Open this document in the playground · Download source
Fate Core — one invocation, +2 or reroll?
R14 · Advanced · Scoped component
Question and scope
In Fate Core you have one available fate point and a relevant aspect whose invocation is agreed. Against fixed passive opposition +2, your skill is +2. After seeing 4dF, choose +2 or reroll all dice. The objective is strictly exceed opposition without a cost, not merely obtain a tie. Predict which initial rolls need a reroll rather than a guaranteed +2 rescue.
Source sections are linked under Rules below; primary-source review is recorded in the release dossier. Independent human and browser/usability review are separate gates.
Inputs and model limits
Integer skill and fixed passive opposition, default both +2; one available paid invocation and an applicable agreed aspect. Objective is strict success without cost.
Complete runnable model
A policy is a function from information already seen to a decision. Keep an already successful roll without spending. If +2 guarantees strict success, choose it. Otherwise reroll and accept the replacement. The callback may inspect the potential reroll only after the decision to reroll has been made.
skill = 2
opposition = 2
f = die_faces([-1, 0, 1])
roll = f + f + f + f
bands = scale().step("Objective missed").step("Strict success")
def policy(first, replacement):
if first > opposition:
return "Strict success"
if first + 2 > opposition:
return "Strict success"
if replacement > opposition:
return "Strict success"
return "Objective missed"
result = joint_classify(roll + skill, roll + skill, bands, policy)
output("One-invocation policy", result)
output("Strict success chance", result.pmf("Strict success"))
output("Chance of spending the point", (roll + skill).cdf(opposition))
One-invocation policy · Outcomes
| outcome | % | frac | X/1076 |
|---|---|---|---|
| Objective missed | 11.4 | 123/1076 | 123 |
| Strict success | 88.6 | 953/1076 | 953 |
Strict success chance · Prob
| outcome | % | frac | X/1076 |
|---|---|---|---|
| Strict success chance | 88.6 | 953/1076 | 953 |
Chance of spending the point · Prob
| outcome | % | frac | X/1076 |
|---|---|---|---|
| Chance of spending the point | 61.7 | 50/81 | 664 |
Reading the report and independent checks
Initial totals with dice ≥−1 are already successful or can be rescued by +2: 66/81. The remaining 15/81 need a reroll, which strictly succeeds 31/81 of the time. Thus success is 5811/6561. A point is spent on the original non-successes: 50/81. Ties may achieve an overcome goal at a minor cost, but deliberately do not meet our stricter objective.
Safe changes
Change skill to +3. The policy now rerolls only original dice −3 or −4. State the objective and budget again before comparing probabilities.
Composition and boundary checks
Could “look at the replacement, then decide whether to take +2” do better? Yes, but it is illegal future information in this model. Answer check: at the original skill, always taking +2 after any miss succeeds only 66/81; the fixed policy adds the reroll rescue term, not hindsight.
Do not combine separate marginal reports and assume they preserve the same rolled faces. When an event depends on several properties of one roll, keep those properties together inside a single classifier or pool callback.
Rules
Fate Core: Invoking Aspects and Overcome. One paid invocation, no free invokes, defender responses, later resources, or further invocations. Relevant aspect and point availability are assumptions, not outcomes.
Related learning
Lesson L30 explains the modelling idea step by step. Cookbook index offers both game and mechanic views of these same recipes.
Content ID: R14 · Review status: pilot
Prerequisites: L30