Archive for the ‘design’ Category


Pandemic Legacy Design

March 25, 2017


Aside from the Pandemic games being pretty awesome, it’s really fascinating to watch folks develop campaign play from a very different mindset.  It also highlights some really useful things we can backport to tabletop rpg design as well.

Campaign Commitment Time

This was one of the major concerns in the design discussions and a focus for a lot of the indie rpgs that came out of the Forge – that the longer the campaign commitment, by nature, the smaller your target audience is who can/will commit that time in.  I think it’s really interesting that they talk about 12-15 play sessions being their top expected amount of content in a Legacy game, and the value of having a specific number laid out to players from the beginning.  (This was one of the big points I’ve written about for years, as simply making games more functional).

Linear Story for a Linear Metric

One of the biggest differences in a boardgame and tabletop RPG is the fact that while you can drench the boardgame with fictional elements to encourage roleplaying and enjoyment of the theme, the fiction doesn’t feed back into changing the outcomes of play as it does with a roleplaying game.  No one is under the illusion that they can “do anything the character can do” as a viable option in a boardgame, so it also means the directions a story can go is highly limited, and makes things like the linear story set up they’ve created work.

It’s funny that they mention Robin’s Laws, since Robin Laws’ version of HeroQuest 2 attempts to create the up and down beats, except without the additional considerations of players gaming the system (“incentivizing failure”) or engaging the mechanical aspects beyond turning a difficulty dial up and down.  I suspect Jason Morningstar’s Juggernaut might be a better parallel, but having not played it yet, and not trying to spoil myself before doing so, I could be wrong.

The Power of Unlockables and Physical Artifacts

Although sometimes people have given their players a document or a puzzle box to open, it’s not the same as fully integrating unlockable, physical cues/objects as a key part of play.  In the video, they mention that people KNOWING there would be a box to open incentivized action to either open it or avoid opening it, depending on whether the assumption was that it was good or bad.

In tabletop RPGs, either mechanical rewards are well known ahead of time (“If I get X number of points, I can buy Y power”) or exist solely as a verbal description (“The sword glows with magic power…”) but not as a physical object to taunt, tease, threaten you.  Adding uncertainty into advancement hits Pavlov’s intermittent reward cycle, but also because these rewards are thematic/plot based, and not just “get +X to ability Y”, they tend to be more interesting and emotionally fulfilling to the group.


I think this is some of the best information on tabletop playtesting out there.  They’re looking for: a) when the group has confusion, questions or forgets rules, b) emotional responses to play, c) why a group is making certain choices in the game.  (Notably, this was effectively what the Forge folks pushed for from actual play accounts in regards to playtesting).

You’ll also note that they’re talking about hundreds of hours of playtesting from several groups, including strangers.  While the indie RPGs pioneered the idea, it was Paizo’s Pathfinder playtest that really made it popular for tabletop games – however, having video or recordings where you can see the actual play rather than depend on memory, filters, or bias in self reporting is vastly superior.


Loops vs. Grind

February 24, 2015

I try not to do too many “pure theory” posts, but recent gaming and some conversations with Quinn Murphy, has given me some thoughts I think are key mostly to design, though given the way a lot of RPGs work, that also influences how you run a campaign in the long term.

Moment to Moment Fun

Bungie, the developers of the Halo videogames, are often quoted as pointing out their design method is to try to make a few seconds of fun, then keep doing that, over and over, until you have a full game.   This is emphasizing the moment-to-moment kind of fun, and videogames are quite good at it.  Tabletop roleplaying games have this too, even if the moments are more “minute to minute” rather than seconds.

The trick in this, is identifying what is supposed to be fun in this experience.  Videogames will have stuff like, “Running and jumping on things is fun!” “Making things explode is fun!” while tabletop RPGs have a different focus such as, “Being Spiderman is fun!”, “Smart tactics in battle is fun!”, “Showing courage in the face of adversity is fun!” and so on.

The problem is, of course, while videogames can focus on the literal moment to moment – where what you have your character do constantly is the “kick off” of the fun, in tabletop, it’s a structured conversation – you say things, the other players say things, and this mediates back and forth to create the actual experience.

And that experience isn’t coded into a computer that will reliably do the same things over and over, so you have to design procedures that allow groups to consistently get to the thing that is fun for this game.

What is the moment to moment fun supposed to be for your game?

What do you have in the rules to support that?  What choices are players required to make?  What input are they supposed to give?  What does the system/rules do to create that?

Play Loops

Beyond the momentary fun, you talk about larger loops of decisions, rewards, and experience.  This is actually the place RPGs have had the biggest effect in game design – the idea of leveling up, gaining resources, and long term play in this manner is pretty much a big thing which has ported over to many other games.

Play loops are on larger scales than moment to moment – we can talk scene to scene, over several scenes, over several sessions.  You can have several loops working inside each other or parallel.  Figuring out how fast a loop should complete and what decisions or actions it encourages is critical.

When it’s done right – you’re rewarded for doing the thing that is fun, so it’s doubly reinforced.  When it’s done wrong, too slow, or against the grain of the point of the game, then it’s a grind (see below).

Focusing play or nothing at all?

A key consideration RPGs have to take into account is managing the fiction – how do you keep the group contributing and pushing the imaginary events and conflicts into the ways that make for the goal of that you want with your game?

This is the baseline issue of whether your RPG is well designed or not.  Unfortunately, many games simply lack this and it becomes “style” or “experience” to find ways to make up the difference, which is why a lot of games have problems and fall down a lot – if you can’t consistently meet and create at the same fun spot – the fun is sporadic or absent.  If you are working to get to different kinds of fun, you can be fighting against each other and again – sporadic or absent.

A clear loop of procedures allows people to more consistently hit the expected fun zone, which is why a lot of narrowly designed games, or even boardgames or videogames have eaten up a lot of the RPG crowd as time goes on.  The more reliable fun, wins.


The difference between a loop and a grind is fun vs. boredom.  When a loop is kicking off fast enough, and in the right way, you have a great reward loop. When a loop is taking too long, or is about doing something that isn’t fun or the point of the game, you have a grind.

The easy example might be to point to videogames.  In Final Fantasy games,  you mostly fight and explore.  That’s pretty much the core fun of those games.  But they started including mini-games, which, instead of being side extras, became things you HAD to do as part of play.   Now you have to learn a different game, that is different than the core experience… and often put in lots of time.  If it happens to be a game you like – great.  But for many, it turned into a grind.

This is true also of loops that take too long.  Fighting the monsters might be fun, but fighting the SAME monsters, over and over, for hours on end, without any changes to the situation, isn’t.   Videogames fall into this pitfall because they often use it to artificially extend playtime without having to add real content.   RPGs usually fall into this category by designing for campaign length play that people aren’t able to fit into their lives.  The loop has to close sooner, and change the variables (difficulty, give you new options via powers, etc.) or it becomes boring.

This is actually a massive pitfall for RPGs.  You have a slow moment-to-moment play, commitment times are usually 3-4 hours of play, weeks on end, months on end, and you can’t really play alone.   Having a grind is killer to sustaining play.  If you don’t have a reliable loop of what play is supposed to be, effectively all play becomes a grind, because the struggle is just in coordinating to play in the first place.

A useful question to ask about loops for any game you design, or play is: “Will I ever see this happen?  Under what conditions?  Is this reasonable to ever expect?”

If you can’t keep a group together for more than a month or two, do you need to buy tons and tons of supplement books for high level play?  Do you need a mega dungeon?  Do you need to prep for months of campaign?

What is the expected rate of this loop turning over?  What actions does the group have to take to make this happen, and how likely are they to do so?  Is this fun enough to make it worth it?


Morality Mechanics

September 25, 2014

Morality mechanics are rules and systems designed to have play deal with morality in some way.   There’s quite a few ways to do it and with some interesting results.

Direct Morality

Direct Morality in a game is where the system explicitly lays out what is good vs. bad and enforces it along those lines.  These are actually pretty hard to do – since all of humanity’s history we’ve been still working out ethics and morality, so having a set system usually means working in a very narrow clear-cut range of ideas or something where genre tropes are simple and well established with regards to morality.

The old Marvel Superheroes RPG is an example of one that works pretty well.  It’s somewhat of a narrow path to walk to get it right.  If the game is too defined in producing a code, you can end up with situations like Legend of the Five Rings’ Honor system where you run into places where the honorable action costs you points or the dishonorable action gains you points.   If the game is too vague about the scale or actions of morality, you get the endless discussions of D&D’s alignment system.

Divergent Morality

Divergent Morality is where the game is designed such that the mechanically coded systems of value and morality are clearly not the real issue of morality – the point is to explore where they meet, and more often, where they diverge.   These work much better because nearly always we’re talking about a system where the mechanically coded “morals” are actually just pressures to constantly set up a choice between following it, breaking it, or finding ways to work around/with it towards real morality.

Polaris and Thou Art But a Warrior both use their particular codes of honor to trigger the character’s advancement along their tragic future, which, often enough, real moral choices are often strongly at odds with the code itself.  Dog Eat Dog is pretty much a game built on the scathing criticism of colonialism and it’s form of cultural genocide set up as morality code.  Poison’d has Sins, which include some things that are clearly wrong, some things which are situational, and other things like Paganism or homosexuality which simply aren’t.

Emergent Morality

Emergent Morality are games where the moral code isn’t directly given mechanics, but rather the mechanics as a whole are set up to pressure you to finding yourself having to make moral choices all the time.  These are often subtle from a read through, but brutally powerful when you play them.

Steal Away Jordan uses it’s Worth mechanic, which, naturally puts white slave owners at the top and everyone else progressively further down the scale, and when you’re closer to the bottom you are forced to make alliances, to seek help, to even align with the oppressor just to survive.  The Drifter’s Escape sets up the Drifter in a situation of horrible choices and morality emerges from that.  Trollbabe explicitly sets up where your character’s effectiveness is based in risking allies to injury or death.  Lacuna and 3:16 Carnage Among the Stars has advancement mechanics that explicitly set player goals against each other even as they’re supposed to be working together, which creates an extra level of friction.


Procedural Consequences

Procedural consequences list out a series of actions or activities which fill or fail the moral code.  “If X then Y” which makes these often inflexible and somewhat tricky to use well.  This is most often what people think of when they think of morality mechanics – Vampire’s Humanity lists, Legend of the Five Rings’ Honor codes, etc.

Directive Based Consequences

Directive consequences involve points added/lost when an action is taken that fulfills the spirit of the value/morality code as determined primarily on the basis of a player making that judgment.  Instead of listing out a code, it primarily sits on a player (the GM, a chosen player, the group) to decide when you’ve fulfilled or violated a moral position or code.  Notice that for this to work well, the thing being judged and the basis of doing so needs some kind of guidance or clarity – or you end up with the same problem that sends people spiralling around D&D alignment arguments.

Sorcerer’s Humanity mechanic fulfills this, though Bliss Stage uses it to an even stronger degree with it’s intermission scenes.

Randomized Consequences

Randomized consequences are pretty interesting.  These can be things like rolling a die to see if you gain or lose points despite having done actions towards that effect, or it can be a roll at the end of your character’s arc to see what the effects are.

1001 Nights uses it’s Safety/Ambition/Escape tracks as gambles you take between scenes, which sometimes results in the most bastard player characters living happily despite their behavior.  Poison’d uses the Salvation roll after your character dies to see if they make it to Heaven, or fall to the depths of Hell for their deeds.

If you find my blog entertaining and valuable, consider supporting me on Patreon.


Stakes and Outcomes

September 16, 2014

Let’s start with the easiest rule: “I say it and it happens.”

That’s the easiest rule to roleplaying and it’s used all the time.  Rules, mechanics, etc. are basically add-ons to try to create something more interesting than that.   Part of what makes roleplaying as an activity conceptually difficult is that you are doing all the things other fiction does, in terms of constructing a narrative, except you’re coordinating it between several people.  Play requires consistently answering these questions on the spot:

– What kind of events fit this game we’re playing?

– What methods/actions can characters take in this game that fits with the mood/expectations/”realism”(feeling) of it?

– What kind of outcomes are reasonable?

(add in an extra hurdle for many games, which is, how do I translate this abstraction of the rules into these things?).

Anyway, beyond the simplest rule, mechanics and systems of resolution basically fold into three types:

Fiat Outcomes

A dice roll is made, a point is spent, something is done which provides some limitation on the outcome but the rest is narrated and decided by one person (in traditional games, that’s the GM, nearly always.  In narration trading games, it’s explicitly passed around.)

The positive to this side is that whoever has fiat for this has a guaranteed input into play, and their vision can get into the events in play.  The negative to this is that if the group isn’t tightly coordinated on what fits in their game (from the questions listed above), then you have all kinds of miscommunication and confusion.   “Wait, I thought the fall was like 5 feet, not 500 feet?!?” “What do you mean a failure destroys the whole kingdom?  I thought it’d only mess up the castle?” etc.

Negotiated Outcomes

Negotiated Outcomes sets up a negotiation process as part of figuring out the outcomes.  The most common example is stakes setting before dice are rolled or cards played, etc.  This lets everyone know the outcomes and players can make better choices about how much resources to spend or how important the outcome will be.  There are games like Polaris, Thou Art But a Warrior where negotiation is actually part of the process of resolution, and there are games where you negotiate after the fact (“You win the argument but you give up a concession.” etc.)

Negotiated outcomes are quite flexible, but they are also a little kludgy in many places and there has to be effort on part of the group not to spiral too deep into pre-playing the events before they actually play out, or pushing outcomes beyond the scope of the situation.

Hardcoded Outcomes

Hardcoded Outcomes appear in many places in many games, though usually they’re in combat or magic and sporadically in other types of conflicts.   A Hardcoded Outcome means that a dice roll adds a specific event into the fiction – in combat, that would be an injury or death, usually.

When done well, these make it easy to play and shape the fiction in great ways – it helps answer the 3 questions as part of the system itself and makes it so that the group doesn’t have to spend time figuring out what “feels right”.  When done poorly, it either breaks the expectations of the fictional world you’ve set up (“Wait, I can jump off a cliff and survive without much trouble?”) or it makes it very hard to translate into the fictional world.

The more modern design take on this would be things like Monsterhearts’ move sets – you roll and pick from a list of pre-set options of outcomes.

If you find my blog entertaining and valuable, consider supporting me on Patreon.


Designing Easy-To-Modify Games

July 8, 2014

One thing that’s pretty interesting is the level to which tabletop RPGs expect the groups playing them to hack and change them.  There’s pretty much three ways to address that, and it’s worth considering how those work:

“Design it yourself” (AKA anti-design)

This would be “Rule 0” or “if you don’t like these rules, make up your own”.  This has been in rpgs for decades, and specifically is a non-helpful way to do things.  It also gets used a lot as a defense when the rules didn’t do what they’d claim they were supposed to do.

I’ve heard arguments in the past “no, no, really the game was made poorly to FORCE you to become a better gamer”… which… is ridiculous.  You might make that argument about certain forms of training and the need to ramp up difficulty, but in games?  Difficulty should be about a challenge factor in playing the game, not a challenge to making the game work – the challenge in basketball is playing basketball, not tying people’s legs together and blindfolding them and giving them a half-flat ball.


This is pretty much the easiest way to make an easy-to-hack game, though the changes are all effectively superficial.  The Pool, HeroQuest, octaNe, Universalis, Fudge, FATE…  all of these games have easy to swap labels while keeping the mechanics identical.  What’s the difference between a heavy armored robot and a small fast robot?  The labels “Heavy” “Armored” vs. “Small” “Fast” applied to them.

On the other hand, while this allows easy genre and element swapping, it doesn’t actually make for any mechanical differences in hacking by itself.  Often these systems use a universal resolution and there’s not a lot more to hang on it.

Unified Principles

Vincent Baker has a nifty chart about how rules work vs. how we actually play at the table. (you should read the whole post, it’s almost 9 years old but still very relevant).

This is an idealized design-to-play set up.  The turquoise is the actual rules in the game book/text.  You’ll notice it’s like 98% within the other circle – which is “how we play at the table”.  In other words, nearly all of the rules or advice in the book are actually useful in play.   You’ll notice of “How We Actually Play”, the part that isn’t covered by the game text circle, is broken into two sections – Ad Hoc decisions, and Principled Decisions.

Principled decisions are the choices you make because they fit in with the principles the game has communicated to you on how things generally work in the game.

Say you’re playing D&D and there’s a gas explosion in a mine.  “Oh gee, the book doesn’t have rules for that… but it’s a LOT LIKE a Fireball or Dragon’s Breath.  Let’s say it does this much damage and you make a Saving Throw for half damage”.   You can make that call and it works with the game because it’s based on common principles within the game.

So part of a game design is how well you communicate the principles of your game – which doesn’t have to be pages of theory and designer notes, but it does need to be consistent across your rules.

Hacking a game and considerations

So, from the other side of this, as a gamer, it’s worth looking at what you need to consider when hacking a game, which then shapes things people should think about when we’re talking about game design for modification.


How much does modding the game risk throwing things way out of wack?  Some games are pretty open to tossing stuff in without too much trouble.  Some games are an exacting system of currency and bumping it around breaks things quickly.  If you are designing a game with the intent for people to modify it, you need to try to aim for less fragility and be clear about which parts are more/less able to be fiddled with before breaking.


How much effort does it take for me to hack things into your game?  Is it a few minutes or nearly hours of trying to put together numbers?  Less crunch makes easier hacking.

Complex Combinations

How many OTHER rules do I have to think about this hack intersecting with?  Are there ways it could do things very different than what I intend because of a complex interaction?

Existing Tools

What’s already in the game as rules that I could use right now or with the most minimal changes?  Why should I use one as opposed to another?  (Example: when should you make this an attribute roll vs. when should you make it a skill roll?)


What are general goals of the rules?  What are rules they follow over and over?  What sorts of things work against this?

So, here’s an example: Apocalypse World has a principle in the rules of “Soft Moves” vs. “Hard Moves”.  A Soft Move is something the GM does that warns about trouble coming, a hazard or a threat.  A Hard Move is something the GM does that has consequences, now and lasting – whether that’s injury, an NPC getting killed, etc.   A core principle is that you have to give warning, you have to go through at least one Soft Move and give the players a chance to act, before jumping into Hard Moves.

Because this is generally a good principle for a lot of action games (forewarn danger, then enact danger) as opposed to “gotcha!” traps, it is something people caught on to and find AW very hack-able to many other things (Monsterhearts, Dungeon World, etc.).

On the flipside, the D20 flood of bad materials is a pretty good example of what happens when you have both high complex combinations and little in the way of principles communicated.



Tactical RPGs and Stunt Systems

July 1, 2014

Had a pretty great conversation with Quinn Murphy about tactical rpgs yesterday and I figured I wanted to put my thoughts about stunt systems down to share.

Key Ideas

First, it’s important to note a couple of key things that shape how you do stuff in tabletop rpgs.

Tactical RPG Pitfalls

There’s basically two major pitfalls for a lot of rpgs that want to do tactical play.   One sits on the player’s side, the other on the GM’s side.

“All the fun choices happened before we started playing…”

What many games do, is front load all the interesting tactical choices into character generation.  You pick a “build” and you do the same thing,  over and over, maybe with only 1-2 really viable options, and everything else just being a bad idea to do outside of very specific circumstances.  Much like tic-tac-toe, this isn’t that interesting to play at all.  What you want is tactical choices to be made during play, not before it.

“It’s so hard to come up with interesting outcomes… I have to prep so much every week.”

The other issue is that a lot of games make improvisation rather difficult, requiring the GM to prep a lot ahead of time.  If there’s not a lot of tactical choices to make during play, it’s basically the players’ front loaded choices vs. the GM’s frontloaded encounter choices, requiring a lot of fiddling and guesswork to how it’s going to turn out.

Of course, this is why fudging becomes a common technique for dealing with this.  You either become a game design master for balancing things OR you just lie to make up the difference because you couldn’t pick out the obscure math relationships at play…

A good stunt system, properly applied, handles both of those problems easily.

Stunt Systems

Let’s be very clear that when I’m talking about stunt systems, I’m talking about mechanics that help you take unexpected actions and translate them into mechanics that will engage a tactical rules set.   There’s plenty of games that don’t have any real tactical heft, and the players can describe their actions in whatever way without it making any real mechanical difference (Primetime Adventures, Wushu, HeroQuest, FATE, etc.).

When you have a good stunt system, players can improvise actions in play that help as well or better than any choices they’ve made during character generation.  Being able to improvise means the GM can run with a lot less statistics needing to be prepped – it becomes easy to set up environmental hazards, have monsters do specialized attacks without statting up everything, etc.

Good stunt systems have to be flexible, easy to improvise without too much work.  Here’s some examples worth considering:

Tunnels & Trolls “Saving Throws”

Tunnels & Trolls Saving Throws is one of the oldest and easiest to understand.  You say what you want your character to do – the GM assigns a “Saving Throw Level” based on how tough it sounds, and which stat uses it.  You roll and succeed or fail.  If the GM wants to have a monster improvise something, or have an environmental hazard?  She simply assigns a Saving Throw Level and stat as usual.  What T&T did well was explicitly laying out this logic in the book, and not leaving it as something to be orally passed down, as a lot of D&D’s “Rulings Not Rules” kind of play did.

D&D 4E’s Page 42

When 4E was first in development and folks were just getting bits and rumors, the design tidbit that eventually became pg. 42 had me quite excited.  I remember it was someone detailing how if you wanted to stunt by kicking a table out from under a monster standing on it, you’d roll a Strength attack vs. a Reflex Defense, and how basically every stunt could be built out of Attribute vs. Defense depending on what made sense.

Now, if you put folks in a stunt-rich environment, a whole lot of options open up from this.  It also provides other viable options than your usual 4E power set.  A lot of 4E play suffered from dragged out fights because usually the 2nd half of any given encounter the players had played out their interesting options, and then it was just a lot of back and forth with basic attacks, over and over.  Aside form reducing enemy HP, the other half is giving players plenty of stuntable environments to work in.

Riddle of Steel’s Terrain Rolls

Riddle of Steel’s main tactical element is managing your Combat Pool – you spend dice out of it, and the focus of play was making sure to spend enough to succeed and/or not get hammered, but at the same time, you didn’t want to spend out too much and leave yourself open for a counter attack.  The Terrain rolls simply required players to choose how many dice they would gamble on the side in order to do useful things like: maneuver around multiple opponents so you don’t get rushed, avoid falling prone on rough ground or up/down stairs, etc.

Pretty much anything you wanted to do during combat, could be given as a Terrain roll in addition to your normal skills.

Rulings Not Rules

Things like the Primer for Old School Gaming shoves all the responsibility into the GM’s hands: make it up.  The advantage here is that it’s mechanically simple (GM fiat), but it also means the GM has to think about things and consider what is fair given the system pretty often.  Despite being mechanically easy, this is not actually necessarily newbie-friendly – it actually depends on the GM’s system mastery of everything else in the game, and their ability to communicate well with the group.


Differentiation: Mechanical vs. Fictional

March 17, 2014

There’s an idea I think about a lot when it comes to RPGs – “How does THIS particular game teach you to take the in-game events and turn them into mechanics?  How does it teach you to take the mechanics and make events in play?” (see also: Vincent’s “How RPG Rules Work” or  Quinn Murphy’s “Thingification” idea)

When a game makes this easy to understand and smooth for the group playing, it’s really easy to take anything that happens in play, and figure out how it fits with the mechanics, or to take anything the mechanics kick out and make it an event in the imaginary events.   (As I tend to say, the easiest rule is “I say this happens, and then it does”).  When this is complex, challenging, or just poorly explained, you end up having lots of problems for groups – you can see a lot of this around D&D 4E’s mechanics and the complaints with it.

Anyway, that translation process either differentiates the “things” in the game by fiction, or by mechanics and it’s important because these are two very different ways to do a game.

Differentiation by Fiction

So, let’s say we’ve got a game, and it’s got dragons: ice dragons and fire dragons.

Here’s one example:

Fire Dragons: 20 hitpoints, 1d6 fire damage per round

Ice Dragons: 20 hitpoints, 1d6 ice damage per round

So, in this example, the dragons are basically the same, and the only difference is we’ve swapped one word for another.  The only thing that makes them different is fiction based – we can describe that ice attacks work different than fire attacks, but there’s nothing mechanically that necessarily makes them that much different.

This is differentiation by fiction.

You can see some games are built entirely on this idea – Primetime Adventures, Inspectres, HeroQuest, FATE games, Universalis, etc.  The mechanics are universally identical, you just swap descriptor words to change one to another.

It’s really useful if you want a game where the mechanics are less deep in specifics, easier to get system fluency to make good choices in play, and it’s very easy to improvise or adapt – just swap some labels for another and there you go.

Differentiation by Mechanics

Now let’s try a different version:

Fire Dragons: 25 hitpoints, 1D6 fire damage a round, and target takes 2 more fire damage until doused

Ice Dragons: 20 hitpoints, 1D6 ice damage a round, and target is frozen in place for 2 rounds.

You can see here, that the dragon types work a little different and probably will require different ways to deal with in play.  Specific mechanics result in specific differences, result in specific tactics.

This is differentiation by mechanics.

Most traditional rpgs use this – specific class/race powers, feats, a power tree, spells – things that give you access to mechanics or methods of play that you can’t get using the other options.

It’s really useful if you want a game where having deeper system familiarity is more important, where tactics or strategy might be useful, or if you want to push play choices in differing directions (“Doing X feels completely different than doing Y”).

What it means for play

The reason to look at this is that these are two very different ways to come at a game, and usually gets subsumed into some vague talk about “crunch”* (which has about 3-4 other factors piled into it as well).

With Differentiation by Fiction, you don’t have to think about mechanics much, but then the only thing you have to really push play, to make one thing behave functionally different than another is judgment calls at the table and agreement of the group (“Well, I’ve got Big Sword 4, but since we’re fighting underwater, it should have a penalty or not be useable at all.”).

With Differentiation by Mechanics, different choices matter much more, but it also means players need to develop some familiarity and fluency with the system to make that happen.  It means a longer time spent learning the options, and trying to piece together why things work and what strategies work best.  It also becomes a lot harder to improvise things if you find yourself having to suddenly come up with new rules or slam together systemic differences on short notice.

The way these two design philosophies operate is so different that players who prefer one type will not find systems that do the other worth touching at all.    So this becomes important to know if you like only one type, if your group is a mix, or if you are designing a game.

(*One might think that fiction-differentiation automatically means “low crunch” but that’s not necessarily the case.  Universalis, for example, has a good amount of procedural steps, and the latest version of HeroQuest also has quite a bit of crunch options that go along with it as well.  Contrary, there’s also a few mechanical differentiation games that have relatively low crunch, such as basic D&D and many of it’s clone rpgs .)