Section 297 of 440

Complete canonical tutorial. This reader section contains the same teaching body as PWR-115 · Probabilistic reasoning. Open the Power dossier.

PWR-115 · Reasoning, Judgement & Strategy

Convert benign events to frequencies, forecast, and score every outcome

This tutorial teaches the full forecast-record cycle on ten benign fictional events. Define the event and reference class, convert favourable cases to a natural frequency, lock a probability, reveal the staged outcome, calculate the Brier score, preserve every row, inspect the direction of the largest errors and write one method update for a fresh set. The exercise is not financial, medical or safety-critical forecasting.

Start the tutorial Open the canonical Power page All full tutorials

Tutorial curriculum · current public edition · Revision 7T · Full Tutorial Edition

Revision 7T · Full Tutorial Edition · complete governed practice tutorial

Convert benign events to frequencies, forecast, and score every outcome

Authority and safety

Use only the fictional or mechanical events below; do not adapt the exercise to gambling, finance, medicine, safety, or another consequential decision.
Do not alter an event definition, probability, or resolution after opening its outcome.
Eight events are a calculation exercise, not a stable estimate of a person's calibration or judgement.

Exact materials

The ten benign event rows below
A natural-frequency and probability worksheet
The staged outcome-and-calculation key, kept closed until all ten probabilities are locked
A method-review sheet with largest overforecast, largest underforecast, counter-reason and next-set change

Set up in this order

Copy each exact event and reference class. Do not replace it with a vague event such as ‘soon’ or ‘likely.’
Write favourable outcomes out of total equally likely outcomes before converting to a decimal probability.
For each row, write one counter-reason: a way the event could fail even when it seems plausible, or occur even when it seems unlikely.
Lock all ten probabilities before opening any staged outcome.
Reveal one outcome at a time and calculate Brier = (probability - outcome) squared. Preserve every forecast exactly as locked.
Add all ten Brier scores and divide by ten. Keep the individual rows; the mean does not explain the direction of an error.
Identify the largest overforecast and largest underforecast, then write one narrow method update such as checking denominators before a fresh set.
Use a new benign ten-event set for any later review. Do not edit, delete or recycle the resolved rows.

Supplied fixture

All elementary outcomes in each row are equally likely. Lock all ten forecasts before opening the staged outcome key.
Event id	Event	Reference class	Learner natural frequency	Learner probability before reveal	Learner Brier after reveal
E1	Draw a blue tile	Bag: 3 blue, 7 white; one draw	[record]	[lock]	[calculate]
E2	Roll an even number	One fair six-sided die	[record]	[lock]	[calculate]
E3	Draw a star card	20 cards: 5 stars, 15 circles	[record]	[lock]	[calculate]
E4	Land on green	8 equal sectors: 1 green, 7 orange	[record]	[lock]	[calculate]
E5	Draw a red bead	Jar: 18 red, 12 yellow	[record]	[lock]	[calculate]
E6	Get at least one head	Two fair flips: HH, HT, TH, TT	[record]	[lock]	[calculate]
E7	Draw a prize ticket	100 tickets: 2 prize, 98 non-prize	[record]	[lock]	[calculate]
E8	Get three heads	Three fair flips: 8 ordered outcomes	[record]	[lock]	[calculate]
E9	Roll a 1	One fair ten-sided die numbered 1–10	[record]	[lock]	[calculate]
E10	Draw a vowel card	10 cards: A, E, I and seven consonants	[record]	[lock]	[calculate]

Check your work only after you finish

Lock every response first. The checking page is separate so the answers are not exposed inside this exercise.

Open the separate checking page →

Scoring rule

Award 1 point for each correct natural frequency, probability and Brier score across ten rows: 30 points. Add 1 for the correct sum, 1 for the correct mean, 1 for the largest overforecast, 1 for the largest underforecast, 1 for a counter-reason and 1 for a concrete fresh-set method update: 36 total. Deleting or editing a locked forecast makes the aggregate unscorable.

Worked example

For E6, list HH, HT, TH and TT as four equally likely outcomes.
Count HH, HT and TH as favourable: 3 of 4, so lock 0.75.
Open the staged outcome: E6 resolves 1. Calculate (0.75 - 1)² = 0.0625.
After all ten rows, identify error direction rather than calling every large Brier score the same kind of mistake.
Write one next-set update, such as checking the complete outcome list before converting counts to a decimal.

Correct result: E6 has a locked probability of 0.75 and Brier 0.0625. The complete method retains all ten rows and uses error direction to improve a later fresh set.

Right and wrong

Right and wrong comparison
Moment	Right / safer	Wrong / riskier	Why
Defining E6	At least one head within exactly two flips.	A head will appear soon.	The wrong event lacks a trial boundary.
Counting E6	Use 3 of 4 ordered outcomes.	Use 1 of 2 because there are two symbols.	The event concerns two-flip sequences.
After reveal	Score the locked probability.	Change 0.75 to 1.00.	Post-outcome editing destroys accountability.
Interpreting the mean	It summarises these eight rows only.	It certifies expert financial forecasting.	The fixture is small and non-consequential.

Common mistakes and fixes

Common mistakes and corrections
Mistake	Fix
Using the wrong denominator.	List all equally likely outcomes before counting favourable ones.
Writing a percentage without a reference class.	Write counts first, such as 5 of 20.
Deleting badly resolved forecasts.	Preserve E1-E8 and divide the full sum by eight.
Calling one low score skill.	Report event type, sample size, scoring rule, and narrow boundary.

Evidence boundary

What the tutorial may—and may not—claim.

Supportable claim

Probabilistic reasoning can improve through representation training and configured forecasting practice, with gains bound to problem form and system design.

Measurement boundary

Declare event set, base rates, elicitation format, calibration, resolution, scoring rule, aggregation, missing outcomes and decision consequences.

Myth

Predict the future with certainty

Metric

Calibration, resolution and proper score on a predeclared event set

Boundary

Good forecasts in one domain do not establish foresight, certainty or safe high-stakes authority.

Negative and limiting findings

Frequency format did not add the expected benefit after nested-set training.
Tournament performance intertwined training, teams, aggregation, selection and attrition.

Prohibited wording

predict the future
become a human probability engine
Bayesian training eliminates uncertainty
forecast score proves wisdom

Direct source register

Evidence that supports—and limits—the route.

The psychology of intelligence analysis: Drivers of prediction accuracy in world politics

Barbara Mellers; Eric Stone; Pavel Atanasov; Nick Rohrbaugh; S. Emlen Metz; Lyle Ungar; Michael M. Bishop; Michael Horowitz; Ed Merkle; Philip Tetlock · 2015 · PRIMARY_RESEARCH

Supports only the bounded empirical proposition in the declared configurations. Constrains generalisation, transfer, certainty or efficacy; it is not optional context.

How to train your Bayesian: A problem-representation transfer rather than a format-representation shift explains training effects

Miroslav Sirota; Lenka Kostovičová; Frédéric Vallée-Tourangeau · 2015 · PRIMARY_RESEARCH

Supports only the bounded empirical proposition in the declared configurations. Constrains generalisation, transfer, certainty or efficacy; it is not optional context.

Tutorial delivery controls

Learn, adapt, troubleshoot and resume

Estimated timeEstimated 18 min reading and worksheet pass

DifficultyIntroductory

EquipmentBasic stationery or digital tools

SpaceDesk / seated

Method qualityEstablished10 of 10 structural checks present. Automated method-readiness band; human editorial sign-off is separate.

Evidence contextG1; Focused research depthScientific support is evaluated separately from teaching-method structure.

Editorial reviewPending manual sign-offNo human approval is claimed until reviewer, date and content hash are recorded.

Your tutorial progress0 of 8 steps complete

0 of 8 steps complete

Download learner worksheet

Progress is saved only in this browser on this device.

Step-by-step learner mode

Each activity includes its success check, a nearby accessible alternative and an “I’m stuck” correction path. Alternatives preserve the target where possible; when they change the task, Titan labels them as related rather than equivalent.

Copy each exact event and reference class

Copy each exact event and reference class. Do not replace it with a vague event such as ‘soon’ or ‘likely.’

Why this step exists

“Copy each exact event and reference class” creates the auditable value, condition or statement needed for the tutorial’s later comparison and conclusion.

Success check

A legible entry directly completes “Copy each exact event and reference class” and records the requested condition, decision or boundary without adding a new claim.

I’m stuck on this step

Reset: Re-read this authored instruction — “Copy each exact event and reference class. Do not replace it with a vague event such as ‘soon’ or ‘likely.’” — and its success check, then attempt only this step.

Possible snag: Writing a percentage without a reference class.
Correction: Write counts first, such as 5 of 20.
Possible snag: Calling one low score skill.
Correction: Report event type, sample size, scoring rule, and narrow boundary.

Stop / get help: A governed step-by-step lesson may be completed independently inside its stated limits. Stop if the declared configuration cannot be maintained, uncertainty becomes material, adverse effects appear, or qualified authority is required.

Mark step 1 complete

Write favourable outcomes out of total equally likely outcomes before converting to…

Write favourable outcomes out of total equally likely outcomes before converting to a decimal probability.

Why this step exists

“Write favourable outcomes out of total equally likely outcomes before converting to a decimal probability” creates the auditable value, condition or statement needed for the tutorial’s later comparison and conclusion.

Success check

A legible entry directly completes “Write favourable outcomes out of total equally likely outcomes before converting to a decimal probability” and records the requested condition, decision or boundary without adding a new claim.

I’m stuck on this step

Reset: Re-read this authored instruction — “Write favourable outcomes out of total equally likely outcomes before converting to a decimal probability.” — and its success check, then attempt only this step.

Possible snag: Using the wrong denominator.
Correction: List all equally likely outcomes before counting favourable ones.

Mark step 2 complete

For each row, write one counter-reason: a way the event could fail…

For each row, write one counter-reason: a way the event could fail even when it seems plausible, or occur even when it seems unlikely.

Why this step exists

“For each row, write one counter-reason: a way the event could fail even when it seems plausible, or occur even when it seems unlikely” creates the auditable value, condition or statement needed for the tutorial’s later comparison and conclusion.

Success check

A legible entry directly completes “For each row, write one counter-reason: a way the event could fail even when it seems plausible, or occur even when it seems unlikely” and records the requested condition, decision or boundary without adding a new claim.

I’m stuck on this step

Reset: Re-read this authored instruction — “For each row, write one counter-reason: a way the event could fail even when it seems plausible, or occur even when it seems unlikely.” — and its success check, then attempt only this step.

Possible snag: The result from “For each row, write one counter-reason: a way the event could fail even when it seems plausible, or occur even when it seems unlikely.” does not yet meet this declared check: A legible entry directly completes “For each row, write one counter-reason: a way the event could fail even when it seems plausible, or occur even when it seems unlikely” and records the requested condition, decision or boundary without adding a new claim.
Correction: Return to the start of “For each row, write one counter-reason: a way the event could fail…”, reduce complexity or pace, and repeat only the part needed to satisfy: “A legible entry directly completes “For each row, write one counter-reason: a way the event could fail even when it seems plausible, or occur even when it seems unlikely” and records the requested condition, decision or boundary without adding a new claim.”

Mark step 3 complete

Lock all ten probabilities before opening any staged outcome

Lock all ten probabilities before opening any staged outcome.

Why this step exists

“Lock all ten probabilities before opening any staged outcome” carries the authored Probabilistic reasoning method into a reviewable output without adding an unstated task, dose or claim.

Success check

Evidence of completion shows the learner followed this exact instruction without adding an unstated step: “Lock all ten probabilities before opening any staged outcome.”

I’m stuck on this step

Reset: Re-read this authored instruction — “Lock all ten probabilities before opening any staged outcome.” — and its success check, then attempt only this step.

Possible snag: The result from “Lock all ten probabilities before opening any staged outcome.” does not yet meet this declared check: Evidence of completion shows the learner followed this exact instruction without adding an unstated step: “Lock all ten probabilities before opening any staged outcome.”
Correction: Return to the start of “Lock all ten probabilities before opening any staged outcome”, reduce complexity or pace, and repeat only the part needed to satisfy: “Evidence of completion shows the learner followed this exact instruction without adding an unstated step: “Lock all ten probabilities before opening any staged outcome.””

Mark step 4 complete

Reveal one outcome at a time and calculate Brier = (probability -…

Reveal one outcome at a time and calculate Brier = (probability - outcome) squared. Preserve every forecast exactly as locked.

Why this step exists

“Reveal one outcome at a time and calculate Brier = (probability - outcome) squared” creates the auditable value, condition or statement needed for the tutorial’s later comparison and conclusion.

Success check

The work shows the value requested by “Reveal one outcome at a time and calculate Brier = (probability - outcome) squared”, retains the supplied units or signs and leaves any unsupported value unknown.

I’m stuck on this step

Reset: Re-read this authored instruction — “Reveal one outcome at a time and calculate Brier = (probability - outcome) squared. Preserve every forecast exactly as locked.” — and its success check, then attempt only this step.

Possible snag: Deleting badly resolved forecasts.
Correction: Preserve E1-E8 and divide the full sum by eight.

Mark step 5 complete

Add all ten Brier scores and divide by ten

Add all ten Brier scores and divide by ten. Keep the individual rows; the mean does not explain the direction of an error.

Why this step exists

“Add all ten Brier scores and divide by ten” isolates the named evidence distinction so the learner can interpret Probabilistic reasoning without extending the claim.

Success check

The output directly answers “Add all ten Brier scores and divide by ten”, names the relevant distinction and stays within the supplied record.

I’m stuck on this step

Reset: Re-read this authored instruction — “Add all ten Brier scores and divide by ten. Keep the individual rows; the mean does not explain the direction of an error.” — and its success check, then attempt only this step.

Possible snag: The result from “Add all ten Brier scores and divide by ten. Keep the individual rows; the mean does not explain the direction of an error.” does not yet meet this declared check: The output directly answers “Add all ten Brier scores and divide by ten”, names the relevant distinction and stays within the supplied record.
Correction: Return to the start of “Add all ten Brier scores and divide by ten”, reduce complexity or pace, and repeat only the part needed to satisfy: “The output directly answers “Add all ten Brier scores and divide by ten”, names the relevant distinction and stays within the supplied record.”

Mark step 6 complete

Identify the largest overforecast and largest underforecast, then write one narrow method…

Identify the largest overforecast and largest underforecast, then write one narrow method update such as checking denominators before a fresh set.

Why this step exists

“Identify the largest overforecast and largest underforecast, then write one narrow method update such as checking denominators before a fresh set” creates the auditable value, condition or statement needed for the tutorial’s later comparison and conclusion.

Success check

A legible entry directly completes “Identify the largest overforecast and largest underforecast, then write one narrow method update such as checking denominators before a fresh set” and records the requested condition, decision or boundary without adding a new claim.

I’m stuck on this step

Reset: Re-read this authored instruction — “Identify the largest overforecast and largest underforecast, then write one narrow method update such as checking denominators before a fresh set.” — and its success check, then attempt only this step.

Possible snag: The result from “Identify the largest overforecast and largest underforecast, then write one narrow method update such as checking denominators before a fresh set.” does not yet meet this declared check: A legible entry directly completes “Identify the largest overforecast and largest underforecast, then write one narrow method update such as checking denominators before a fresh set” and records the requested condition, decision or boundary without adding a new claim.
Correction: Return to the start of “Identify the largest overforecast and largest underforecast, then write one narrow method…”, reduce complexity or pace, and repeat only the part needed to satisfy: “A legible entry directly completes “Identify the largest overforecast and largest underforecast, then write one narrow method update such as checking denominators before a fresh set” and records the requested condition, decision or boundary without adding a new claim.”

Mark step 7 complete

Use a new benign ten-event set for any later review

Use a new benign ten-event set for any later review. Do not edit, delete or recycle the resolved rows.

Why this step exists

“Use a new benign ten-event set for any later review” isolates the named evidence distinction so the learner can interpret Probabilistic reasoning without extending the claim.

Success check

The output directly answers “Use a new benign ten-event set for any later review”, names the relevant distinction and stays within the supplied record.

I’m stuck on this step

Reset: Re-read this authored instruction — “Use a new benign ten-event set for any later review. Do not edit, delete or recycle the resolved rows.” — and its success check, then attempt only this step.

Possible snag: The result from “Use a new benign ten-event set for any later review. Do not edit, delete or recycle the resolved rows.” does not yet meet this declared check: The output directly answers “Use a new benign ten-event set for any later review”, names the relevant distinction and stays within the supplied record.
Correction: Return to the start of “Use a new benign ten-event set for any later review”, reduce complexity or pace, and repeat only the part needed to satisfy: “The output directly answers “Use a new benign ten-event set for any later review”, names the relevant distinction and stays within the supplied record.”

Mark step 8 complete

Correct versus incorrect execution

These accessible process diagrams are built from the tutorial’s own right/wrong teaching. They are not anatomical illustrations and do not add technique beyond the canonical tutorial.

Defining E6 — The wrong event lacks a trial boundary.

Correct / safer

At least one head within exactly two flips.

Wrong / riskier

A head will appear soon.

Counting E6 — The event concerns two-flip sequences.

Correct / safer

Use 3 of 4 ordered outcomes.

Wrong / riskier

Use 1 of 2 because there are two symbols.

After reveal — Post-outcome editing destroys accountability.

Correct / safer

Score the locked probability.

Wrong / riskier

Change 0.75 to 1.00.

Interpreting the mean — The fixture is small and non-consequential.

Correct / safer

It summarises these eight rows only.

Wrong / riskier

It certifies expert financial forecasting.

Method-structure checklist

10 of 10 structural checks present

Ordered, Power-specific instructions — present
Every activity has a success check — present
Materials or supplied records are declared — present
Measurement or assessment rule is present — present
Tutorial-specific troubleshooting is present — present
Stopping or escalation boundary is present — present
Every activity has an adjacent alternative — present
Correct-versus-incorrect comparison is present — present
Evidence context is bound to the Power record — present
Planning metadata is present — present

The method-readiness band and presence checklist assess tutorial presentation and are separate from evidence quality for the underlying Power. They are automated editorial aids, not human approval.

Manual editorial sign-off: Pending. This tutorial must not display a human-approved state until an identified editor signs the exact content hash.