mechanic is simple - 2d6+stat (very few modifiers). 10+ is a complete success, 7-9 is a partial success, 6- things go very wrong. But after choosing that you need to pay attention to exactly how each move works - especially on the partial success list, a key factor is who chooses which part of the success you get (so on a partial success at Going Aggro, whoever you targetted needs to do something but they decide whether they retreat, fort up, suck it up, or what). There's a lot of design work to be done here after choosing your core mechanic.

D&D has a morality system called alignment.

D&D has two morality systems called alignment that were haphazardly jammed together. oD&D was a game about frontier exploration, conquest, and settlement and had Law (urbanisation, settlement) vs Chaos (untamed wilderness, nature). There is also an entirely different High Fantasy Good vs Evil morality system. And they were just kinda jammed together in 1E AD&D without checking how well they fitted. Confusion all round.

Realism vs. Numbers

I think what you're looking for is simulation vs game.

D&D took real medieval society roles to create character classes, and assigned a character attribute for each class to create their core mechanic.

Nope, nope, nope, nope. Utter misunderstanding of what happened. In the beginning D&D started out as a hacked tabletop wargame with two classes. The fighter, and the wizard (who was the fantasy equivalent of field artillery - and could only cast each of their spells once). The Cleric came in in the early 1970s specifically to combat a vampire PC who had got a little overpowered - the Vampire was called Sir Fang and based on Hammer Horror Studios - so they based the vampire fighter on Hammer Horror Studios Jonathan Helsing and gave him magic that would help deal with vampires. And then gave him clerics' robes that are somewhat less authentic than the average costume at a Rennaissance Faire.

The hit point, saving throw, and attack systems revolved around this core mechanic.

Again, nope. oD&D was produced to work with Chainmail rules. The core D&D rules came later.

D&D has a terrible, realistic representation of shields in their Armor Class system. In realistic shield-wielding combat, the shield is the primary defense against attacks, and in D&D it can be a simple +1 to AC compared to full plate armor that gives a +8. This mechanic can be addressed with house rules and the expanded rules that more recent editions provide, but in first edition D&D, reflecting realistic shield combat was not as important as reflecting realistic medieval societal roles.

Neither were remotely important in any way, shape, or form. This was one of the big reasons Gygax railed against realism - and where games like Chivalry & Sorcery were made (and generally sank without trace ). What was important was that the game itself was fun and provided useful and engaging things for the PCs to do. D&D as written wasd a game about managing platoon sized units of followers, exploring hostile environments, and getting rich to the point you might rule a kingdom. An utterly artificial setting - and the very artificiality of the dungeon was part of its appeal.

Also a desire for realistic combat is a common feature of Fantasy Heartbreakers. D&D wasn't about that either.

Basic Rules vs. Detailed Rules

I like the idea of a game with few and simple rules, such as Guardians of the Order games or something like Ghost Echo, but the simple games I play are only fun for a inadequate amount of time due to their limited content and complexity.

This, I'm sorry to say, is begging the question. What is "an adequate amount of time"? I've spent more time playing a single RPG campaign than it would take to watch Breaking Bad from start to finish. Actually there is a completely different issue going on here.

Between 1974 (oD&D) and 2003 (My Life With Master), with the notable exception of Pendragon, there were almost no RPGs that gave a sense of closure to the story. oD&D had the Adventurer/Conqueror/King arc in which the very nature of levelling up and your relationship to the world changed at around level 10 and you retire into a king or statesman. Pendragon was generational. And My Life With Master always has the same story (although a lot of variations). You start off with all the PCs as minions of an evil master who treats them badly. Eventually one of them snaps and tries to kill the Master. The Master fights back and tries to kill the minion. Win or die, the story came to an inexorable conclusion and is satisfying. In only a couple of sessions. Monsterhearts lasts five or six sessions per season, and that's very satisfying because it has growth and transformation built into the rules. But a game designed round procedural mechanics frequently doesn't have this. You grow upwards as you level, but don't fundamentally change. There's no narratively satisfying conclusion.

Heavily detailed games can be fun, eventually, but the initial cost of time for reading and learning one is a barrier that prevents many gamers from ever playing the game.

In addition, simpler game systems rely heavily on the ability of the game master. An experienced GM is free from the shackles of strict rules to craft a story, but an inexperienced GM can become frustrated, often ignoring some of the rules of the game systems, to simplify the entire game as a result, providing the players with a less desirable game experience.

I'm going to once again poing you at Apocalypse World or Monsterhearts. Leading GMs to become good GMs is a part of a good ruleset.

 

[attevil]

Thanks for reading!

I'd rather differentiate between Hacking and Design

I think some designers begin with hacking, but then create a new system after they have changed the rules to an existing one and found that wasn't enough. That is why I called it bottom up, but whatever its called, its an approach to game design.

Apocalypse World & Dungeon World

I've read the rules for both, they are innovative and I like some rules which I've integrated into Cyber Run.

In the beginning D&D started out as a hacked tabletop wargame with two classes.

I think the history of D&Ds creation is helpful to design, but I was focusing on the first edition game mechanics, explaining their relationships between subsystems, not explaining how those game mechanics came to be. The first edition classes each had an attribute as its primary, attributes also influenced combat and hit points, so a game designer could look at the attributes as the core game mechanic and classes as well as combat ability as subsystems branching off of the core. I would recommend designers to create a core rule mechanic and then play test. Through play testing they can add things like the Cleric class you mentioned above.

Gygax railed against realism.

I think other designers struggle with reflecting reality compared to making the game itself fun.

What is "an adequate amount of time"?

There isn't, it is up to the intentions of the designer. "KOBOLDS ATE MY BABY!" is great for a night of entertainment, but doesn't have the depth that D&D has to allow players to invest years of their life. Just something for designers to keep in mind for the type of game they want to make.

 

[nomotog]

I think that there is a very important distinction round here, but missed quite what it is. I'd rather differentiate between Hacking and Design whith hacking mostly equating to your bottom up design and Design being your top down design. Hacking is changing the rules of a pre-existing game, and it is entirely possible to reach the "Same old axe, three new handles, two new blades" situation through hacking.

I don't really think there is that much of a difference between hacking and designing. Hack something enough and you do end up with a different thing. I am kind of partial to hacking because it's kind what I do. I also get vaguely annoyed by the massive verity of core(roll) mechanics we have. We have a billion ways to roll to hit and often there doesn't seem to be that much of a point for the verity.

My thinking is, the interesting elements of a design come right after that core design. Like if we were using d20 as our base, the interesting stuff comes from the elements you tack on top of that d20+something roll mechanic. You know elements like skills feats ect. The close to core mechanics. (Actually, how do you tell what a core mechanic is?)

 

[attevil]

We have a billion ways to roll to hit and often there doesn't seem to be that much of a point for the verity.

I agree, a few articles back

http://grumpogames.com/blog/table-top-games-the-d20/

I went through several game systems to determine the best rolling mechanic. I finally decided that the D20 was the best for me, because the D20 is basically a percentile broken into 5% parts. I felt 5% is the lowest percentage that we would want to modify a roll by, a 1% change isn't meaningful enough, and its easy to modify a D20 with higher percentages with a +2 for 10%. The only advantage for other systems was tackling the probability problem, they use additional dice for that and I personally wanted to keep the roll to one die.

This is why the D20 system is nice, a set of rules already based around the D20 die. I find "hacking" that system is a good way to try new design ideas.

how do you tell what a core mechanic is?

I don't think all games need a core mechanic, but I wanted to share my experience that creating a core mechanic made everything else fall into place, well after a design overhaul. For me a core mechanic is a single aspect of the game that determine or influence other parts of the game.

For Cyber Run, I made the core mechanic the character strategies. There is a list of about 9 strategies, for example Charm and Steal. The character traits and skills either gives a bonus or unlocks a specialization of a strategy. The traits and skills are different sub-systems that are still connected to the core mechanic.

 

[nomotog]

I agree, a few articles back

http://grumpogames.com/blog/table-top-games-the-d20/

I went through several game systems to determine the best rolling mechanic. I finally decided that the D20 was the best for me, because the D20 is basically a percentile broken into 5% parts. I felt 5% is the lowest percentage that we would want to modify a roll by, a 1% change isn't meaningful enough, and its easy to modify a D20 with higher percentages with a +2 for 10%. The only advantage for other systems was tackling the probability problem, they use additional dice for that and I personally wanted to keep the roll to one die.

This is why the D20 system is nice, a set of rules already based around the D20 die. I find "hacking" that system is a good way to try new design ideas.

Ditto all of that for the reason I like the d20 system. It's just really basic and really easy to understand and then hack up.

OK here is a thought. Mechanical feel. This is something I have just started looking at is how a mechanic feels and if that matches up with how it should feel. You know dose combat feel like combat, or dose picking a lock feel like picking a lock. (Should they feel alike?) This is what I use rather then reasim. I tend not to try and replicate the reality of an action, but more the feeling and choice that the player expects from it.

 

[attevil]

(Should they feel alike?) This is what I use rather then reasim. I tend not to try and replicate the reality of an action, but more the feeling and choice that the player expects from it.

You Bring up a good point, its easier for new players to learn a game, if each game mechanic is similar, rolling a D20 for a skill check is same as rolling a D20 for an attack.

But is rolling a die in both these cases actually fun or why is it fun?
That question can be it's own article, so I wont go in to it.

To back up your idea, we could think why is picking a lock fun or how can it be more fun. The answer is creating the feeling that picking a lock is different from just rolling a D20. I'm curious at what you have come up with for that.

For combat, I certainly feel we can do more with the system, part of the fun of combat is the strategy aspect rather than the actual die rolling. Often strategy is limited to only a few variables, for example cover, flanking or special attacks. Creating combos, team work attacks or adding moral to combat can open up new strategies to expand the choices players have in combat.

I do think fun is the most important aspect of a system design, but I think fun and a system having some reflection of reality have a positive correlation with each other.

 

[nomotog]

You Bring up a good point, its easier for new players to learn a game, if each game mechanic is similar, rolling a D20 for a skill check is same as rolling a D20 for an attack.

But is rolling a die in both these cases actually fun or why is it fun?
That question can be it's own article, so I wont go in to it.

To back up your idea, we could think why is picking a lock fun or how can it be more fun. The answer is creating the feeling that picking a lock is different from just rolling a D20. I'm curious at what you have come up with for that.

For combat, I certainly feel we can do more with the system, part of the fun of combat is the strategy aspect rather than the actual die rolling. Often strategy is limited to only a few variables, for example cover, flanking or special attacks. Creating combos, team work attacks or adding moral to combat can open up new strategies to expand the choices players have in combat.

I do think fun is the most important aspect of a system design, but I think fun and a system having some reflection of reality have a positive correlation with each other.

I haven't thought of too much with lockpicking. It kind of runs into the issue of brain and time space. You know how much effort should you put into a single element. I had a kind of complex system involving different tool metals and how they interact with different kinds of locks/traps. It was kind of complex though. Another thought I have is to have a tension building thing where the player lists the things they are doing before they roll. The idea to kind of ground what is happening and build a little tension. I haven't tested this stuff though.

Lockpicking ends up being kind of a low priority, so most of the time people are ok with the simple dynamic in that some players can lockpick and some can't. Your a rouge, you do rouge things and picking locks makes you feel unique.

I don't like to use fun as a goal, because fun is just really hard to define and everyone will define fun differently. Like I find in depth rule systems fun, other people find in depth fluff fun.