Having a background in statistics, I like evaluating the mechanics of roleplaying game systems. I have previously blogged how 3d6 is not less swingy than d20 comparing different types of systems at a high level.
This post is a more detailed dive into Fate Core (the 4th edition), which has a bell curve based dice result distribution.
Fate Core is an open source roleplaying game, available under both OGL and Creative Commons licences, with several published books (available both printed and as nicely formatted PDFs). The Fate Accelerated Edition (FAE) version is a short (only 40 pages) and lightweight variant of the system, with the recent Fate Condensed a streamlined version of the full system with clarifications and minor updates such as a safety tools section. (All are considered part of the same edition.)
There are several highly rated settings that use different versions of Fate, including Dresden Files, Fate of Cthulhu, Disapora, and Spirit of the Century.
Dice probability distribution
Fate, which is based on an earlier open gaming system called Fudge, uses special dice, a regular dice shape (cube) with two + (plus) faces, two blank faces, and two - (minus) faces. You roll 4 dice, and add them up, each giving a +1, 0, or -1, generating a result between -4 and +4. This can also be expressed as 4d3-8.
You can also use regular dice (1-2 counts as -1, 3-4 as 0, and 5-6 as +1), or a special deck of cards with the same probability distribution.
You add the dice roll to your skill rating (usually in the +0 to +4 range), with the options of various bonuses (usually +2), and compare to either a difficulty or an opposing roll.
The dice distribution can be generated at https://anydice.com
F: {-1, 0, +1}
output 4dF named "Bell curve 4dF distribution"Reminder: there is no swingyness
As discussed in my earlier post, for a given roll the chance of success or failure (or other outcomes) usually is a binomial probability.
e.g. If you need a +3 to succeed, then you have a 6% chance, and the variance (or standard deviation) of a 6% chance is the same, irrespective of whether you are rolling 4dF or a flat d100.
What is useful when analysing, is not the statistical swingyness of statistical results, but what the effect is of modifiers and other changes in difficulty. In other words, how the chance of success varies by skill level, difficulty, and modifiers.
Success probabilities
A test has one of four results: succeed with style, succeed, tie, or fail (which could still be a success at a major cost).
The following chart shows how the probability of each result varies based on skill level, against a standard difficulty of Fair (+2).
The dashed line shows the effective chances of being able to get a Tie, or better, if you are in a position to invoke an aspect after you have rolled.
The +4 to -4 range of 4dF, along the bell curve, gives a good range of success probability points, with 9 possible probability steps (plus 0% and 100%).
At the ends it has steps of 1% — 5%, with the middle having larger steps of 10% — 20%, because while the different between 1% and 6% is a lot, the difference between 61% and 66% is relatively less discernible.
Outcomes
The system has a fairly streamlined categorisation of four different actions, with clear definitions of minor costs and benefits due to succeed with style.
It has a narrative system of aspects, powered by fate points, where success at a style usually results in a boost — a free one use temporary aspect, essentially a +2 on the next roll, if it can be appropriately included narratively.
The chance of outright success at an average (Fair) task for an average (Fair) skill is only 38%, but when you include success-at-a-cost ties, that becomes 62% (but with a cost).
Note that where appropriate, even a failed roll can be awarded a successful outcome by the GM, but at a major cost, so the chance of 'success' is 100% where needed, such as where clues are important for the story to progress. The test doesn't determine if you get the clue, but what the cost of it is.
Degree of success
Where a numerical degree of success is needed, e.g. for damage, the number of shifts between the opposing rolls is used. While this does involve two rolls (attack and then defend) they are opposed, rather than being rolled separately, and the system is not too complicated
The ability to trade off a shift (1 damage) in return for a boost (+2 to the next roll) is a nice trade off, as is the ability to gain a boost when your defend succeeds with style.
Effect of modifiers
Difficulty
Fate is a symmetrical system, in that a change in difficulty, e.g. Fair (+2) to Mediocre (+0), has the same effect as a change in skill, e.g. Fair (+2) to Good (+4), just in opposite directions. This means the same success curve applies to all tests, just shifted as necessary.
Other modifiers
Other modifiers have the same effect (for a given skill level) on success probability as changing difficulty.
After you roll
Invoking aspects can be used to modify results after the roll has been made, getting a +2 (or a re-roll) by spending a fate point on a narratively appropriate aspect.
As this only needs to be done after rolling, it means your effective probability of getting a success — either outright or by invoking an aspect when necessary — is two points higher than your base skill.
This means a player has a lot of ability to modify the outcome through spending fate points, turning a 62% chance of success (for a Fair vs Fair contest) into 94% chance, assuming they are willing to spend fate to alter the narrative.
Earning tokens
While compels often rely on the Game Master, it is possible to self-compel, to build up fate points for later use. A player can also gain fate points by conceding a contest (especially if they have also taken consequences).
Free invocations on aspects depend on results help, as they are kind of like getting a fate point, but only spendable on that specific aspect. This means that aspects regularly have ways to invoke them, however there are no random triggers for earning tokens.
Strategic options
Like many narrative games there are not a lot of strategic options during conflicts, it comes down to narratively framing your actions to trigger bonuses (which leads, intentionally, to a very narrative description).
The four actions are relatively straight forward to understand when to apply them. Working out some of the results, e.g. is a tie a success or fail for an action, may initially require checking the rules, but there are only 4-5 variations, so can be quickly learned.
A general strategy in narrative games can be to take losing narrative options early on (compels, consequences, conceding) to build up a store of tokens to then use in order to win the final encounter.
There are some nice tips on running the game in how to make an encounter against a major opponent strategically interesting.
Character complexity
For character options, the base Fate Core system has 19 skills, which feels like a good spread. Skill names are fairly straight forward, and even the Accelerate approaches are relatively easy to understand and apply.
There is an explicit emphasis on niche protection during creation, allowing characters to excel at a particular aspect, e.g. if you choose to be the Great at Drive, you can only be at best Good at Stealth, keeping options for other characters to have their own niche.
Narrative aspects are open ended, and part of world building (they are true statements that define parts of the world). Stunts are also free form, but there are suggested patterns to follow, generally a +2 in particular circumstances.
System customisation
While the skills in the base system can be swapped out for a different list of characteristics, e.g. the 7 approaches used in the Accelerated version. Other suggested lists are given in the Condensed and Core books.
There are also other complexity dials you can tweak in the system, to better represent specific game settings.
Conflict alternatives
In evaluating systems I like to see how well they handle alternative goals to simply slugging it out until the last person standing, such as a goal to take the characters hostage, or an attempt to get away.
An escaping villain could be run as a contest of overcome actions, but you might want to make it interesting by having some environment aspects provide a level of danger.
An attempt to kidnap is probably still a conflict, with attack and defend actions, but with taken out resulting in being kidnapped (one of the examples in the book), rather than killed.
You may be able to use an event compel, or offer the chance to concede, bribing players with fate points to end the scene — but be prepared for them, and the dice, to decide the kidnapping does not succeed.
There are also optional rules like countdowns that can help model complex situations.
Scale of play
The game is oriented towards episode style play, with short and longer term arcs.
Action is scene-based, making it easy to split the party when necessary, and with flexible time and movement between scenes.
There are no explicit systems for downtime, but the system is quite customisable so they could be added if desired.
Character development
Character advancement is fairly flat, with end of session milestones focused more on character change, e.g. swapping skills/approaches, trading refresh for stunts, re-writing aspects.
Actual increase in power (adding to skills, refresh, or stunts) occurs on less frequent end of story arc breakthroughs.
Death
Like most narrative systems, Fate easily caters for games where death is not on the table unless specifically decided by a player. Being taken out means you do lose the contest, but not necessarily that you die.
This does depend on the tone of the game, so for a more gritty game it is certainly possible to have death as a possible outcome. Make sure this is clear with the group before you play.
Overall
Fate is a highly rated narrative system, has been refined over time, and is fairly streamlined. All rules are available for free under open gaming licences, and if you want a print copy the Accelerated version is very lightweight and very affordable. Rolls can use custom dice (or a card deck), but can be easily substituted with regular dice (6-sided) if needed.
The explicit inclusion of safety tools in the latest variant, Fate Condensed, is a nice touch.
The mechanics use a bell curve distribution (effectively 4d3-8), which gives a nice spacing of 9 result levels, with 1% - 5% difference at the ends, and 10% - 20% steps in the middle range. Outcomes include a fairly wide band of succeed at a cost (for ties).
The system is straight forward, with symmetric skill/difficulty adjustments, and it easy to work out the likely chances to succeed at an action.
A typical contest falls around the middle range, around 60% chance for some measure of success outcome. Failed tests should still have an interesting narrative, and even a failed roll can give a successful result where necessary, but at a major cost.
Players have a large opportunity to affect narrative direction through plot points and aspects, which can be invoked after a roll is made (meaning a much higher effective chance of success when needed).
Strategic options are limited, with some optional rules for adding small adjustments, but still far less than other games (due to the focus on narrative elements). The game has an explicit mention of character niche protection.
The system skill/attribute list can be tailored for the type of game being run, with different complexity levels. Character death can be either always on-the-table or a purely narrative choice, depending on the tone of the game.