Tuesday 10 January 2012

How to write a free RPG - Chapter 5: System

The mechanics of your game provide the players with tactical and strategic choices. They will spend resources, take risks, win and lose. The mechanics must mesh with the concept of the game and support the setting. The first question you must answer is:
Do I use an existing system?
There are hundreds of free RPG systems available, all of which can be extended and modified to meet your needs. By choosing an existing system (especially a popular one), you begin with a solid, playtested base. However, you then rely upon your setting and rule extension being novel enough for people to want to play.

Before you dive in and create your own system, check the list of systems I gave in Part 2 and make sure yours is truly novel. It is much better to extend an existing system that start from scratch.

How to make mechanics

Only include mechanics for things you want the players to do in the game. Reward for the style of play you want to foster. Mechanics are a set of steps that achieve a specific outcome. You do not need to use dice, the steps can be purely conversational or use bartering with resources.

The three points of the mechanics spectrum are resource, traditional and verbal. Resource mechanics are where the player trades an in-game currency for control of the game or success. Traditional mechanics involve rolling dice and comparing to a target number. Verbal mechanics reward good ideas and rhetoric with success.

Your mechanic can be a mix of these three things. Traditional mechanics are the most widely used.

Your mechanic must give the player choice. They must choose to do something and be able to understand the likelihood of an outcome. Avoid mechanics where a single roll can cause the sudden death of the character. You do not need to have a random element to a character's action but avoid making everything automatically easy.

Make the player earn a success though clever use of their brain, either by manipulating the mechanics, setting or putting effort into the game.

Working backwards

Often it is wise to work from the mechanic you are trying to achieve back to the statistics of the character. This will avoid you getting dump statistics. Write down what part of your concept (you decided in Chapter 1) you are trying to emulate then decide on a mechanic in words that satisfies that. Finally work out what combination of skills, statistics, feats and randomness will achieve it.
I want the characters to be able to hurt the monster in imaginative ways. They must be able to inflect more damage by clever use of their gadgets, skills and environment - shooting it over and over should not lead to a win. I will need a statistic for using gadgets, skills for different gadgets (to allow specialisation) and a mechanic to make it worthwhile combining ideas and gadgets together. I'm going to use a shared dice mechanic, so the players should able able to gain more dice for working together.

The Meta Game

The Meta Game is what player-to-player interaction is called. If two players are talking about the situation from their point of view then they are Meta Gaming. If the players talk in character about the situation then that is regular roleplaying. All mechanics sit on a scale between the in character point of view and the Meta Game.

Meta Gaming mechanics can add variety to your game but must be used with care as they are often outside the sphere of knowledge of the character. The upshot is that the character may be taking actions for which they have no justification. A mantra for many roleplay groups is "What would your character do?" and that is often broken by the meta-game.

Only you can be the judge of whether Meta Game mechanics fit into your system. The mechanic types listed below include how "meta" they are.
For Chgowiz, I have a Meta Game mechanic where the players share dice. The characters do not know about the pool of dice that is being used for their actions, so it sits firmly in the Meta Game. A description of it is at the end of the mechanic types section below.

What to make mechanics for

Only make mechanics for things you want the players to do in your game. Some typical ones are:

Character creation

The creation of a character sets the benchmark against which everything in the world is judged.

Unopposed actions

The character interacting with the world alone are actions. These include riding a horse, sailing and navigation. In these cases, there is no-one opposing the character, its just the character against the world. These actions will be performed a lot.

Opposed actions

Where a character is trying to do something and another character is trying to stop them. These include persuading an NPC to open the gate to the castle. A character is trying to get the NPC to do something and the NPC's sense of duty is opposing it.


Combat can come in may different forms: unarmed, with close quarters weapons, ranged, vehicle, space craft and so on. Combat also should have a method of doing harm to the opponent. This can be the same mechanic as an opposed action. Combat is usually broken up into rounds where each character takes it in turn to do an action.

Help the players make quick choices and keep the game moving by presenting the choices in a combat action clearly. The player can then spend their thinking time working out what their character would do rather than what options there are.

Wounding and healing

Invulnerable characters are less interesting to play than vulnerable ones. Have a mechanic to keep track of how much hurt the character has been through and how much more they can take before they can take no more actions. Having a decreasing point value (Hit Points) is a traditional solution but you can also choose narrative effects that affect the player's decisions. Taking damage might also incur a penalty to performing actions. Ensure you include a way for the characters to heal too.
For Chgowiz, I want the effect of being damage to be more narrative. As characters take damage, they can pick up disadvantages - making it more difficult to play. As the characters are clones, dying is not a problem, so the damage affects can be outrageous.


Measure lethality as the amount of game time it takes for a healthy character to die with average weapons/equipment in an average fight. Is lethality so high that the player will never get a chance to retreat? If that's not part of your concept then consider changing it.

Performing "Magic"

Magic (or doing technical actions in modern/Sci Fi) does not need its own system but you can add novel flavour to your system with it. Ensure that the magic system related to the setting - a society based on magic being easy should not have a system where runes need to be drawn accurately over several turns.

Controlling the narrative

Controlling the narrative means that the players get to decide the outcome of things.

Character improvement

If the roleplaying game is designed to be run over a series of sessions, then it is important to hand out a reward that can be used to improve the character.

Character Creation

Character creation is the cornerstone of any system. Do not stint on it. A prospective GM might well make a few characters to see what it is like. Ensure it is slick, majestic and well described. For the players, it is the first time they will interact with your game and it is important that the process is well explained. Character creation does not need to be quick, some players enjoy an involved creation session, especially if the character will last through a long campaign. Be true to the concept you laid out.

Characters tend to made of some or all of the following parts:
  • Attributes - a fixed number of inherent abilities of the character. Include: Strength, Intelligence and so on.
  • Skills - a list of learnt abilities, often picked from a list in the setting.
  • Feats/Traits - extraordinary abilities that the general populace do no possess, these can be both disadvantageous too.
  • Health - a way of tracking the amount of damage the character can take before they fall over.
  • Fluff - description, character name, organisations they belong to, age or anything pertinent. It's the only place in the RPG where fluff is acceptable and prompts the player to flesh out the character.
If you want the game to be learnt quickly, try and keep to familiar terms. If your game is more epic in scale, feel free to break out the Thesaurus and pick words more familiar to your genre.

Avoid adding one of the above unless there is a rule that makes use of it. If you have a 'Psyonic strength' ability and no psyonic rules then the Attribute will be useless.

Ensure you include an example character creation, explaining the choices made at each point.

Random roll vs Point assign

Most roleplaying games use either random roll, point assign or a combination of the two (sum the rolls of 10 dice and assign). Random roll mechanics lead to faster character generation but can leave the player with a character they didn't want to play. Point assign creation tends to be slower, leads to optimisation but leaves the player with the character they think they want to play.

Backstory creation

Flow charts or randomly rolled tables can be used to create the backstory of your characters. Some players might find it too restrictive, others liberating. If you include one these mechanisms, I recommend it is optional.

Collaborative creation

Some character creation mechanisms use play a way of creating a character. In these collaborative methods, players play out scenes. The outcomes of those scenes determine or modify the facets of the character.
For Chgowiz, the players will create a 'Genome' - a root from which each clone is then generated. The Genome will have attributes and skills and will be chosen with point assign. Advantages and Disadvantages will then be randomly rolled per clone. If the clone is difficult to play because of a tough combination Disavdantages and Advantages, it is ok because Clones are expendable.

Types of mechanics

There are an enormous number of variants of dice, resource and narrative mechanic. Below are just a taste of four of the simplest mechanics many systems build upon. Most roleplaying games depend on mechanics using character properties (attributes and skills) combined with a random element.

Target number

  • Used for: Unopposed actions, Opposed actions, combat, magic
  • Format: Character Properties + Modifiers + Dice roll >= Target number
A target number mechanic is the simplest form of mechanic. A Character's Properties are combined (such as the sum of appropriate Attribute and Skill) with modifiers and a die roll. The result is then compared to a target number that is set by the Gamesmaster. In most cases, the higher the target number, the more difficult the task. For opposed rolls, the target number is a roll of the opponent. This can be slower as two dice are rolled, two equations summed before the comparison can be done.

As long as the properties are kept in low digits, the calculations are easy. Avoid applying too may modifiers. Some systems use tables to set the target numbers, this improves simulation of the mechanic but can be slow.Speed can be maintained by having the result of some calculations written on a character sheet. These are sometimes written down as secondary statistics.


  • Easy to balance
  • Quick
  • Versatile


  • Linear probability scale
  • Mathematics can be difficult with large numbers
  • There is a temptation to add many modifiers elsewhere (such as modifiers on weapons)
  • Does not foster communication at the table

Meta Gaming?

This is not Meta Gaming because the rolling of dice represents the actual actions of the character.

Dice pool

  • Used for: Unopposed actions, Opposed actions, combat, magic
  • Format: Roll as may dice as you have in character properties, remove dice for modifiers, count the number of dice that roll over a given number. To succeed, you need a number of successes.
Dice pool mechanics rely on counting the number of dice that successfully roll over a number. This can be a length process when you are rolling 20 dice but the mathematics remains simple because you are not performing additions or subtractions. Modifiers are applied by removing dice (either before or after the roll).


  • Quick
  • Modifiers do not involve maths
  • Versatile
  • Feels good to heft cupped hands full of dice


  • Can need a lot of dice
  • Counting can take longer than comparing a single number
  • Balance is more tricky
  • Does not foster communication at the table
  • Probability of success more difficult to estimate than for target number rolls

Meta Gaming?

This is not Meta Gaming because the rolling of dice represents the actual actions of the character.

Resource Pool

  • Used for: Boosting actions, controlling the narrative
  • Format: Character has a pool of points that they can spend when required
Resource pools reduce the randomness in your game by giving the player a tactical choice whether to spend the points from their pool or save them for later. This mechanic is sometimes used to allow the player to control the narrative. It can also be used to re-roll dice, boost outcomes.


  • Gives the player an tactical choice
  • Simple to understand
  • Player feels an element of control
  • Fosters communication at the table


  • Slower than dice rolling
  • Balance difficult

Meta Gaming?

Resources management tends to be a Meta Gaming task because it is not the character who is spending a point to boost an action, or taking hold of the narrative. The player is the one that is deciding to spend the pool point. If you use a resource pool for something that the character controls (such as a magical pool of energy) then this is not a Meta Game mechanic.


  • Used for: Controlling the narrative
  • Format: Players vote on the outcome
Voting reduces the randomness of outcomes by putting those back into the hands of the players. Some voting mechanisms are used with resource pools so that players have to use their votes tactically. Voting can be secret or public. This mechanic can add a level of competition at the table, make sure that fits in with the concept of your game.


  • Gives the players the feeling of more control
  • Adds tension and atmosphere to the table
  • Fosters communication at the table


  • Slower than dice rolling
  • Slows the pace of the whole game if used liberally
  • Secret voting even slower!

Meta Gaming?

This is a Meta Game mechanic. Players voting on outcomes is detached from the characters themselves.
Chgowiz uses a modification of the Target Number mechanic. To do an action, they add Attibute, Skill and a die roll versus a target number. Even if the Attribute and Skill combined are more than the target number, they still much roll a die. Where it differs is that all the players share a pool of dice in the middle. When someone does an action, they take a die from the middle. By doing so, they are depriving other players of dice. This is a Meta Game mechanic because the character do not realise that they are about to fail because the players have run out of dice!


Crunch is the name given to the feeling that there are a lot of rules to remember to play the game. You should try and strike a balance between a simple system where the tactical decisions are quick and a crunchy system where there are lots of options, modifiers and special rules. Too few rules and you're giving the player fewer tactical options, there is less game system to manipulate. Too many options and the system becomes overwhelming. Lite rules tend to be quicker to player whereas crunchy rules do a better job of representing the game world. Only you can decide whether the system fits the concept you decided on at the start.

Crunch often creeps into a system in the form of special rules for spells, monsters or equipment. These extra rules might look innocuous on their own but when the GM tries to apply all the caveats from different parts of the rules then the game grinds to a halt.


Joshua Macy said...

I'd add a couple of Cons to the mechanics list:

Dice Pool - Cons : the probability of success is not intuitive and hard to calculate. Your players may spend a lot of time not knowing whether what they're about to attempt is hard or easy. It's easy for designers to make a mistake, particularly if there are rules for varying the success number or splitting the dice into sub-pools and arrive at very counter-intuitive results (such as more skilled characters being *more* likely to Botch, as was the case with early versions of White Wolf's Storyteller system).

Resource Pools - Cons: some players object when the Resources don't represent anything the character could know or think about. This type of "meta" knowledge of, e.g. exactly how much Luck the character has left can be a distraction when trying to figure out what the character would choose to do.

Voting - Cons: Similar objections about "meta" decisions as Resource Pools; it fosters communication between the players at the table, but downplays communication between the characters. Can lead to the most charismatic/persuasive *player* always determining the outcome, or in games where it's tied to a Resource Pool the player with the best understanding of the exact rules (the rules-lawyer).

Joshua Macy said...

Oh, and Roll v. Roll probably deserves a mention, considering how common it is. Hero Games, Runequest, T&T, TriStat, even (iirc) many of the dice pool games like VtM, StarWars rely on both offense and defense rolling. While it's roughly equivalent to Roll v. Target Number if you assume the target rolls an exactly average result before modifiers, the difference in variability can be big...particularly where you subtract one from the other for degree of success.

anarchist said...

I've always thought that 'crunch' was rules, as opposed to 'fluff' which is setting or description, whereas you seem to define 'crunch' as 'having too many rules'.

Rob Lang said...

@Joshua, thanks for the feedback. When I come to update the page, I'll find a way to discuss the meta game and its affects on the system. Also, I'll be more explicit about roll vs roll for combat. I can't imagine why I didn't before!

@anarchist, thanks for your comment. Sadly there isn't a standard lexicon for talking about games on the internet. I define my terms so as to avoid confusion. Crunch really is a measure of the amount of rules, what constitutes too many is subjective.

Fluff as far as I am concerned are words that are not required to get your point across. Setting and background isn't fluff for me - it's required. Stating the obvious phrases such as "The Gm can ignore this rule if they want" is fluff because it's obvious.

Emmett said...

I find it interesting that RPG designers often decide to make their own rules (including me). There is so much that can alter how a story is told by how the mechanics influence play of the game. Sometimes a designer thinks that their mechanics explain all of the game and the implications of them are obvious. This is often not the case. Players sometimes don't understand what the purpose of a rule is even after playing for some time.

This is a huge subject for one chapter. You did a good job covering the main concepts. Maybe there could be an appendix section tied into this that discusses some number theory? The difference in the difficulty of mathematical operators for the human mind to handle and the advantages and disadvantages of larger number scales (larger numbers give more granularity while smaller numbers are simpler to handle).

Thought said...

To play off Emmett's comments, I think one of the reasons why so many game writers also craft their own system is because there is a belief that THAT is part of making a game. To speak in character, if I might: If I create a fun setting for FATE, then I have created a setting, not a game.

Of course, a lot of games end up mirroring or mangling existing systems anywho, so this isn't the sturdiest justification, but I think it is one of those underlying assumptions that influences people. It might be useful, then, if Rob were to include a Fight Club-esq line assuring would-be game writers that "Your game is not your setting. Your game is not your system. It is not the graphics it has. It is not the typeface it uses. It is not the number of downloads it has. It's the all-playing, all-socializing game of the world."

Another possible reason why so many people create their own system (although I think this is far less common) is that, ideally, the system and setting should be carefully crafted to fit each other. As extreme version of this is Rob's example of having a psychic mechanics without any actual use in game. I would propose that the difference between FATE points and Force Points centers on a similar schism between system and setting. Now, it’s been over a decade since I played either, so I may be recalling poorly. That aside, to my memory both basically make it so that a roll is more likely to be a success. Fate points, however, have no setting justification. They're a fun mechanic, but don't really mirror a concept in the world in which characters can interact (somewhat appropriately, since FATE isn't really tied to one setting). Force points, on the other hand, have the setting correlation of the Force itself. A character can use the Force, and so the player can use force points. Force Points, then, is the more elegant of the two, as it is the one that marries system and setting together. I think some writers create their own system to do the same sort of thing.

Tamás Kisbali said...

Regarding the "meta-knowledge" problem of resource pools: one could always design a system where the content of the resource pool is unknown or partially unknown to the players! It can be simple, e.g. the GM has the number of resources avaible written down on a piece of paper and players have to "request" resources from him. A problem remains: how to generate the resource pool?

Another idea that just came to my mind:
Imagine a game that uses, for example, a shared pool of three dice per player (with four players = 12 and so on). The dice are put in a bag or box. When it comes to action, it gets divided between the players, they blindly reach into the box and pull out three dice.

The catch is - dices are different! For every three dice a "rotten dice" is put into the box. You can use variously coloured dices, reds are normal, blacks are rotten. A rotten dice doesn't add up to the result, or even gets subtracted from it!

Sounds like a novelty system, but might be interesting.

Tamás Kisbali said...

Oh, and a question:
can you name a few RPGs that use a voting system for resolution?
(The UN General Assembly doesn't count...)

anarchist said...
This comment has been removed by the author.
anarchist said...

"To speak in character, if I might: If I create a fun setting for FATE, then I have created a setting, not a game."

Or the reverse: If I create a novel dice mechanic then I've created a game, even though it has no setting, focus, or any reason whatsoever for anyone to want to play it.

Rob Lang said...

Great comments everyone! Lots to unpack here... let's give it a go...

@Emmett. Making mechanics is great fun but I feel that there are enough generic systems out there. Mechanics are definitely there to support story telling. I had a quick go at writing a post-ette on probability theory (dice and stuff) but it got complex (and not useful) very quickly. I've added it to my things to blog list and will add it in.

@Thought: You can make a setting for FATE and it would be a setting. However, it would be a very thin setting. With FATE, Fudge, 5x5 and other generic systems you have to do a fair amount of mechanic tweaking before it feels like the systems hits your setting. As soon as you start adding feats or addition rules then you're building mechanics. Using FATE/Fudge as a base and then extending it with a setting and setting specific rules is making a whole game.

I like the thoughts about meta game and justification of mechanics, I'll improve the guide given those.

@Evil Scientist: Very cool mechanics ideas - get them onto 1km1kt and let's have a bigger discussion!

@Anarchist: You have indeed created a game but you've left a lot for the GM to do. I believe that those RPGs that do well online are those that make life easy for the GM. Being novel makes the game easier for the GM to sell to the players and a mechanic has to be very novel to warrant a GM convincing their group off the back of it. Normally, a GM would explain a game through its setting first and then explain the mechanics to support it.

Joshua Macy said...

@evil scientist - the problem with meta stuff is often not that it provides certainty where none should exist, but that it *requires* thinking out of character. Having to ponder whether to risk invoking luck or depleting a shared pool of resources may actually make that worse. I could see it becoming a source of inter-player conflict as well; players who might be ok with your character making clearly sub-optimal mechanical decisions because "that's what he'd do" might find it a step too far if your character is at the same time depleting party resources. That can happen anyway over things like healing potions or wishes, but a shared pool puts it front and center. With the right group I could see it being a lot of fun, but as a designer I think it's something to think about.

Rob Lang said...

@Joshua The meta game should be acknowledged but it is not a necessarily a detriment to the game. It can be dangerous for groups but I think most can see the meta game mechanics as a bit of fun and separate that from what the character's actions are.

It's definitely worth explaining, thanks!

Thought said...

I'm not really sure why the metagame resource knowledge is a problem. Take the Marvel Universe RPG as an example: sure, Cyclops might not know exactly how many energy stones he has left, but chances are, he's familiar enough with his powers and their limits to recognize when he's becoming fatigued, and thus can roughly gauge how many optic blasts he has left. Same with HP: a character might not know the exact number, but they'll still have a good idea of how much something hurts and how badly injured they are themselves (especially if they get hurt often).

And if it is a completely metagame resource, like Fate Points, then the characters would never even be aware of it anyway, and so the player is no more justified in using a resource in a given instance as not using it. Is there really any metagame knowledge if there is no possibility of in-game knowledge?

Joshua Macy said...

@Thought - knowledge isn't meta if it maps roughly into something the character could think about; things like HP and mana points or spells remaining aren't a problem, as you point out, since characters could be thinking about how hurt they are or how many spells they can cast before exhausting their mojo. Purely meta things like Fate Points *are* a problem for a lot of players precisely because the character could never be aware of it. To use them at all requires taking a step back from thinking about what the character knows and what the character would do in that situation to a more "authorial" stance about what you, the player, would like to see happen and what you're willing to spend. At best this requires some extra thinking and firewalling, at worst you come up with opposed answers.

Michael Taylor said...

"To use them at all requires taking a step back from thinking about what the character knows and what the character would do in that situation to a more "authorial" stance about what you, the player, would like to see happen and what you're willing to spend."

I'd actually say that this is a good thing. Too many games suffer from players only thinking about their characters as individuals, rather that being party of a 'group-adventure-story' meant to entertain each other.

In that sense it's not "meta-gaming" at all. It's just 'good sportsmanship.'